WO1995004428A1

WO1995004428A1 - Apparatus and method for creating and transmitting confidential documents

Info

Publication number: WO1995004428A1
Application number: PCT/US1994/008684
Authority: WO
Inventors: Yitzchak Pomerantz; Emanuel Menczer
Original assignee: Solomon, Hadar; Aliroo Ltd.
Priority date: 1993-08-02
Filing date: 1994-08-01
Publication date: 1995-02-09
Also published as: AU7478594A; JPH09504660A

Abstract

A method for scrambling and descrambling a source image (190) including a first plurality of first image portions (191), the method including generating a scrambled image (figure 17D), by generating a second plurality of second image portions (192), wherein at least one of the second image portions (192) in an individual location in the second plurality includes a first image portion (191) in a different location in the first plurality and a passpartout (196) surrounding at least some of the first image portions (191).

Description

APPARATUS AND METHOD FOR CREATING AND TRANSMITTING CONFIDENTIAL DOCUMENTS

The present invention relates to creation and transmission of confidential documents.

There has long existed a need for secure handling of confidential documents. It is well known to protect information during transmission along unprotected channels by scrambling or enciphering the information upon transmission and unscrambling or deciphering the information upon receipt, using a common confidential secret key which is known both to the transmitter and recipient.

Modern telefax machines such as the NTTFAX-43 of NTT, Japan, offer protection for confidential transmissions by locking confidential transmitted messages in the receiving machine and releasing them only, in response to an appropriate access code which identifies the intended recipient.

Automatic scrambling systems for documents are described, for example, in "A confidential message handling facility for facsimile communication" by Tominaga et al, in Transactions of the Institute of Electronics and Communication Engineers of Japan, Nov. 1982, the disclosure of which is hereby incorporated by reference. In such systems, scrambling is carried out on electrical signals bearing the confidential information. Once the document appears in a hard-copy format, it is no longer protected. U.S. Patent 4,459,611 to Arai describes a method for scrambling and unscrambling of documents, by dividing their area into sub-areas and shuffling the sub-areas according to a random key. The same key is employed for reconstructing the original document in a unscrambling function. In order for unscrambling to take place, the scrambled document is placed on a scanner for image digitization of the scanned image.

U.S. Patent 4,459,611 fails to take into account significant differences between the scrambled image as created by the scrambler and the scanned image, as the result of handling of the hard copy of the scrambled image and/or of scanning errors.

These differences may include the following: 1. Dislocation and rotation of the image due to poor alignment of the scrambled document on the scanner;

2. Shrinkage or expansion of the image due to humidity and temperature effects on the hard copy of the scrambled image.

3. Distortion of the image due to optical errors in copying machines that may have been used to copy the scrambled image.

4. Missing horizontal rows in the image due to loss of synchronization during fax transmission.

5. Loss of data on the scrambled image due to erasing, stains, folds, and tears that may have occurred on the hard copy.

U.S. Patent 4,459,611 suggests that alignment registration marks be employed to ensure precise alignment of the scrambled document on the scanner. This solution is not, however, practical, due, inter alia to the following reasons:

a. The alignment accuracy required for reasonable unscrambling quality is typically 1/250 of an inch, which is not normally realized when loading a document onto office equipment such as a fax machine. b. No solution is provided for errors outside of the vicinity of the alignment registration marks.

c. Loss of any of the registration marks may render unscrambling impossible.

The P.A.S. port system, commercially available from Shrink-Wrap Software Ltd., 32 Hashachar St., Raanana, Israel, provides direct connection between a computer and a fax machine for printing and for scanning.

A facsimile with encrypted hard copy is described by Myers, R. A and Toupin, R. A. (IBM Technical Disclosure Bulletin, Vol. 20, No. 11B, April 1978. Myers and Toupin state that "The present concept comprises simply printing a hard copy without decryption. Thus, at the input station, one would enter in the cryptographic "key", then scan the document, compress it, encrypt it, perform error correction/detection encoding, send it, and print it. Thus, the output would be an unintelligible combination of black and white dots. Once the still encrypted hard copy is printed out, it can be handled relatively casually by uncleared people and even sent by mail. In order to get a "clear" text, one would then take this encrypted text, enter the key into another local facsimile terminal having a decryption feature built in, whereby a clear printed version of the text would be provided. Note that, conceptually, the present method requires no more hardware than conventional facsimile with a cryptographic feature."

US Patent 5,159,630 to Tseng et al is directed to a facsimile message encryption system.

Tseng et al state: "Preferably, the encryption/decryption module 25 is an independent unit that can be connected to a facsimile machine via its input/output ports. (Alternatively, the encryption unit can be integrated into the main control board of a facsimile machine.) Also it is preferred that the encryption unit includes a plurality of user selectable encryption algorithms for providing selection of the level of security during transmission. In practice, for each encryption algorithm, a different pattern of the scrambled information is created. The encryption algorithm, normally is a series of four or more digits, can be as simple as a random number generator or as complicated as a data encryption standard (DES)."

Tseng et al also indicate that "a scrambled message need not be decoded in real time - - nor, online. Instead, a scrambled message can be decoded at a later time by keying-in a decoding key and running the copy mode at the transmitting/receiving facsimile machine. Also, the encrypted message can be decoded by another facsimile machine that has been equipped with the above described security system. Then, by usingthe "copy" function combined with the decoding key, the original message can be recovered (i.e., decoded)."

Tseng et al claim, inter alia, a system for maintaining the security of information which is transmitted between facsimile machines which, among other features, has an encryption/decryption module that "is an independent unit that can be connected to a receiving/transmitting facsimile machine via its input/output ports."

The present invention seeks to provide improved apparatus for handling confidential information in hard copy format.

There is thus provided in accordance with a preferred embodiment of the present invention apparatus for scrambling documents including apparatus for providing output signals representing the contents of a document, apparatus for operating on the output signals to produce modified output signals representing a scrambled version of the document including a system of reference marks to be used during later unscrambling for at least one of registration, scaling, rotation, shifting and defect compensation and generator apparatus receiving the modified output signals and producing a scrambled version of the document.

It is to be appreciated that references to "a system of reference marks" or the like or to "a system of targets" are intended to include, as well, a single distributed reference mark, such as a grid.

There is also provided in accordance with a preferred embodiment of the present invention apparatus for unscrambling documents including a scrambled document signal source, providing signals that represent receiving a document to be unscrambled and providing output signals representing the graphical contents of the document, apparatus for operating on the output signals to produce modified output signals representing a unscrambled version of the document, the apparatus for operating employing said system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation and generator apparatus receiving the modified output signals and producing a unscrambled version of the document.

In accordance with a preferred embodiment of the present invention the apparatus for operating is controlled by a coded input to provide a selected one from a plurality of possible modifications to the output signals.

Additionally in accordance with a preferred embodiment of the invention, the apparatus is embodied in a digital photocopier.

Further in accordance with a preferred embodiment of the invention, the apparatus is embodied in a facsimile machine.

Additionally in accordance with a preferred embodiment of the invention, the apparatus is embodied in a computer.

Additionally in accordance with a preferred embodiment of the present invention the apparatus for operating is operative to change the relative positions of multi-pixel regions of a document, without modifying the information content within each of the multi-pixel regions.

Further in accordance with a preferred embodiment of the present invention, the apparatus for operating is operative to leave unchanged certain predetermined regions of the document.

Additionally in accordance with a preferred embodiment of the present invention, the apparatus for providing includes a scanner receiving a document to be scrambled and providing output signals representing the contents of the document.

Further in accordance with a preferred embodiment of the invention, verification of correct unscrambling is provided. Such verification may be carried out by the use of a registration area in the document or by edge correlation operations on adjacent scrambled pixels.

Throughout the specification and claims, the term "scrambling" is used in an extremely broad sense, to include any suitable reordering of the information in the document. Scrambling may include encryption, but need not involve encryption.

Preferably the reference marks are distributed throughout the area of the document.

Alternatively the reference marks are distributed about the periphery of the area of the document.

In accordance with a preferred embodiment of the present invention, the identification and correct location of each of the reference marks can be determined by its appearance.

Preferably, the reference marks are characterized by a local color modification reversal of the image information thereat.

In accordance with one embodiment of the invention, the reference marks define a frame around a scrambled area.

Preferably, the correct location of each of the reference marks is indicated by the pattern of the frame.

Alternatively, the reference marks are arranged in a predetermined grid.

There is thus provided in accordance with a preferred embodiment of the present invention a method for scrambling and unscrambling a source image including a first plurality of first image portions, the method including generating a scrambled image by generating a second plurality of second image portions, wherein at least one of the second image portions in an individual location in the second plurality includes a first image portion in a different location in the first plurality and a passpartout surrounding at least some of the first image portion.

Further in accordance with a preferred embodiment of the present invention at least one of the portions of the scrambled image contains target information which was not present in the source image.

There is still further provided in accordance with a preferred embodiment of the present invention a method for scrambling a source image while maintaining readability for a designated recipient, the method including scrambling the source image using a scrambling key which is identifiable by applying a confidential transformation to the key identifying information, and generating and outputting a message including the scrambled image and the key identifying information.

Additionally in accordance with a preferred embodiment of the present invention the method also includes automatically extracting the key identifying information from the message and automatically retrieving the scrambling key therefrom.

There is further provided in accordance with a preferred embodiment of the present invention a method for unscrambling a scrambled image to reconstruct a source image including a plurality of image portions, at least one of which is blank, the method including aligning the scrambled image including recognizing at least one target marking on at least one of the blank image portions, computing a correct location of the target marking within the scrambled image, and rubber-sheeting at least a portion of the scrambled image to obtain correspondence between the actual location of the target marking and the correct location thereof.

Still further in accordance with a preferred embodiment of the present invention there is provided a scrambling method including receiving a source image and generating a scrambled image of at least a portion of the source image, providing at least one label image including label information which is not included in the source image, generating an output image including a scrambled image of at least a portion of the source image and the label image.

Also in accordance with a preferred embodiment of the present invention the label information includes unscrambling information. Further in accordance with a preferred embodiment of the present invention the label information includes operator-readable information.

Still further in accordance with a preferred embodiment of the present invention the label information includes alphanumeric information.

Yet further in accordance with a preferred embodiment of the present invention the label information includes machine-readable information.

There is additionally provided in accordance with a preferred embodiment of the present invention an information bearing substrate including a scrambled image of at least a portion of a source image, a label image including label information which is not included in the source image.

Still further provided .in accordance with a preferred embodiment of the present invention is a document generating system including a user-generated information processor operative to transfer user-generated information onto a document, and a document labeler operative to incorporate a label onto the document including processor identification information, not generated by the user, which is operative to identify the information processor.

Further in accordance with a preferred embodiment of the present invention the information processor includes an image scrambler.

Still further in accordance with a preferred embodiment of the present invention the processor identification information includes a serial number identifying the information processor.

There is additionally provided in accordance with a preferred embodiment of the present invention a method for scrambling a source image including generating a template image including a plurality of reserved locations and a plurality of nonreserved locations, wherein at least some of the reserved locations are arranged in a meaningful pattern, copying a plurality of portions of the source image into the plurality of nonreserved locations respectively, and filling the reserved locations which form the pattern with information which is visually distinguishable from the information in the nonreserved locations.

Also in accordance with a preferred embodiment of the present invention the meaningful pattern includes a non-grid pattern.

Further in accordance with a preferred embodiment of the present invention the meaningful pattern includes the pattern of at least one character.

Still further in accordance with a preferred embodiment of the present invention the pattern identifies an individual user and the method also includes authenticating the user based on the user-identifying pattern.

Also in accordance with a preferred embodiment of the present invention the key identifying information is generated by employing the designated recipient's public key to conceal the scrambling key.

There is further provided in accordance with a preferred embodiment of the present invention a method for scrambling a source image including identifying blank portions in the source image, replacing at least some of the blank portions in the source image with target-bearing portions which each bear a target whose location within the target-bearing portion is known to a unscrambling processor, and scrambling the modified source image.

There is additionally provided in accordance with a preferred embodiment of the present invention a method for generating a unscrambled image by unscrambling a scrambled image, generated from a source image including at least one line of text, the method including copying a rectangular portion located in an individual location within the scrambled image into a different individual location within an area allocated for representing the unscrambled image, the rectangular portion having long and short edges, the short edges being generally parallel to the text lines and the long edges being generally orthogonal to the text lines, and repeating the copying step for additional rectangular portions.

Further in accordance with a preferred embodiment of the present invention the ratio between the lengths of the long and short edges exceeds 2.

Yet further in accordance with a preferred embodiment of the present invention the ratio between the lengths of the long and short edges exceeds 3.

There is provided in accordance with a preferred embodiment of the present invention a method for generating a unscrambled image by unscrambling a scrambled image, generated from a source image including at least one line of text, the method including copying a portion located in an individual location within the scrambled image into a different individual location within an area allocated for representing the unscrambled image, the portion having edges which are generally parallel to the text lines and edges which are generally orthogonal to the text lines, the length of each orthogonal edge being an integer multiple of the pitch of the text lines, and repeating the copying step for additional scrambled image portions.

There is still further provided in accordance with a preferred embodiment of the present invention a method for scrambling a source image including a first plurality of first pixel arrays, the method including compressing the source image by identifying at least one sequence of blank first pixel arrays, deleting the sequence and adding an indication of the length of the sequence, and scrambling the compressed source image by generating a second plurality of second pixel arrays, wherein at least one of the second pixel arrays in an individual location in the second plurality includes a first pixel array in a different location in the first plurality.

There is additionally provided in accordance with a preferred embodiment of the present invention a method for scrambling a source image including generating a template image including a plurality of reserved locations and a plurality of nonreserved locations, copying a plurality of portions of the source image into the plurality of nonreserved locations respectively, and filling at least some of the reserved locations with hollow registration marks.

Further in accordance with a preferred embodiment of the present invention the source image includes a text and each copied portion of the source image is similar in size to the size of the text characters.

Still further in accordance with a preferred embodiment of the present invention the size of the cavity in each hollow registration mark exceeds the size of the cavities of the text characters.

There is additionally provided in accordance with a preferred embodiment of the present invention a method for unscrambling documents using a facsimile machine and a computer, having a fax modem and being in telephone communication with the facsimile machine, the method including providing the computer with a scanning error correction and unscrambling function, feeding a scrambled document into the facsimile machine and communicating its contents as a facsimile message via the telephone to the computer, and employing the computer's correction/unscrambling function to correct for scanning errors of the facsimile machine and to unscramble the document.

Still further in accordance with a preferred embodiment of the present invention the sending step includes sending the contents of the scrambled document to the computer via a telephone line simulator. Yet further in accordance with a preferred embodiment of the present invention the sending step includes sending the contents of the scrambled document to the computer via the serial port of the computer.

There is further provided in accordance with a preferred embodiment of the present invention portable image modification apparatus operative in conjunction with a facsimile machine and including a connector operative to transmit a digital representation of an image to and from the facsimile machine, and

a dedicated image modifier operative to receive the digital representation from the connector, to modify the image and to return the modified image to the connector.

Still further in accordance with a preferred embodiment of the present invention the dedicated image modifier has no graphics display.

Also in accordance with a preferred embodiment of the present invention the dedicated image modifier has no alphanumeric input device.

Further in accordance with a preferred embodiment of the present invention the dedicated image modifier includes an image scrambler.

Additionally in accordance with a preferred embodiment of the present invention the dedicated image modifier includes an image unscrambler.

Further in accordance with a preferred embodiment of the present invention the image modifier includes an optical character recognition device.

Still further in accordance with a preferred embodiment of the present invention the connector is operative to transmit dialed numbers from the facsimile machine and wherein the image modifier is operative to perform image modification on the basis of information conveyed by the dialed numbers.

There is provided in accordance with a preferred embodiment of the present invention a mass- distributed printed product including a scrambled image of confidential information characterized in that the confidential information is accessible to designated recipients who possess a unscrambling key by scanning and unscrambling the scrambled image.

There is also provided in accordance with a preferred embodiment of the present invention a method for mass dissemination of confidential information to a first multiplicity of designated recipients while maintaining confidentiality versus a second multiplicity of nondesignated recipients, the method including providing a unscrambling key to the first multiplicity of designated recipients, and separately disseminating a scrambled image of confidential information to the first and second multiplicities of recipients.

Further in accordance with a preferred embodiment of the present invention the method also includes scanning the scrambled image, and unscrambling the scanned image.

Still further in accordance with a preferred embodiment of the present invention the disseminating includes publishing in a mass-disseminated printed medium.

There is additionally provided in accordance with a preferred embodiment of the present invention a method for monitoring use of a software program by an individual computer platform, the method including earmarking a data storage component of the computer substantially without harming the computer's processing capabilities, thereby to provide a long-term record of engagement between the software program and the individual computer platform.

Further in accordance with a preferred embodiment of the present invention the earmark is a confidential function of a characteristic of the computer platform environment.

Still further in accordance with a preferred embodiment of the present invention the earmarking step is repeated each time the program is launched and wherein the earmark is changed for each repetition.

There is further provided in accordance with a preferred embodiment of the present invention a method for cracking a scrambled graphics image including a plurality of scrambled image portions, the method including forming a unscrambled area by identifying a neighbor, in the source image, of a first scrambled image portion along a first edge thereof from among other scrambled image portions by identifying a scrambled image portion having an edge whose cross-correlation with the first edge is high, and enlarging the unscrambled area by repeating, at least once, the forming step so as to identify additional source image neighbors of the image portions within the unscrambled area.

Further in accordance with a preferred embodiment of the present invention the image modifier includes a sketch straightener.

Yet further in accordance with a preferred embodiment of the present invention the image modifier includes an image merger operative to merge an input image with a stored backdrop image.

There is also provided in accordance with a preferred embodiment of the present invention a method for scrambling a text image including loading into a memory a first swathe of the image including a single row of gixels in which a plurality of gixels is arranged in a first sequence, generating output signals representing the same portion of the image represented by the swathe by outputting a second single row of gixels in which the plurality of gixels is arranged in a second sequence, and repeating the loading and generating steps for at least one more swathe of the image.

There is additionally provided in accordance with a preferred embodiment of the present invention apparatus for unscrambling documents including a scrambled document signal source, providing signals that represent the contents of a document, receiving a document to be unscrambled and providing output signals representing the contents of the document, an unscrambler for operating on the output signals to produce modified output signals representing a unscrambled version of the document, the unscrambler employing a system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation, and an unscrambled document generator receiving the modified output signals and producing a unscrambled version of the document.

Further in accordance with a preferred embodiment of the present invention the apparatus includes a unscrambling verifier for verification of correct unscrambling.

Still further in accordance with a preferred embodiment of the present invention the unscrambling verifier includes an edge correlator for carrying out edge correlation operations on adjacent scrambled pixels.

Yet further in accordance with a preferred embodiment of the present invention the unscrambler is controlled by a coded input to provide a selected one from a plurality of possible modifications to the output signals.

Also in accordance with a preferred embodiment of the present invention the apparatus may be embodied in a photocopier.

Additionally in accordance with a preferred embodiment of the present invention the apparatus may be embodied in a telefax.

Further in accordance with a preferred embodiment of the present invention the apparatus may be embodied in a computer.

Still further in accordance with a preferred embodiment of the present invention the unscrambler is operative to change the relative positions of multi-pixel regions of a document, without modifying the information content within each of the multi-pixel regions.

Additionally in accordance with a preferred embodiment of the present invention the unscrambler is operative to leave unchanged certain predetermined regions of the document.

There is also provided in accordance with a preferred embodiment of the present invention apparatus for unscrambling documents for use in conjunction with a document receiver and a scrambled document signal source receiving a document to be unscrambled and providing output signals representing the contents of the document, the apparatus including an unscrambler operating on the output signals to produce modified output signals representing a unscrambled version of the document, the unscrambler employing the system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation, and to provide the modified output signals to the document generator, for producing an unscrambled version of the document.

Further provided in accordance with a preferred embodiment of the present invention is a method for unscrambling documents including the steps of receiving a document to be unscrambled including a system of reference marks and providing output signals representing the contents of the document, operating on the output signals to produce modified output signals representing a unscrambled version of the document, employing the system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation, and receiving the modified output signals and producing a corrected unscrambled version of the document.

Further in accordance with a preferred embodiment of the present invention the reference marks are distributed throughout the area of the document. Still further in accordance with a preferred embodiment of the present invention the reference marks are distributed about the periphery of the area of the document.

Yet further in accordance with a preferred embodiment of the present invention the correct location of each of the reference marks is determined by its appearance.

Additionally in accordance with a preferred embodiment of the present invention the reference marks are characterized by a local color modification of the image information thereat.

Further in accordance with a preferred embodiment of the present invention the reference marks are arranged in a predetermined grid.

Still further in accordance with a preferred embodiment of the present invention the reference marks are distributed throughout the area of the document.

Also in accordance with a preferred embodiment of the present invention the reference marks are distributed about the periphery of the area of the document.

Further in accordance with a preferred embodiment of the present invention the correct location of each of the reference marks is determined by its appearance.

Still further in accordance with a preferred embodiment of the present invention the reference marks are characterized by a local color modification of the image information thereat.

Additionally in accordance with a preferred embodiment of the present invention the reference marks are arranged in a predetermined grid.

There is further provided in accordance with a preferred embodiment of the present invention a method for unscrambling documents for use in conjunction with a document receiver and a scrambled document signal source receiving a document to be unscrambled including reference marks and providing output signals representing the contents of the document, the method including the steps of operating on the output signals to produce modified output signals representing a unscrambled version of the document, employing the system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation, and providing the modified output signals to the document generator, for producing an unscrambled version of the document.

Still further in accordance with a preferred embodiment of the present invention the step of employing includes employing the reference marks to determine an average gixel size, and employing the reference marks to find a reference point for each of a plurality of gixels in the unscrambled image and copying a vicinity of average gixel size around the reference point into an area reserved for the unscrambled image, at a location determined by the key which generated the scrambled image.

The present invention will be more fully appreciated from the following detailed description, taken in conjunction with the drawings and appendices in which:

Fig. 1 is a generalized illustration of apparatus for creating and transmitting signals which represent scrambled documents, which is constructed and operative in accordance with a preferred embodiment of the present invention;

Fig. 2 is a generalized illustration of "paper flow" in accordance with a preferred embodiment of the present invention;

Fig. 3 is a block diagram illustration of the use of a digital copier in a preferred embodiment of the present invention;

Figs. 4A, 4B, 4C and 4D together constitute an illustration of target insertion and document scrambling provided in accordance with a preferred embodiment of the present invention;

Figs. 5A, 5B, 5C and 5D are illustrations of validation of unscrambling in accordance with a preferred embodiment of the present invention;

Figs. 6A, 6B, 6C, 6D, 6E and 6F together constitute an illustration of target insertion, scrambling, reconstruction and unscrambling in accordance with a preferred embodiment of the present invention;

Figs. 7A, 7B, 7C and 7D together illustrate reconstruction and unscrambling of a received scrambled image suffering from dislocation, lateral distortion and the obliteration of part of the image;

Fig. 8 is a diagram illustrating the computation of an error vector for a given pixel in accordance with a preferred embodiment of the present invention;

Figs. 9A and 9B are illustrations of the reconstruction of a stretched image;

Figs. 10A, 10B and 10C are illustrations of the reconstruction and unscrambling of an image having a missing row or rows;

Fig. 11 is an illustration of reconstruction of an image having a rotation distortion;

Figs. 12A, 12B, 12C, 12D and 12E together constitute an illustration of target insertion, image reconstruction and document scrambling in accordance with one embodiment of the present invention;

Figs. 13A, 13B, 13C, 13D and 13E together constitute an illustration of target insertion, image reconstruction and document scrambling in accordance with another embodiment of the present invention;

Figs. 14A, 14B, 14C, 14D and 14E together constitute an illustration of target insertion, image reconstruction and document scrambling in accordance with yet another embodiment of the present invention;

Fig. 15 is a simplified block diagram of a graphics communication system for scrambled images operative to communicate scrambled graphic images over time and/or over space;

Fig. 16A is a simplified flowchart of a preferred swathe-by-swathe method for performing a "scramble and fill" operation;

Fig. 16B is an illustration of a source image;

Fig. 16C is an illustration of the swathe-by-swathe scrambled image generated by applying the method of Fig. 16A to the source image of Fig. 16B;

Figs. 17A - 17B are pictorial diagrams illustrating a preferred method for performing the "scramble and fill" step described above;

Fig. 17C is a pictorial illustration of the source image;

Fig. 17D illustrates a scrambled image generated from the source image of Fig. 17C;

Fig. 18A is a simplified semi-pictorial semi- block diagram of a preferred method for incorporating a key identification label into a scrambled image;

Fig. 18B is a simplified semi-pictorial semi-block diagram of a preferred method for unscrambling a scrambled image including a key identification label;

Fig. 18C is a pictorial illustration of a machine readable label;

Fig. 19 is a tabular depiction of data useful in understanding the block diagram of Figs. 18A and 18B;

Fig. 20 is a simplified semi-pictorial semi-block diagram of another preferred method for performing the "template generation" step;

Figs. 21A - 21C illustrate a method for embedding opportunistic targets into a source image;

Fig. 22 is a pictorial illustration useful in understanding the template generation step of Fig. 18A;

Fig. 23 is a semi-pictorial semi-block diagram of template generation step of Fig. 18A;

Fig. 24 is a pictorial illustration of a template in which some of the reserved locations are arranged in a meaningful pattern;

Fig. 25A is a pictorial illustration of a source text image and gixel markings of rectangular gixels whose height is not a multiple of the text pitch;

Fig. 25B is a pictorial illustration of a portion of a scrambled image generated from the source image of Fig. 25A;

Fig. 25C is a pictorial illustration of a source text image and gixel markings of rectangular gixels whose height is equal to the text pitch;

Fig. 25D is a pictorial illustration of a portion of a scrambled image generated from the source image of Fig. 25C;

Fig. 26A is a source image which includes blank rows of gixels; Fig. 26B is an illustration of a method whereby a blank row indication is incorporated into the source image itself by adding an additional, "service" gixel at a preselected position within the row such as the end of each row;

Fig. 26C is a pictorial illustration of a sample binary code;

Fig. 27A is a source image including hollow patterns such as text;

Fig. 27B is a scrambled image generated from the source image of Fig. 27A;

Fig. 28 is an illustration of a preferred method for unscrambling documents using only conventional office equipment;

Fig. 29 is a modification of Fig. 28 in which the telephone link between a facsimile machine and a computer is replaced by a telephone line simulator;

Fig. 30 is a further modification of the apparatus of Fig. 28 in which the facsimile machine has a serial port connection;

Fig. 31 is a simplified block diagram of portable image modification apparatus;

Fig. 32 is an illustration of a preferred method and apparatus for mass dissemination of confidential information using a mass distributed printed product;

Fig. 33 is a preferred method for monitoring an individual computer platform's use of a software program;

Fig. 34 is a simplified flowchart of a preferred unscrambling method which includes, for each gixel, the step of copying a vicinity of average gixel size around the expected location of the gixel;

Fig. 35 is a pictorial illustration of a distorted scrambled image and of portions thereof which are to be copied in accordance with the method of Fig. 34;

Fig. 36 is a simplified flowchart of a preferred scrambling method which includes preparing a "watermark" template as in Fig. 24 whose gixel size is an integer multiple of the pitch as in Fig. 25A - 25D, the passpartout copying feature of Fig. 17, compiling a service label and incorporating key identification information therein as in Figs. 18, 22 and 23, compressing blank lines as in Figs. 26A - 26B, and generating opportunistic targets as in Figs. 21A - 21C; and

Fig. 37 is a simplified flowchart of a preferred unscrambling method which includes retrieving the unscrambling key from the service label, recognizing opportunistic targets as in Figs. 21A - 21C, and restoring blank lines as in Figs. 26A - 26B.

Appendix A is a hexadecimal listing of a preferred software implementation of the scrambling method of Fig. 36 except that the opportunistic target feature, the watermark feature and the blank line compression feature are omitted;

Appendix B is a hexadecimal listing of a preferred software implementation of the unscrambling method of Fig. 37 except that the opportunistic target feature, the watermark feature and the blank line compression feature are omitted; and

Appendix C is a printout of a sample set-up file used for de/scrambling.

Reference is now made to Fig. 1, which is a generalized illustration of apparatus for creating and transmitting scrambled documents constructed and operative in accordance with a preferred embodiment of the present invention. Ordinary, conventional office machines, such as a computer printer 10 and a typewriter 12 may provide a hard copy document, which alternatively may be handwritten. The document is readable by any person without required authorization and is normally human-readable.

In accordance with a preferred embodiment of the present invention, the hard copy document is provided to a scrambling copier 15, preferably a modified version of a digital copying machine such as a Canon 8580, which is capable of scanning hard copy documents into its memory and then printing them to make copies. In accordance with a preferred embodiment of the present invention, the digital copying machine is modified to scramble the contents of the hard copy document which is stored in its memory in accordance with a predetermined scrambling protocol, an example of which is described hereinbelow.

Additionally in accordance with a preferred embodiment of the present invention, the scrambling copier is also operative to distribute a system of mutually location coordinated reference marks, hereinafter "targets" in the scrambled image to assist in later reconstruction of a scrambled image which has undergone distortion.

Alternatively, a computer generated document may be transmitted directly from a computer 16 via an electronic fax machine 18, such as a fax modem, which generates a readable document which is supplied to the scrambling copier 15.

As will be described hereinbelow, the scrambling copier 15 provides a scrambled hard copy document, which can be handled in any conventional office procedure, such as copying, filing, mailing and faxing, without having the information contained therein disclosed to an unauthorized reader.

An authorized reader can, at any time, take the scrambled document, or a copy thereof and "copy" it on an unscrambling copier 20 and thus turn it into an ordinarily, human readable document.

Unscrambling copier 20 is preferably a digital copying machine such as a Canon 8580, which is capable of scanning scrambled hard copy documents into its memory and then printing them to make unscrambled, preferably human readable, copies. In accordance with a preferred embodiment of the present invention, the digital copying machine is modified to unscramble the contents of the hard copy document which is stored in its memory in accordance with a predetermined unscrambling protocol, an example of which will be described hereinbelow, and which is the inverse of the scrambling protocol used to scramble the document.

Additionally in accordance with a preferred embodiment of the present invention, the unscrambling copier is also operative to identify the mutually location coordinated targets in the scrambled image and to use them in reconstruction of a scrambled image which has undergone distortion.

It will be appreciated that preferably the scrambling copier 15 and the unscrambling copier 20 are respectively capable of scrambling and unscrambling a document in a multiplicity of different ways, which are selected by the input of a given code to the copier. The scrambling and unscrambling codes may need to be known to both the transmitter or recipient and may be configured in accordance with any suitable known scrambling and encryption technique. Alternatively, if a public key or any other information which identifies a key to selected parties is used, the recipient need not memorize the scrambling code. It is appreciated that the term "copier" includes integrated digital copiers such as the Canon 8580 and more generally includes any apparatus, such as a facsimile machine or scanner, which is operative to perform a scanning operation and a printing operation and which is capable of processing the scanned image prior to printing. Examples of devices having the above capabilities are shown and described below with reference to Figs. 28 - 31. Preferred methods for processing the scanned image prior to printing are shown and described herein with reference to Figs. 3 - 27.

Fig. 2 graphically illustrates a dual paper flow path in a typical office environment in accordance with the present invention. A regular document 50 is dealt with in an entirely normal way in all normal office procedures, such as copying, faxing, mailing and filing. If such a document is desired to be maintained confidential, it is copied on a target insertion/scrambling copier, such as an enciphering copier and is thus converted into an enciphered document 60 which cannot normally be read by a person. This enciphered document can be dealt with in all office procedures identically to the regular document, but cannot be read by unauthorized persons.

An authorized person receiving the enciphered document 60 and being in possession of an appropriate unscrambling code may readily copy the enciphered document 60 on a reconstruction/unscrambling copier 20, which may or may not be the same copier as scrambling copier 15 and may be remotely located therefrom. All that is required is that the unscrambling copier 20 operate in an inverse manner to the operation of scrambling copier 15. The result of unscrambling is a regular document, which may be entirely identical to original regular document 50 or may contain some indicia to indicate that it has undergone enciphering and deciphering.

In certain operations copiers 15 and 20 operate on fixed scrambling and unscrambling protocols and thus do not require the application of scrambling and unscrambling codes. Alternatively and preferably, the copiers 15 and 20 are capable of operation in a multiplicity of different scrambling and unscrambling modes of operation, which are selectable by appropriate codes.

In accordance with one embodiment of the invention, the unscrambling code, or part thereof may be carried by the document itself and may be readable by the unscrambling copier. Any other suitable technique of transmitting unscrambling code information may alternatively be employed.

Reference is now made to Fig. 3, which illustrates the general structure of the scrambling and unscrambling copiers 15 and 20. In each case, the copier comprises a document scanning unit 70 which outputs to a memory 72 which is interactively connected with a scrambling or unscrambling program 74, which may be an enciphering or deciphering program, as appropriate, which is typically embodied on a VLSI chip and which outputs to a printing unit 76. Program 74 preferably provides target placement and document reconstruction functions as well.

At the scrambling or enciphering end, a human readable document is scanned by the scanning unit 70 and an enciphered document is produced by the printing unit 76. An enciphering key is provided to the enciphering program 74 for this purpose. At the unscrambling or deciphering end, a scrambled document is scanned by the scanning unit 70 and a deciphered document is produced by the printing unit 76. A deciphering key is provided to the deciphering program 74 for this purpose.

Figs. 4A - 4D illustrate a typical target insertion and scrambling transformation wherein a multiplicity of multipixel areas 80, here termed "gixels", are transformed from their original relative spatial orientation on a human readable document 82 to a different, scrambled, relative spatial orientation on a scrambled document 84 which also includes target markings 85. The contours of the gixels are indicated herein as dotted lines, it being appreciated that such dotted lines do not have to appear on either the original or scrambled document. It is appreciated that within each gixel, the pixel arrangement is unscrambled.

In the illustrated embodiment, nine targets 85, marked 1 - 9 and having a predetermined spatial relationship therebetween are employed. Each target typically occupies a single gixel, causing the target inserted scrambled image of the 4 × 4 gixel human readable document 82 to have typically 5 × 5 gixels. Alternatively, no gixels are reserved for targets and instead, targets are superimposed into gixels which contain data.

The size of the gixel may be predetermined at the design stage. Larger gixels provide a more readable scrambled document, thus enabling faster processing. Small gixels provide a more scrambled document which is harder to read, but requires more processing time. The minimum gixel size is a single pixel. Information regarding the gixel size can be incorporated as part of the scrambling code, or alternatively it may appear on an unscrambled part of the document or be ascertained empirically by an examination of a scrambled document.

Arrows 90 indicate a typical spatial transformation, it being appreciated that any suitable spatial transformation may alternatively be provided.

Preferably part of the document area is not scrambled. The unscrambled area normally contains information that is intended to be readable by all persons, such as the date, addressee, identification of the sender, company logo and general instructions on how to unscramble the document. The non-scrambled area may include a heading portion, as well as a peripheral edge strip 88.

The peripheral edge strip 88 and the contours of the gixels 80, indicated in dashed lines may provide a registration frame, which although it may be considered to be a single mark, actually is considered for the purposes of the invention as a system of reference marks. Alternatively, any other suitable registration marks may alternatively or additionally be provided. It is appreciated that the registration frame or other registration marks do not normally appear on the original document, but rather are added to the scrambled document by the scrambling program and removed by the unscrambling program.

In accordance with a preferred embodiment of the present invention, verification of correct scrambling and unscrambling is provided by the scrambling and unscrambling copiers. As illustrated in Figs. 5A - 5D, such verification can readily be performed by considering the edges of adjacent gixels to confirm matching between light and dark regions thereon. Conventional spatial correlation techniques may be employed for this purpose.

Reference is now made to Figs. 5A - 5D, which illustrate one technique of validation of correct unscrambling.

Figs. 5A and 5B illustrate a case of valid unscrambling, in respective normal and enlarged views. The normal view illustrates the text divided into four gixels. Fig. 5B illustrates an enlarged part 91 of Fig. 5A, which lies along two adjacent gixels. It is noted that the two pixel lines 92 and 93 adjacent to the border between adjacent gixels are identical or nearly identical.

Figs. 5C and 5D illustrate a case of invalid unscrambling, in respective normal and enlarged views. The normal view illustrates the text divided into four gixels. Fig. 5C illustrates an enlarged part 94 of Fig. 5C, which lies along two adjacent gixels. It is noted that the two pixel lines 95 and 96 adjacent to the border between adjacent gixels are not even nearly identical.

It is appreciated that when a gixel boundary falls on the edge of a character or image, there may be a difference between lines 92 and 93. The average correlation over an entire page should nevertheless provide a reasonably accurate indication of valid unscrambling. Thus, by determining the correlation between pixel lines along the border between adjacent gixels, a measure of the accuracy of the achieved unscrambling may be obtained.

Reference is now made to Figs. 6A - 6F, which illustrate the present invention in general terms. An original image is shown at Fig. 6A, it being appreciated that normally the original image may include a large multiplicity of signals representing alphanumeric symbols, images or the like.

Fig. 6B illustrates a target map which typically includes nine target markings arranged in a predetermined arrangement. Fig. 6C illustrates a scrambled image having embedded therein target markings, preferably in the manner illustrated in Figs. 4A - 4D and described hereinabove.

The scrambled image of Fig. 6C is transmitted or handled in any usual manner, e.g. my mail, fax, etc. and in the course of transmission is subjected to various distortions, including, for example, skewing and staining. A typical received document having some of these distortions appears in Fig. 6D.

By employing the target markings, or targets, a receiving copier or other suitable apparatus provides reconstruction of the distorted scrambled image. The reconstructed scrambled, distorted image is shown in Fig. 6E. The reconstructed image is unscrambled to produce the image of Fig. 6F. The steps of target placement and reconstruction employing a rubber sheet mapping technique are summarized hereinbelow:

1. Placement of targets at predetermined locations throughout the scrambled image or about its periphery.

2. Upon reconstruction, for each target in the image to be reconstructed, calculate the X and Y error in location and prepare a list of vectorial errors for the targets.

3. For each pixel in the image to be reconstructed, find the N nearest targets. N can be any desired number; the larger that N is, the more accurate is the reconstruction.

4. For each pixel calculate a weighted vector of error by adding the vectorial errors of the N nearest targets, each weighted according to its proximity to the pixel. The weighting function should be such that as the pixel approaches a target, its error vector converges to that of the target.

5. For each pixel of the reconstructed scrambled image, calculate the error vector, and read the value of the corresponding pixel from the distorted scanned image.

Alternatively, if the targets are arranged on a grid, bilinear interpolation between targets may be employed to perform rubbersheeting.

Computer programs for scrambling, unscrambling and verifying correct unscrambling appear in Appendices A and B. These computer programs can readily be incorporated in conventional digital photocopiers such as those mentioned hereinabove.

Reference is now made to Figs. 7A, 7B, 7C and 7D, which together illustrate a process of unscrambling an image suffering from three interrelated types of distortions: dislocation, later distortion and the presence of a large stain or erasure of information at a given location. Fig. 7A illustrates a received scrambled image as scanned into unscrambling apparatus. As seen in Fig. 7B, the targets are identified, it being noted that two of the typically nine targets have been obliterated. The X and Y error vectors for each of the identified targets are determined.

A local error vector is assigned to every pixel in the image to be reconstructed, as discussed above and as exemplified in Fig. 8, discussed hereinbelow. In the illustrated embodiment, the error vector is determined by the three targets nearest to each pixel 98. Consideration of the distances to the neighboring targets and their error vectors provides local error vectors of each pixel according to the formula given in Fig. 8. It is appreciated that due to the obliteration of targets 5 and 6, the error vectors of the pixels adjacent thereto are based on relatively remote targets.

A reconstructed scrambled image produced by the foregoing reconstruction methodology is illustrated in Fig. 7C. It is noted that the sequence of pixels for calculation can be selected according to the order of the reconstructed image (pulling) or according to the order of the scanned image of Fig. 7A (pushing). It is appreciated in the art of raster image processing that pulling is preferred for ensuring that each pixel in the reconstructed image will be properly dealt with.

An unscrambled image, corresponding to the reconstructed scrambled image of Fig. 7C, is shown in Fig. 7D. It is noted that the obliteration is now scattered over the unscrambled image.

Reference is now made to Fig. 8, which illustrates one possible methodology for computing the error vector of a given pixel 100. In this case, the error vector is calculated employing the error vectors of the three nearest targets. It can be seen that as the pixel approaches any of the targets, its error vector converges to the error vector of that target. For example, as the pixel 100 approaches target no. 2, the distances d1, d2 and d3 converge to d12, 0 and d23 and the vector V converges to V2.

It is appreciated that various other methodologies may be employed for this purpose using any number of targets. It has been found, however that using the three nearest targets enables every pixel of the image to be located inside one triangle of targets. The three vertices of the triangle represent points 1, 2 and 3 of the following formula: V = d3 × d2 × V1 / d12 / d13 +

d1 × d3 × V2 / d12 / d23 + d1 × d2 × V3 / d13 / d23 Reference is now made to Figs. 9A and 9B, which illustrate the process of reconstruction of images having scale errors. In Fig. 9A, the scrambled image has been expanded horizontally and shrunk vertically.

As seen in Fig. 9B, the targets are identified and their error vectors are computed. It is noted that inasmuch as central target 5 is correctly placed, having a zero error vector, and thus indicates that displacement of the entire image is not present. The remainder of the unscrambling of the image follows the techniques described hereinabove.

Reference is now made to Figs. 10A, 10B and 10C, which illustrate the process of reconstruction and unscrambling of images having lost rows, as occurs frequently in fax transmissions. Fig. 10A illustrates such an image wherein a significant portion of the image is lost, including most of the gixels adjacent targets 6 and 7.

As seen in Fig. 10B, the first steps are to identify the targets and to compute their error vectors. It is seen that targets 1 - 5 are in their correct places, but targets 8 and 9 have moved upward. As seen in Fig. 10C, reconstruction of the image causes a strip 110 lying between target 5 and targets 8 and 9 in the distorted scrambled image of Fig. 10A to be vertically stretched to fill the gap between the targets. This causes local distortions in some gixels, as indicated by reference numbers 112 and 114 in Fig. 10C.

Reference is now made to Fig. 11, which illustrates the error vectors of the targets of an image which is distorted by rotation about its center. It is appreciated that the rotation distortion may be superimposed on the distortions of the scrambled image of Fig. 10A or any other distortions encountered in the received image to be reconstructed.

It is to be appreciated that normally a relatively large number of targets are employed in an image, preferably at least about 5% of the number of gixels defined for scrambling.

Reference is now made to Figs. 12A - 12E, which illustrate target insertion, image reconstruction and document scrambling employing peripheral targets. Fig. 12A shows an original image and Fig. 12B shows a scrambled image. Fig. 12C shows peripheral binary coded targets 120 arranged along the periphery of the scrambled image.

Fig. 12D illustrates a received scrambled image having missing rows, prior to reconstruction. Fig. 12E shows a reconstructed image, wherein the grey portion represents missing information. The reconstruction may be carried out using the methodology described hereinabove with respect to Figs. 4A - 11 wherein target identification is carried out according to the binary coding. Reference is now made to Figs. 13A - 13E, which illustrate target insertion, image reconstruction and document scrambling employing selected gixel reversal. Fig. 13A shows an scrambled image and Fig. 13B shows a pattern of gixels to be reversed (black-white reversal). Fig. 13C shows the scrambled image of Fig. 13A having selected gixels thereof reversed according to the pattern of Fig. 13B. Generally, reversed gixels can be recognized by the percentage of white or black therein.

Fig. 13D illustrates a received scrambled image corresponding to that of Fig. 13C having missing rows, prior to reconstruction. Fig. 13E shows a reconstructed image, wherein the grey portion represents missing information. The reconstruction may be carried out using the methodology described hereinabove with respect to Figs. 4A - 11, modified so as to identify the targets as reversed gixels.

Reference is now made to Figs. 14A - 14E, which illustrate target insertion, image reconstruction and document scrambling employing peripheral targets in accordance with a further alternative embodiment of the present invention. Fig. 14A shows an original image and Fig. 14B shows a scrambled image. Fig. 14C shows peripheral targets 130 arranged along the periphery of the scrambled image. The targets have a varying white/black ratio which enables their X and Y coordinates to be determined.

Fig. 14D illustrates a received scrambled image having missing rows, prior to reconstruction. Fig. 14E shows a reconstructed image, wherein the grey portion represents missing information. The reconstruction may be carried out using the methodology described hereinabove with respect to Figs. 4A - 11 wherein target identification is carried out according to the sensed white/black ratio.

Reference is now made to Fig. 15 which is a simplified block diagram of a graphics communication system for scrambled images operative to communicate scrambled graphic images over time and/or over space. The graphic communication system of Fig. 15 manipulates a carrier 154 bearing a scrambled graphics image 156 and, typically, a template 158. The carrier may comprise a sheet or page or alternatively may comprise an electromagnetic signal. The carrier may be generated by scanning a document 159 into the memory of a scanner 160 and subsequently digitally scrambling the contents of the document, by means of a digital scrambler 162. Alternatively, the carrier may be digital, originally, in which case scanner 160 is omitted.

Typically, the digital scrambler is operative to embed

a target into the scrambled image, as described in more detail above with reference to Figs. 4, 6 and 7. A preferred mode of operation for the digital scrambler 162 is described below with reference to Fig. 36.

The carrier may then be manipulated in one or more different ways, such as but not limited to being scanned, faxed or otherwise communicated to a remote location, photocopied, stored, and/or physically handled. The carrier, in the course of being processed by the graphics communication system of Fig. 15, may undergo one or more of the following physical processes, each of which typically degrades the scrambled graphics image being communicated:

a. Processing by a scanner, fax machine or photocopier having an inoperative LED, which results in lost columns.

b. Long term storage in the course of which the document may expand, shrink or become distorted, resulting in scale errors.

c. Noisy signal or signal cut-off momentarily due to faulty transmission link over which a document is being faxed or otherwise transmitted to a remote location.

d. Crooked feeding of the document into a scanner, fax or photocopier, resulting in skew or rotation of the image on the new carrier; e. Careless handling, resulting in stains, folds, holes and/or tears of the document; and/or f. Optical errors in the lenses incorporated in a copier, which distort the image.

A particular feature of the graphics communication system of Fig. 15 is that physical degradations of the scrambled graphics image are corrected before and/or after the scrambled image is unscrambled. In the illustrated embodiment, a degradation correction unit 170 is provided upstream of a unscrambler 180. Unscrambler 180 is operative to unscramble an upgraded scrambled image 190 provided by the degradation correction unit 170. The output of unscrambler 180 is a digital representation of the unscrambled and upgraded image 200 which may be displayed electronically, as via a CRT, or may be printed on a printer 210 or other output device.

A preferred mode of operation for the digital unscrambler 180 is described below with reference to Fig. 37.

The scrambler/target embedding unit of Fig. 15 is operative to perform the following steps:

a. TEMPLATE GENERATION: Generate a template image, on any suitable carrier, with targets, also termed herein "scheduled targets", and additional service information, as described in detail below, at reserved locations which are known to the unscrambling processor. A sample template image is illustrated in Fig. 4A.

b. SCRAMBLE AND FILL: Fill nonreserved locations of the template image with portions of the source image, using a scrambling key to determine the correspondence between locations of the portions within the source image and the nonreserved locations of the template image.

Suitable methods for implementing the above steps are described above with reference to Figs. 12A - 14E.

It is appreciated that, as described above with reference to Figs. 13A - 13E, the targets are sometimes not in reserved locations but rather are implemented as predetermined modifications of predetermined portions of the source image.

Reference is now made to Fig. 16A which is a simplified flowchart of a preferred swathe-by-swathe method for performing the "scramble and fill" step described above. Fig. 16B is an illustration of a source image. The swathe-by-swathe scrambled image generated by applying the method of Fig. 16A to the source image of Fig. 16B is illustrated in Fig. 16C. The method of Fig. 16A preferably includes the following steps, described with reference to Figs. 16B - 16C:

a. A first swathe 170 of the source image, including a plurality of scan lines 172 and representing a first array 174 of source image gixels, is loaded into computer memory. The source image is illustrated in Fig. 16B.b. Output signals are generated which result in a swathe 175 representing the same portion of the image represented by the swathe 170. The output signals are generated by outputting a second plurality 176 of gixels of the image which is a different arrangement of the first plurality 174 of gixels.

c. Steps a and b are repeated for at least one more swathe 178 of the source image, resulting in at least one more swathe 180 in the scrambled image.

It is appreciated that, alternatively, each swathe may include only a single row of gixels, as indicated by reference numerals 184. The advantage of having a single gixel-row swathe is that blank rows of pixels appearing in the source image are preserved in the scrambled image, because scrambling is only performed along the horizontal dimension. Preservation of blank rows is desirable because it facilitates compression of the scrambled image for fax transmission. A particular advantage of the method of Fig. 16A is that much less computer memory is required, relative to an embodiment in which any source image portion may be assigned to any nonreserved location anywhere within the entire template image. In the embodiment of Figs. 16A - 16C, only one swathe of the source image need be in computer memory at any one time. Therefore, the amount of memory required is less than that Required to store the entire source image. shown in Figs. 16B and 16C, the unit of unscrambling need not necessarily be an entire conventionally sized page but rather may be a horizontal swathe of the conventionally sized page. Therefore, in the present specification and claims, the term "page" is intended to refer to a portion of the scrambled image or the entirety thereof, which is unscrambled as a single unit.

Reference is now made to Figs. 17A - 17B which are pictorial diagrams illustrating a preferred method for performing the "scramble and fill" step described above. Fig. 17A is a pictorial diagram of a source image, illustrated as including only 4 portions or gixels 191 for simplicity. The outlines of the gixels are indicated by a solid line. Fig. 17B is a pictorial diagram illustrating a scrambled image generated from the source image.

The scrambled image of Fig. 17B is generated by generating an arrangement of portions or gixels 192, each of which includes a source image gixel 191 and a passpartout 196 thereof. The term "passpartout" refers to a vicinity of a gixel in the source image which vicinity at least partially surrounds the gixel. Preferably, at least one of the targets of the scrambled image represents target information which was not present in the source image, as described below with reference to Fig. 24. It is appreciated that the passpartout may either completely surround the source image gixel, as in Figs. 17A - 17B, or only partly surround the source image gixel. For example, in facsimile applications, the accuracy along the horizontal scan lines is very high, in contrast to the accuracy along the paper motion (vertical) axis which is much lower. Therefore, in these applications, it is often sufficient for the passpartout to include only the vertically adjacent vicinity of the gixel, as illustrated in Fig. 17C which is a pictorial illustration of the source image. Fig. 17D illustrates a scrambled image generated from the source image of Fig. 17C in which each gixel in the scrambled image includes a gixel from the source image as well as the vertical passpartout thereof.

Reference is now made to Fig. 18A which is a simplified semi-pictorial semi-block diagram of a preferred method for performing the "template generation" step described above. According to the method of Figs. 18A-18C, a location 201 within the scrambled document 211 is reserved for key identifying information which allows a scrambling key 220, employed to scramble the source image 230, to be identified, preferably automatically, by a designated recipient. The apparatus includes a clue generator 222 which is employed during scrambling to transform the scrambling key 220 into key identifying information which is then placed at location 201.

The clue generator is also employed during unscrambling, by the designated recipient, to transform the key identifying information back into the scrambling key. The clue generator typically applies a transformation to the key identifying information displayed in location 201, in order to extract the scrambling key 220. The transformation is a confidential secret between the sender and the designated recipient. Fig. 18B is a simplified semi-pictorial semi-block diagram of a preferred method for unscrambling the scrambled image of Fig. 18A. To unscramble, the key identifying information is interpreted by a human operator or alternatively, if the label 201 is machine readable as shown in Fig. 18C, by an automatic clue interpreter 232. The key which is produced by the interpretation process is used by a unscrambler 234 to generate a unscrambled image 236 from the scrambled image 211.

For example, as shown in Fig. 19, the designated recipient and the scrambling party may both have access to a confidential key list 238 in which each of a plurality of keys 240 is associated with a serial number 250. The key identification information then comprises the serial number of a scrambling key selected by the scrambling party from among the scrambling keys 240.

Another example is shown in Fig. 20. A public/private coding process, such as the RSA method, described in an article by Asael Dror which appears in Byte, June 1989, pp. 267 - 270, is employed to conceal the scrambling key in a message using the public key of the designated recipient. In this application, the key identification information comprises the message which conceals the scrambling key. The scrambling key can only be extracted from the message by using the designated recipient's private key which is confidential.

Optionally, "opportunistic targets" are embedded onto at least some of the portions of the scrambled image which correspond to blank portions of the source image, during the "scramble and fill" step. The "opportunistic targets" augment the scheduled targets which appear at the reserved locations, as described above. The additional targets provide a higher density of targets without further enlarging the overhead in the scrambled image. The "opportunistic targets" may or may not have the same shape as the scheduled targets.

Reference is now made to Fig. 21A which illustrates a blank portion 260 of a source image, the portion 270 corresponding thereto in a scrambled image, on which is embedded an opportunistic target 280, and the corresponding portion in the unscrambled image on which the opportunistic target does not appear. Fig. 21A is best understood by comparison with Figs. 21B and 21C. Fig. 21B illustrates a non-blank (data bearing) portion 290 of the source image which remains the same within the scrambled and unscrambled images. Fig. 21C illustrates a scheduled target 300, which appears in a reserved location 310 within the scrambled image which reserved location typically does not correspond to a location within the source or unscrambled images.

A preferred method for performing the "scramble and fill" step of the source image scrambling process so as to incorporate opportunistic targets into the scrambled image includes the following steps:

a. Read the source image, partition into gixels and identify blank gixels.

b. Replace at least some of the blank gixels with target-bearing gixels which bear opportunistic targets whose location within the gixel is known to the unscrambling processor. For example, the opportunistic targets may be positioned at the center of some of the blank gixels. The opportunistic targets may be differently shaped than the scheduled targets.

c. Scramble the modified source image to obtain the scrambled image with opportunistic targets.

A preferred method for unscrambling a scrambled image includes the following steps:

a. Find the scheduled targets.

b. Rubber-sheet the scrambled image so as to align the locations at which the scheduled targets are found with their prescribed locations within the scrambled image which are known. A preferred method for rubber-sheeting an image is described above with reference to Fig. 8.

c. Search for opportunistic targets in the rubber-sheeted scrambled image.

d. For each opportunistic target found:

i. Identify the gixel in which the opportunistic target resides.

ii. Compute the prescribed location of the opportunistic target within the scrambled image, based on the prescribed location within the scrambled image of the gixel in which the opportunistic target resides, and on the prescribed location of the opportunistic target within the gixel.

iii. Add the opportunistic target to a list of targets to be used for a second rubber-sheeting process. The list of targets initially includes only the scheduled targets.

f. Rubber-sheet the scrambled image using all targets in the list, i.e. all available scheduled targets and opportunistic targets. Preferably, in the template generation step, the template image includes service information other than targets. For example, the template image may include, apart from the scrambled data and the targets, unscrambling information such as scrambling key identification information. Reference is made to Fig. 22 which illustrates a source image, the corresponding scrambled image and the corresponding unscrambled image. As shown, all three images include a header section 310 which bears a header which appears clear in all three images. The three images also include a data section 320 which bears clear data in the source and unscrambled images and scrambled data in the scrambled image. The data section 320 in the scrambled image is typically somewhat larger in area than the data sections of the source and unscrambled image due to the overhead required for targets. As shown, the scrambled image includes a label section 330 which bears service information, typically clear, such as:

a. scrambling key identification information or other information which is useful in unscrambling the image; and/or

b. scrambler identification information operative to identify the serial number of the individual scrambler employed to scramble the source image.

More generally, the label section may bear any type of information which is not included in the source image. The information may comprise operator-readable information or machine readable information, such as a bar code. The information may or may not represent alphanumeric information.

It is appreciated that the applicability of the above feature is not limited to scrambling applications. More generally, as illustrated in Fig. 23, the above feature is useful in any type of document generating system which includes a user-generated information processor 350, such as but not limited to a word processor, copier, printer or fax. According to the above-described embodiment of the present invention, a document labeler 360 is provided which incorporates a label 370 onto each document 380 produced by the information processor. The label bears the serial number of the information processor. For example, the serial number of the information processor may be programmed into the document labeler.

Alternatively or in addition, service information may appear within the scrambled image itself rather than in a separate label or in addition to service information provided in a label.

Fig. 24 is a pictorial illustration of a template in which some of the reserved locations, which are intended to store scheduled targets, are arranged in a grid, as indicated by reference number 400. Others of the reserved locations, as indicated by reference number 410, are arranged in a meaningful pattern. In the illustrated example, the meaningful pattern comprises the patterns of the letters "J" and "K". However, more generally the meaningful pattern may comprise any other meaningful shape, such as a logo or known pictorial symbol. Fig. 25A is a pictorial illustration of a source text image. Lines 450 are used to indicate gixel borders although no such lines 450 are actually part of the image. Fig. 25B is a pictorial illustration of a portion of a scrambled image generated from the source image of Fig. 25A. If the source image is a text image it is preferable that the nonreserved locations or gixels 460 be rectangular such that the edges parallel to the text lines are shorter than the edges orthogonal to the text lines, as shown in Fig. 25B. For example, the ratio between the two sides of each gixel may be 2 or 3. One reason why rectangular pixels are advantageous is that the text is highly correlated along the text lines and much less correlated along the direction perpendicular to the text lines. Therefore, fine partitioning into gixels is advantageous along the dimension parallel to the text lines. The same fineness is not required along the perpendicular dimension and preferably is not provided in order to reduce computational complexity.

Preferably, as shown in Figs. 25C and 25D, if the image comprises a printed text, the length of the perpendicular edge is equal to, or is an integer multiple of, the pitch p of the text lines. For example, in Figs. 25C and 25D the length of the perpendicular edge is equal to the pitch p. If this is the case, the scrambled image will include horizontal blank strips 460, as does the source image, which expedites transmission of the scrambled document by fax. If the length of the long edge is not an integer multiple of the pitch then, as shown in Fig. 25B, there are no blank strips extended across the scrambled image and therefore, fax transmission is slower.

It is appreciated that, when scrambling a source image or when unscrambling a scrambled image, the repeated copying of gixels which effects the scrambling or unscrambling may be carried out piecewise on each gixel. The term "copying", as used in the present specification and claims, is intended to include piecewise copying, for example, square-by-square copying of a rectangular gixel comprising a plurality of square gixel portions.

Reference is now made to Figs. 26A and 26B. Fig. 26A is a source image which includes 4 blank rows 480 of gixels. According to a preferred embodiment of the present invention, the source image is compressed in the following manner before being scrambled:

a. For each non-blank row of gixels of the source image, an indication is stored, and is subsequently made available to the unscrambler, of the number of blank rows following the non-blank row. Fig. 26B illustrates a method whereby this indication is incorporated into the source image itself by adding an additional, "service" gixel at a preselected position within the row such as the end of each row. A sample binary code which allows the "service" gixel to indicate numbers is illustrated in Fig. 26C. In the illustrated example, the first three rows are each followed by zero blank rows, as indicated by the service gixels 490 thereof, whereas the fourth row is followed by four blank rows, as indicated by the service gixel 500 thereof.

b. Remove the blank rows.

The image is then scrambled as usual, including "service" gixels if provided. The resulting scrambled image is generally smaller than the original image.

In the course of ordinary unscrambling of the compressed source image, the "service" gixels, if provided, will return to the ends of the rows, thereby to provide a unscrambled compressed image.

To decompress the image, the "service" gixels, if any,

are omitted and the appropriate number of blank rows are reintroduced in the appropriate locations, according to the stored indication of blank rows in the "service" gixels or elsewhere.

Reference is now made to Fig. 27A which is a source image including hollow patterns such as text. Fig. 27B is a scrambled image generated from the source image of Fig. 27A. According to a preferred embodiment of the present invention,

source images including hollow patterns are typically scrambled by means of a template, illustrated in Fig. 27B, which includes hollow targets 520 at reserved locations and nonreserved locations or gixels 530 which are similar in size to, or smaller than, the size of the text characters.

Provision of hollow targets facilitates target recognition because the hollow patterns in the source image are typically broken up in the gixel permutation process, so that a hole or cavity is highly indicative of the presence of a target. Preferably, the holes or cavities in the targets are larger than the typical size of hole or cavity in the text letters to further facilitate target recognition.

Reference is now made to Fig. 28 which illustrates a preferred method for unscrambling documents using only conventional office equipment. Specifically, the method of Fig. 28 employs a conventional facsimile machine 550 with no scrambling/unscrambling capability, a computer 560, having a fax modem 570 and being in telephone communication with the facsimile machine 550, and, optionally a printer 580. A particular feature of the method of Fig. 28 is that a high resolution scanner is not required. Instead, it is possible to use a fax machine 550 for scanning, despite the poor quality of scanning which characterizes fax machines, because the method of Fig. 28 is operative to reduce the effects of scanning errors. The advantage of reducing the effects of scanning errors is that, thereby, only low-cost and universally available office equipment is required which is also even currently available in many homes.

The method preferably comprises the following steps:

a. Providing the computer 560 with a scanning error correction and unscrambling function 590 which typically is implemented in software and is loaded onto the computer by conventional means.

The scanning error correction and unscrambling function 590 may for example be based on the scanning error correction and descrambling function described below with reference to Fig. 37.

b. Feeding a scrambled document 600 into the facsimile machine 550 and communicating its contents as a facsimile message via telephone line 610 and local switchboard 620 to the computer 560;

c. Employing the computer's correction/unscrambling function 590 to correct for scanning errors of the facsimile machine and to unscramble the document.

d. Providing a hard copy output 630 of the unscrambled document via the facsimile machine 550 or via the printer 580.

Reference is now made to Fig. 29 which is a modification of Fig. 28 in which the telephone link between the facsimile machine 550 and the computer 580 is replaced by a line simulator 640, such as the P.A.S. port line simulator, commercially available from Shrink-Wrap Software Ltd., 32 Hashachar St., Raanana, Israel.

Reference is now made to Fig. 30 which is a further modification of the apparatus of Fig. 28 in which the facsimile machine has a serial port. connection 650. A commercially available facsimile machine of this type is the F26, commercially available from Acer Computer Turkey, Nispetiye Caddesi, Profesorier Sitesi B-2 D.1, 80630 Etiler - Istanbul, Turkey.

In Fig. 30, the telephone link between the fax and the computer is replaced by communication between the serial ports 650 and 660 of the fax and the computer respectively, such as by means of an RS232 line 670. Reference is now made to Fig. .31 which is a simplified block diagram of portable image modification apparatus which is particularly adapted for use by traveling business executives in that it is dedicated to image modification and therefore lacks one or more space-consuming elements such as keyboard, screen, digital I/O connectors and hard disk. Elimination of these elements also results in apparatus which is particularly easy to use with very little if any initialization processes required. In contrast, use of a laptop computer to perform the functions of the dedicated apparatus of Fig. 31 would result in a more heavy and bulky system which would additionally be more energy consuming and would require initialization processes such as bootstrapping, loading the operating system, accessing the disk, and interactively selecting the specific function to which the apparatus of Fig. 31 is dedicated.

The apparatus of Fig. 31 is operative in conjunction with a conventional facsimile machine 700 which may entirely lack image modification capabilities such as OCR, scrambling, unscrambling, sketch straightening, and scaling (enlarging or reducing). A telephone connector 710 communicates with the facsimile machine 700 via a telephone cable. The connector 710 is operative to communicate to the other components of the apparatus fax signals representing the image 720 to be modified as well as, optionally, instructions on how to modify the image. The connector 710 also communicates the fax signals representing the modified image back to the facsimile machine which provides a hard copy 730 of the modified image.

According to a preferred embodiment of the present invention, the portable image modification apparatus entirely lacks user input means and instead is operative to interpret dialing tone signals received from fax machine 700 as control inputs.

The portable image modification apparatus preferably includes a fax modem 740, an input image memory 750, an image modifier 760, a dial tone recognizer and interpreter 770 and an output image memory 790.

A preferred method of operation for the apparatus of Fig. 31 is as follows:

a. The user feeds a document bearing the image 720 to be modified into the fax machine 700;

b. The user typically inputs a control instruction via the dialing keypad 800 of the fax machine 700. The control instructions typically have a known syntax which may or may not employ a separator signal such as the # key of the fax machine. For example, a separate number code may represent each of the following image modification operations: OCR, scrambling, unscrambling, sketch straightening, and scaling (enlarging or reducing). Some or all of the operation codes may be followed by a parameter number code representing one or more parameters defining the selected operation. For example, dialing of the scrambling or unscrambling operation codes may be followed by dialing of the scrambling key. Dialing of the enlarging or shrinking operation codes may be followed by dialing of the percentage by which the original image 720 is to enlarged or shrunk.

c. The control instruction is communicated, via the telephone cable and the connector 710, to dial tone recognizer and interpreter 770 which is operative to recognize the dial tones as number codes and to interpret the number codes as operation codes and operation parameter codes accordance with the known syntax.

d. The fax modem 740 recognizes the termination of the dialing sequence, and signals to the fax machine 700 via the connector 710 that it is ready to receive a fax message.

e. The fax machine interprets the signal received from the fax modem 740 as indicating that a remote fax machine whose number is as dialed is ready to receive a fax message.

f. The image 720 is transmitted from the fax machine 700 to the fax modem 740, using the conventional fax protocol.

g. The input image 720 is stored in input image memory 750.

h. The input image 720 in memory 750 is modified by image modifier 760 in accordance with the interpreted control instructions provided by dial tone recognizer/interpreter 770.

i. The modified image is stored in output image memory 790.

j. The fax modem initiates a facsimile transmission call to the facsimile machine 700 and transmits the modified image thereto using the conventional protocol.

As described above, the apparatus of Fig. 31 is applicable to one or more of a wide variety of image modification processes such as but not limited to OCR, scrambling, unscrambling, sketch straightening, and scaling (enlarging or reducing).

Another operation which may be performed by the apparatus of Fig. 31 is superimposing an image 770 onto a prestored backdrop image such as a company letterhead or such as a form outline. To provide this operation, a backdrop image memory 810 is provided which stores one or more backdrop images. When dialing the superimposing operation, the user also dials a serial number indicating the backdrop image within memory 810 to be superimposed. Yet another operation is typically provided in addition to the superimposing operation, namely a "store new backdrop" operation in which a new backdrop image scanned by the fax machine 700 is stored in backdrop memory 810.

OCR methods are well known and are described, for example, in U.S. Patent 4,944,022 to Yasujima et al.

Scrambling and unscrambling methods suitable for the embodiment of Fig. 31 are described above in the present specification.

A commercially available sketch straightening device is the Imagio MF530 multifunctional copier, marketed by Ricoh.

Reference is now made to Fig. 32 which illustrates a preferred method and apparatus for mass dissemination of confidential information using a mass distributed printed product.

According to the embodiment of Fig. 32, confidential information is scrambled by a scrambler 830, thereby to generate a scrambled image which may be printed onto a document 840. The document 840 or digital representation of the scrambled image is incorporated into a mass-distributed printed product 850 such as a newspaper. The newspaper is distributed to a multiplicity of recipients including a first multiplicity of designated recipients and a second multiplicity of nondesignated recipients. The designated recipients each have access to the scrambling key.

Each designated recipient can unscramble the scrambled image appearing in the mass-distributed printed product by scanning the scrambled image, using a conventional scanning device 860 such as a conventional fax machine, and unscrambling the scanned scrambled image using the scrambling key. The nondesignated recipients, on the other hand, cannot unscramble the scrambled image appearing in the mass-distributed printed product.

Reference is now made to Fig. 33 which is a preferred method for monitoring an individual computer platform's use of a software program such as a scrambling program when it is desired to provide a fixed maximum number of uses of the program (such as a fixed number of scrambling sessions) under a first set of conditions and to allow subsequent uses of the program under different conditions. The term "trial status" is used herein to denote an individual computer platform which has used the program less than the maximum number of times. Once the program has been used a maximum number of times, an expiration procedure is initiated which typically comprises displaying a message to the user prompting him to decide whether to discontinue use of the program or whether to "sign on", i.e. accept the new conditions under which continued use of the program is being offered. Once the user has signed on, the program assigns "permanent status" to the individual computer platform.

The method of Fig. 33 is typically performed by the program upon launching thereof.

In step 900, the preselected file is preferably one which evolves with use, such as the WIN. INI file in the Microsoft Windows environment.

The preselected data storage component may, for example, comprise the hard disk on which the WIN. INI file is stored.

In step 910, the program distinguishes between a first use of the program and a subsequent use of the program, in which case an earmark is encountered.

The earmark employed by the method of Fig. 33 is preferably a changing earmark so as to impede attempts of a user to detect the earmark and remove it. Preferably (step 920), the earmark and/or its location within the preselected file is repeatedly recomputed as a function of a dynamic parameter, such as the time, of the computer platform environment. Alternatively, the earmark may be varied in any other suitable fashion, such as randomly or serially.

Fig. 34 is a simplified flowchart of a preferred unscrambling method which includes, for each gixel, the step of copying a vicinity of average gixel size around the expected location of the gixel in the scrambled image, into an area in memory reserved for the unscrambled image, at a location determined by the scrambling key. The method of Fig. 34 is less accurate than the rubbersheeting method described above, but is computationally much simpler.

Fig. 35 is a pictorial illustration of a distorted scrambled image having distorted gixels 970 which may be unscrambled using the method of Fig. 34. Targets 980 are used to determine expected locations of gixels. Vicinities 990 of average gixel size around the targets are copied into the memory area reserved for the unscrambled image.

Fig. 36 is a simplified flowchart of a preferred scrambling method which includes: preparing a "watermark" template as in Fig. 24 whose gixel size is an integer multiple of the pitch as in Fig. 25A - 25D, the passpartout copying feature of Fig. 17, compiling a service label and incorporating key identification information therein as in Figs. 18, 22 and 23, compressing blank lines as in Figs. 26A - 26B, and generating opportunistic targets as in Figs. 21A - 21C.

In the present specification and claims, the term "image" refers to a raster representation of information, including graphics and alphanumeric information or to a vector representation thereof and excludes alphanumeric representations of information.

The term "scrambling" refers to gixel-by-gixel treatment of a source image wherein data, representing a particular portion of the source image and occupying a computer memory location assigned to store that particular portion of the source image, is copied from that computer memory location into a computer memory location assigned to store a particular portion of the template image. The location of the template image portion within the template image typically differs from the location of the source image portion within the source image. The term "descrambling" and "unscrambling" refer to the reverse operation.

It is appreciated that the software components of the present invention may, if desired, be implemented in ROM (read-only memory) form.

In the present specification and claims, the terra "average" refers to any central tendency value of a population of values, such as the mean, mode or median thereof.

Any components of the present invention which are described as being implementable in software may, if desired, be implemented in ROM (read-only memory) form or in hardware, using conventional hardware design techniques.

Appendix A is a hexadecimal listing of a preferred software implementation of the scrambling method of Fig. 36 except that the opportunistic target feature, the watermark feature and the blank line compression feature are omitted.

Appendix B is a hexadecimal listing of a preferred software implementation of the unscrambling method of Fig. 37 except that the opportunistic target feature, the watermark feature and the blank line compression feature are omitted.

Appendix C includes source codes and interfaces useful for running the programs of Appendices A and B.

In order to run the scrambling and descrambling programs of Appendices A or B, using the source codes and interfaces of Appendix C, the following steps may be performed: 1. Using a word processor, type in the characters of the hexadecimal printout of Appendices A and B and save them as DOS text files. 2. Run the program entitled TXTTOEXE.C, the source code of which is provided in Appendix C, as follows: txttoexe < text file name> <exe file name> for each of the DOS text files produced in step 1. Call the two resulting executable files scramdll.dll and descrdll.ll, respectively. 3. Create an ASCII file named NORMAL_M.PRI, the contents of which is provided in Appendix C. 4. Create an interface (".h") file named SCRDLAPI.H, the source code of which appears in Appendix C . 5. Provide monochrome PCX files to serve as input for the scrambling and descrambling programs. 6. Invoke the scrambling program by running the program SCRAMBLE, the source code of which appears in Appendix C. 7. Invoke the descrambling program by running the program DESCRAMBLE, the source code of which appears in Appendix C. It is appreciated that the particular embodiment described in the Appendices is intended only to provide an extremely detailed disclosure of the present invention and is not intended to be limiting.

The embodiments of the present invention shown and described herein have been described in the context of scrambling and descrambling, which is an important application of the invention. However, it is appreciated that the invention is equally applicable to other situations in which the spatial arrangement of items of information on a document is of crucial importance but this arrangement is disturbed or may be disturbed due to operations performed on the document such as the operations illustrated in Fig. 15. The disturbance to the spatial arrangement of information typically comprises a distortion of the spatial arrangement of some of the information, relative to the remaining information.

Documents, other than scrambled documents, in which the spatial arrangement of items of information on the document is of crucial importance include scaled technical drawings, maps, color separations and printed circuit layouts. In all these applications, treating the document in the manner shown and described hereinabove, including scrambling the document and introducing registration marks therein, before it suffers the disturbance, allows the document to be accurately restored after it suffers the disturbance. Scrambling the document before it is processed reduces the damage caused by noise inherent in the process because it diffuses the damage caused by a local distortion. This is because gixels which are in the same vicinity during the process (e.g. during fax transmission) and therefore may be affected by a local distortion are not in the same vicinity in the restored document so that the effects of the local distortion are mitigated by being scattered throughout several locations within the document. It is appreciated that targets or registration marks may be superimposed onto data bearing locations within the scrambled image. Alternatively or in addition, targets or registration marks may be incorporated at additional, non-data bearing locations within the scrambled image. A target may be any recognizable item of information such as a predetermined shape or color. If the target is superimposed onto a data bearing location, the shape and/or color of the target is preferably easily distinguishable from the shapes and/or colors characterizing the data.

Throughout the drawings, thin lines are often employed to indicate gixel borders within scrambled or unscrambled images even though the images do not in fact include thin lines at these locations.

It is appreciated that the "scramble and fill" and "template generation" steps described above may be implemented based on one, some or all of the embodiments of Figs. 16 - 27B, since these implementations are not mutually exclusive but rather are independent improvements of the basic scrambling and filling step and the basic template generation step. These independent improvements have a synergistic effect when implemented all together.

It is appreciated that various features of the invention which are, for clarity, described in the contexts of separate embodiments may also be provided in combination in a single embodiment. Conversely, various features of the invention which are, for brevity, described in the context of a single embodiment may also be provided separately or in any suitable subcombination.

It will be appreciated by persons skilled in the art that the present invention is not limited to what has been particularly shown and described hereinabove. Rather, the scope of the present invention is defined only by the claims that follow: APPENDIX A

APPENDIX B

APPENDIX C

source code of SCRAMBLE

#include "scrdlapi.h" void scramble(void)

{

struct SCRDLAPI st;

/* fill structure st data V

strcpy(st.input_file_name,"input.pcx"); strcpy(st.output_file_name,"output.pcx"); st.program_path[0] = '\0'; st.serial_number[0] = '\0'; st.user_name[0] = '\0';

st.help_info[0]= '\0';

st.key_clue[0] = '\0';

st.scramble_code = 12345; st.page_in_document = 1; st.total_pages_in_document = 1;

st.dbeug_level = 0;

st.image_mark_offset = 0; st.always_use_default = 0; st.first_page_header_default = 400;

st.sub_page_header_default = 200;

st.page_struct = ARPS_NORMAL; st.start_of_data = 0;

st.left_margin = 0;

st.right_margin = 0;

st.scrambled_height = 0; st.marker_height = 0;

st.hParent = 0;

st.action_type = ONE_PAGE;

st. view = 0;

st.error = 0;

st.ddx = 0;

st.ddy = 0;

st.memory_check = 0;

st.noise_removal = ARNR_NONE; st.decrease_block_size = 0; scramdll(&st);

} source code of DESCRAMBLE

#include "scrdlapi.h" void descramble(void)

{

struct SCRDLAPI st;

/* fill structure st data as code in scramble above*/ descrdll(&st);

}

source code of SCRDLAPI.H

/*

* scrdlapi.h

*

* the definition of the structure SCRDLAPI which is used to connect

* with scramdll and descrdll

*/

typedef enum { ONE_PAGE, MULTI_PAGE, GET_SEPERATOR, GET_MARGINS } eaction_type; typedef enum { ARPS_NORMAL = 0, ARPS_TALL_NARROW = 1,

ARPS_SHORT_WIDE = 2 } epage_struct; typedef enum { ARNR_NONE = 0, ARNR_LINES = 1, ARNR_AVERAGE = 2 } enoise_removal; typedef struct {

char input_file_name[80];

char output_file_name[80];

char program_path[80];

char serial_number[20];

char user_name[40];

char help_info[30];

char key_clue[80];

long scramble_code; int page_in_document;

int total_pages_in_document;

int debug_level;

int image_mark_offset;

int always_use_default;

int first_page_header_deeault;

int sub_page_header_default; epage_struct page_struct; /* page structure - determines setup */ int start of data; /* bot torn row of unscrambled header */ int left_piargin; /* left margin for scrambled area */ int right_margin; /* right margin for scrambled area */ int scrambled_height; /* height of scrambled area */ int marker_height; /* height of strip used at the bottom */

/* for the general page targets */ unsigned int hParent; /* handle to parent window */ eaction_type action_type; /* what to do */

int view; /* confirm on view status */ int error; /* error code returned by DLL */

/* unscramble supplements */

int ddx;

int ddy;

int memory_check;

enoise_removal noise_removal;

int decrease_block_size;

} SCRDLAPI;

#if defined (_SCRAMBLE) | | defined (_DESCRAMBLE)

#ifdef _cplusplus

extern "C" {

#endif

#ifdef _SCRAMBLE

int _ far _pascal scramdll (SCRDLAPI_far *);

#endif

#ifdef _DESCRAMBLE .

int _far _pascal descrdll(SCRDLAPI _far *);

#endif void LoadSCRDLAPI(SCRDLAPI far *p);

#ifdef _cplusplus

}

#endif

; contents of NORMAL-M.PRI

;original page and supporting files

1 100 header start from top of page when seperation line not found

2 340 header size

3 116 label height

4 80 number of lines in bottom left code strip

5 0 16 0 block size (x), showing pavements

6 0 32 0 block size (y), showing pavements

7 82 number of blocks in a line

8 17 object size in seperation line, same dimention in x and y

9 1020 tolerance in resolution, x and y, in percent

; scrambled page

10 DYNLAB.img file name of label in scrambled page

11 11 number of fields in (x)

12 25 number of fields in (y)

13 8 number of blocks in each field (x)

14 2 number of blocks in each field (y)

15 1450 315 upper right code location (x,y), relative to header

16 200 25 bottom left code location (x,y), relative to bottom of fields

17 10 20 tolerance in resolution, x and y, in percent

18 10 targets size in scrambled page

19 15 code-8 cross size in scrambled page

; unscrambled page

20 labout.img file name of label in unscrambled page

21 0 label height (width is always 1728)

22 4 number of strips in raster buffer

#include <stdio.h>

#include <stdlib.h> int conv(char ch)

{

if ((ch >= '0') && (ch < = '9')) return (ch - 48) ; if ((ch >= 'A') && (ch < = 'F')) return (ch - 55) ; if ((ch >= ' a' ) && (ch <= 'f')) return (ch - 87); return - 1 ;

} void main (int argc,char **argv)

{

int num1,num2;

FILE *inf,*outf;

char ch1,ch2; if (argc<3) return;

if ((inf=fopen (argv [1], "rb")) ==NULL) return;

if ((outf=fopen (argv [2], "wb")) ==NULL) return; while (!feof(inf)) {

if (fscanf (inf, "%c%c", &ch1, &ch2) <2) continue; num1 = conv(ch1);

num2 = conv(ch2);

if ((num1<0) || (num2<0)) {

printf ("Error !\n");

return;

}

fprintf (outf, "%c", 16*num1+num2);

}

fclose(outf);

fclose(inf);

}

Claims

1. A method for scrambling and descrambling a source image

including a first plurality of first image portions, the method

comprising: generating a scrambled image by generating a second plurality of second image portions, wherein at least one of the second image portions in an individual location in the second plurality comprises a first image portion in a different location in the first plurality and a passpartout surrounding at least some of the first image portion.

2. A method according to claim 1 wherein at least one of the portions of the scrambled image contains target information which was not present in the source image.

3. A method for scrambling a source image while maintaining readability for a designated recipient, the method comprising:

scrambling the source image using a scrambling key which is identifiable by applying a confidential transformation to the key identifying information; and

generating and outputting a message including the scrambled image and the key identifying information.

4. A method according to claim 3 and also comprising automatically extracting the key identifying information from the message and automatically retrieving the scrambling key there*from.

5. A method for descrambling a scrambled image to reconstruct a source image comprising a plurality of image portions, at least one of which is blank, the method comprising:

aligning the scrambled image including recognizing at least one target marking on at least one of the blank image portions;

computing a correct location of the target marking within the scrambled image; and

rubber-sheeting at least a portion of the scrambled image to obtain correspondence between the actual location of the target marking and the correct location thereof.

6. A scrambling method comprising:

receiving a source image and generating a scrambled image of at least a portion of the source image;

providing at least one label image comprising label information which is not included in the source image;

generating an output image comprising a scrambled image of at least a portion of the source image and said label image.

7. A method according to claim 6 wherein said label information comprises descrambling information.

8. A scrambling method according to claim 6 wherein the label information comprises operator-readable information.

9. A scrambling method according to claim 6 wherein the label information comprises alphanumeric information.

10. A scrambling method according to any of claims 6 - 9 wherein the label information comprises machine-readable information.

11. An information bearing substrate comprising:

a scrambled image of at least a portion of a source image;

a label image comprising label information which is not included in the source image.

12. A document generating system comprising:

a user-generated information processor operative to transfer user-generated information onto a document; and

a document labeler operative to incorporate a label onto the document including processor identification information, not generated by the user, which is operative to identify the information processor.

13. A document generating system according to claim 12 wherein the information processor comprises an image scrambler.

14. A system according to either of claims 12 or 13 wherein the processor identification information comprises a serial number identifying the information processor.

15. A method for scrambling a source image comprising:

generating a template image including a plurality of reserved locations and a plurality of nonreserved locations, wherein at least some of the reserved locations "are arranged in a meaningful pattern;

copying a plurality of portions of the source image into the plurality of nonreserved locations respectively; and filling the reserved locations which form the pattern with information which is visually distinguishable from the information in the nonreserved locations.

16. A method according to claim 15 wherein the meaningful pattern comprises a non-grid pattern.

17. A method according to claim 15 or claim 16 wherein the meaningful pattern comprises the pattern of at least one character.

18. A method according to any of claims 15 or 16 and wherein the pattern identifies an individual user and also comprising authenticating the user based on the user-identifying pattern.

19. A method according to claim 3 or claim 4 wherein the key identifying information is generated by employing the designated recipient's public key to conceal the scrambling key.

20. A method for scrambling a source image comprising:

identifying blank portions in the source image;

replacing at least some of the blank portions in the source image with target-bearing portions which each bear a target whose location within the target-bearing portion is known to a descrambling processor; and

scrambling the modified source image.

21. A method for generating a descrambled image by descrambling a scrambled image, generated from a source image including at least one line of text, the method comprising: copying a rectangular portion located in an individual location within the scrambled image into a different individual location within an area allocated for representing the descrambled image, the rectangular portion having long and short edges, the short edges being generally parallel to the text lines and the long edges being generally orthogonal to the text lines; and repeating the copying step for additional rectangular portions.

22. A method according to claim 21 wherein the ratio between the lengths of the long and short edges exceeds 2.

23. A method according to claim 21 wherein the ratio between the lengths of the long and short edges exceeds 3.

24. A method for generating a descrambled image by descrambling a scrambled image, generated from a source image including at least one line of text, the method comprising:

copying a portion located in an individual location within the scrambled image into a different individual location within an area allocated for representing the descrambled image, the portion having edges which are generally parallel to the text lines and edges which are generally orthogonal to the text lines, the length of each orthogonal edge being an integer multiple of the pitch of the text lines; and repeating the copying step for additional scrambled image portions.

25. A method for scrambling a source image including a first plurality of first pixel arrays, the method comprising: compressing the source image by identifying at least one sequence of blank first pixel arrays, deleting said sequence and adding an indication of the length of the sequence; and

scrambling the compressed source image by generating a second plurality of second pixel arrays, wherein at least one of the second pixel arrays in an individual location in the second plurality comprises a first pixel array in a different location in the first plurality.

26. A method for scrambling a source image comprising:

generating a template image including a plurality of reserved locations and a plurality of nonreserved locations;

copying a plurality of portions of the source image into the plurality of nonreserved locations respectively; and

filling at least some of the reserved locations with hollow registration marks.

27. A method according to claim 26 wherein the source image comprises a text and wherein each copied portion of the source image is similar in size to the size of the text characters.

28. A method according to claim 26 or claim 27 wherein the size of the cavity in each hollow registration mark exceeds the size of the cavities of the text characters.

29. A method for descrambling documents using a facsimile machine and a computer, having a fax modem and being in telephone communication with the facsimile machine, the method comprising: providing the computer with a scanning error correction and descrambling function;

feeding a scrambled document into the facsimile machine and communicating its contents as a facsimile message via the telephone to the computer; and

employing the computer's correction/descrambling function to correct for scanning errors of the facsimile machine and to descramble the document.

30. A method according to claim 29 wherein the sending step comprises sending the contents of the scrambled document to the computer via a telephone line simulator.

31. A method according to claim 29 wherein the sending step comprises sending the contents of the scrambled document to the computer via the serial port of the computer.

32. Portable image modification apparatus operative in conjunction with a facsimile machine and comprising:

a connector operative to transmit a digital representation of an image to and from the facsimile machine; and

33. Apparatus according to claim 32 wherein the dedicated image modifier has no graphics display.

34. Apparatus according to claim 32 or 33 wherein the dedicated image modifier has no alphanumeric input device .

35. Apparatus according to claim 32 or 33 wherein the dedicated image modifier comprises an image scrambler.

36. Apparatus according to claim 32 or 33 wherein the dedicated image modifier comprises an image descrambler.

37. Apparatus according to claim 32 or 33 wherein the image modifier comprises an optical character recognition device.

38. Apparatus according to claim 32 or 33 wherein the connector is operative to transmit dialed numbers from the facsimile machine and wherein the image modifier is operative to perform image modification on the basis of information conveyed by the dialed numbers.

39. A mass-distributed printed product comprising:

a scrambled image of confidential information characterized in that the confidential information is accessible to designated recipients who possess a descrambling key by scanning and descrambling the scrambled image.

40. A method for mass dissemination of confidential information to a first multiplicity of designated recipients while maintaining confidentiality versus a second multiplicity of nondesignated recipients, the method comprising:

providing a descrambling key to said first multiplicity of designated recipients; and

separately disseminating a scrambled image of confidential information to said first and second multiplicities of recipients.

41. A method according to claim 40 and also comprising:

scanning the scrambled image; and

descrambling the scanned image.

42. A method according to claim 40 or claim 41 wherein said disseminating comprises publishing in a mass-disseminated printed medium.

43. A method for cracking a scrambled graphics image comprising a plurality of scrambled image portions, the method comprising:

forming a descrambled area by identifying a neighbor, in the source image, of a first scrambled image portion along a first edge thereof from among other scrambled image portions by identifying a scrambled image portion having an edge whose cross-correlation with the first edge is high; and

enlarging the descrambled area by repeating, at least once, the forming step so as to identify additional source image neighbors of the image portions within the descrambled area.

44. Apparatus according to claim 32 wherein the image modifier comprises a sketch straightener.

45. Apparatus according to claim 32 wherein the image modifier comprises an image merger operative to merge an input image with a stored backdrop image.

46. A method for scrambling a text image comprising:

loading into a memory a first swathe of the image including a single row of gixels in which a plurality of gixels is arranged in a first sequence; generating output signals representing the same portion of the image represented by the swathe by outputting a second single row of gixels in which said plurality of gixels is arranged in a second sequence; and

repeating the loading and generating steps for at least one more swathe of the image.

47. Apparatus for descrambling documents comprising:

a scrambled document signal source, providing signals that represent the contents of a document, receiving a document to be unscrambled and providing output signals representing the contents of the document;

an unscrambler for operating on the output signals to produce modified output signals representing a unscrambled version of the document, the unscrambler employing a system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation; and

an unscrambled document generator receiving said modified output signals and producing a unscrambled version of the document.

48. Apparatus according to claim 47 and also comprising a descrambling verifier for verification of correct descrambling.

49. Apparatus according to claim 48 and wherein said descrambling verifier comprises an edge correlator for carrying out edge correlation operations on adjacent scrambled pixels.

50. Apparatus according to claim 47 and wherein said unscrambler is controlled by a coded input to provide a selected one from a plurality of possible modifications to the output signals.

51. Apparatus according to claim 47 when embodied in a photocopier.

52. Apparatus according to claim 47 when embodied in a telefax.

53. Apparatus according to claim 47 when embodied in a computer.

54. Apparatus according to claim 47 and wherein said unscrambler is operative to change the relative positions of multi-pixel regions of a document, without modifying the information content within each of the multi-pixel regions.

55. Apparatus according to claim 47 and wherein said unscrambler is operative to leave unchanged certain predetermined regions of the document.

56. Apparatus for descrambling documents for use in conjunction with a document receiver and a scrambled document signal source receiving a document to be unscrambled and providing output signals representing the contents of the document, the apparatus comprising:

an unscrambler operating on the output signals to produce modified output signals representing a unscrambled version of the document, the unscrambler employing said system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation; and to provide the modified output signals to the document generator, for producing an unscrambled version of the document.

57. A method for descrambling documents comprising the steps of: receiving a document to be unscrambled including a system of reference marks and providing output signals representing the contents of the document;

operating on the output signals to produce modified output signals representing a unscrambled version of the document;

employing said system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation; and

receiving said modified output signals and producing a corrected unscrambled version of the document.

58. A method according to claim 57 and wherein said reference marks are distributed throughout the area of the document.

59. A method according to claim 57 and wherein said reference marks are distributed about the periphery of the area of the document.

60. A method according to claim 57 and wherein the correct location of each of said reference marks is determined by its appearance.

61. A method according to claim 57 and wherein said reference marks are characterized by a local color modification of the image information thereat.

62. A method according to claim 57 and wherein said reference marks are arranged in a predetermined grid.

63. Apparatus according to claim 47 and wherein said reference marks are distributed throughout the area of the document.

64. Apparatus according to claim 47 and wherein said reference marks are distributed about the periphery of the area of the document.

65. Apparatus according to claim 47 and wherein the correct location of each of said reference marks is determined by its appearance.

66. Apparatus according to claim 47 and wherein said reference marks are characterized by a local color modification of the image information thereat.

67. Apparatus according to claim 47 and wherein said reference marks are arranged in a predetermined grid.

68. A method for descrambling documents for use in conjunction with a document receiver and a scrambled document signal source receiving a document to be unscrambled including reference marks and providing output signals representing the contents of the document, the method comprising the steps of:

providing the modified output signals to the document generator, for producing an unscrambled version of the document.

69. A method according to claim 57 wherein said step of employing comprises:

employing the reference marks to determine an average gixel size; and

employing the reference marks to find a reference point for each of a plurality of gixels in the descrambled image and copying a vicinity of average gixel size around the reference point into an area reserved for the unscrambled image, at a location determined by the key which generated the scrambled image.

70. Apparatus for descrambling documents comprising:

a document signal source, providing signals that represent the contents of a disturbed document, receiving a document to be restored and providing output signals representing the disturbed contents of the document;

a document restoring module operating on the output signals to produce modified output signals representing a restored version of the document and employing a system of reference marks for at least one of registration, scaling, rotation, shifting and defect compensation; and

a restored document generator receiving said modified output signals and producing a restored version of the document.

71. A method for performing a distorting process on a document bearing a spatial arrangement of information so as to minimize distortion of the spatial information, the method comprising:

scrambling an input document bearing a spatial arrangement of information;

performing a distorting process on the input document, thereby to obtain an output document; and

descrambling the output document.

72. A method according to claim 71 wherein said scrambling comprises incorporating targets into the input document.

73. A method according to either of claims 71 and 72 wherein the document bearing a spatial arrangement of information includes one of the following: a technical drawing, a map, a color separation, and a printed circuit layout.

74. A method according to either of claims 71 and 72 wherein the distorting process comprises scanning of the document and subsequent printing thereof.