WO2006061941A1

WO2006061941A1 - Image reading device

Info

Publication number: WO2006061941A1
Application number: PCT/JP2005/018076
Authority: WO
Inventors: Takehiko Saitoh; Masaaki Yoshida
Original assignee: Seiko Instruments Inc.
Priority date: 2004-12-08
Filing date: 2005-09-30
Publication date: 2006-06-15
Also published as: JP2006166106A

Abstract

An image reading device reads out a document by using a plurality of reading units having a plurality of light receiving elements arrayed in a main scanning direction. Read pixel data on the portion where the pixels in the main scanning direction of adjoining CIS overlap is subjected to an averaging operation, in which the output values of individual light receiving elements at the same position in the main scanning direction are weighted, and the weighting coefficients are not made identical in an overlap range. The image reading device thus provided can suppress the image quality deterioration at the joints of the reading units and has a high reading image quality.

Description

Image reading device

Technical field

TECHNICAL FIELD [0001] The present invention relates to an image reading apparatus, and more particularly, to reading of a document using a plurality of reading units in which a plurality of light receiving elements are arranged in the main scanning direction, and joining image data between the reading units. The present invention relates to an image reading apparatus that improves image quality accurately.

Background art

In a conventional image reading apparatus, as shown in Japanese Patent Application Laid-Open No. 60-31357, a light receiving element in which reading light is irradiated onto an original and reflected light from the original is arranged in a line in the main scanning direction. Image data was obtained by photoelectric conversion in the reading unit that also has a coercive force. As the reading unit, CIS, which is a 1X sensor, is used because of its simple structure and easy miniaturization. In CIS, light receiving elements are usually arranged at intervals of 300dpi to 1200dpi in order to obtain high quality images. The light receiving element combines an optical component such as a self-occ lens and a photoelectric conversion element such as a phototransistor, and the reflected light from the document surface passes through the self-occ lens and forms an image on the photoelectric conversion unit. Up to A4 and A3 sizes, CIS is relatively easy to manufacture and can be supplied at low cost due to its mass production effect. However, a photo detector that covers only Al and AO size documents with a single CIS. It is difficult to manufacture, and the yield becomes difficult, and as a result, it becomes expensive. Therefore, a large-size scanning device such as A1 or AO forms a sensor unit by scanning multiple CISs for scanning A4 or A3 size documents, and reads large-size documents in a single document feed. Making things possible. When a plurality of CISs are arranged in one row in the main scanning direction, as shown in FIG. 15, the distance Ly between the light receiving element at one CIS end and the light receiving element at the other adjacent CIS end is one. Since it will be longer than the distance (Lx) between the light receiving elements of the CIS, it is necessary to perform processing such as interpolating the pixels between the CIS after reading so that the unnaturalness of the joints is not revealed. is there. However, this pixel interpolation is not correct because the pixel to be interpolated is estimated from the pixel information read in the vicinity. Therefore, as shown in Figure 3, The CISs are arranged at different positions in the sub-scanning direction, the ends of adjacent CISs are overlapped, and further, on the upstream side, the read image data is delayed using delay means, and then the downstream side A method is used that eliminates white space at the joints by combining with image data. A specific example of such conventional joining will be described below with reference to FIG. Note that the sub-scanning direction described in this specification refers to the document conveyance direction, and the main scanning direction refers to the light receiving elements in the reading unit that are perpendicular to the sub-scanning direction. It is the direction of being.

[0003] Each of CIS (A) 101 and CIS (B) 102 in Fig. 16 (a) is composed of X light receiving elements. The CIS (A) 101 (X-3) th 111th to Xth 116th light receiving elements overlap the CIS (B) 102 first 121th power to fourth 124th light receiving elements in the main scanning direction. In CIS (A) 101, the (X-2) th 114th light receiving elements are valid, and (X-1) 115, the Xth 116th light receiving elements are invalid. Similarly, in CIS (B) 102, the light receiving element from the third 123 is made valid, the first 121 and the second 122 light receiving elements are made invalid, and they are connected as shown below.

• CIS (A) 101 (X—4) th light receiving element 112, CIS (A) 101 (X—3) th light receiving element 113, CIS (A) 101 (X—2) 1st light receiving element 114, CIS (B) 102 1st light receiving element 121, CIS (B) 102 2nd light receiving element 122, CIS (B) 102 3rd light receiving element 123,

Such a joining method eliminates the need for pixel interpolation when the reading units are arranged in one row in the main scanning direction, and makes it possible to reproduce the original more faithfully at the joints.

[0004] However, even in such an image reading apparatus in which the ends of the CIS overlap, the joints of the read images may become unnatural. As shown in Fig. 16 (b), the CIS overlaps with each other due to slight variations in the CIS mounting accuracy and slight expansion / contraction due to the temperature of the plate to which the CIS is mounted as the reading resolution increases. A Lx shift occurs in the light receiving element that should be at the same position in the main scanning direction. This A Lx made the image joint between CIS unnatural, and when reading natural images, the quality of the read image deteriorated, such as the lines appearing to join. main Regarding the variation in the scanning direction, it is possible to adjust by the unit of the light receiving element interval by changing the position of the pixel on the CIS to be connected, but it cannot be adjusted below the interval of the light receiving element on the CIS. Due to this non-absorbable variation, the joints of the images are displaced, leaving unnaturalness and degrading the image quality.

[0005] In the conventional image reading apparatus using a plurality of CISs described above, the image quality at the joint is poor! ,When! There was a problem.

[0006] Therefore, the present invention is intended to solve the problems of the conventional configuration as described above, and eliminates the unnaturalness of the image at the joint and realizes an image reading apparatus with high read image quality. The purpose is to do.

Disclosure of the invention

[0007] In order to achieve the above object, according to the present invention, read pixel data relating to a portion where pixels in the main scanning direction of adjacent CISs overlap at the time of stitching are respectively received in the same position in the main scanning direction. The output value is averaged by weighting, and the weighting coefficient is not the same in the overlap range. In other words, in each of the reading sections to be connected, the gradation values of the pixels around the joint are read, and both read gradation values are included, so that the discontinuity of the gradation of the joint is made inconspicuous. Is possible.

[0008] Further, according to the present invention, the gradation value obtained by one light receiving element force and the other light receiving element force obtained from a pixel in a portion where the light receiving elements of adjacent reading units overlap each other are expressed by the following arithmetic expressions: This is an image reading apparatus that is used for calculation.

Nout (x) = Na (x) X (ga (x) / G) + Nb (x) X (gb (x) / G)

G = ga (x) + gb (x)

x: Pixel position in the main scanning direction of the joined images

Nout (x): The gradation value after calculation of the pixel at the position in the main scanning direction X.

Na (X): The gradation value of the pixel of one reading unit at the position in the main scanning direction X.

Nb (X): The gradation value of the pixel of the other reading unit at the position in the main scanning direction X.

ga (X): A weighting coefficient of the pixel of one reading unit at the position in the main scanning direction X. gb (X): Weighting coefficient of the pixel of the other reading unit at the position in the main scanning direction X. G: integer greater than or equal to l

Further, according to the present invention, the weighting coefficient used when calculating the pixels in the overlapping range in the main scanning direction of the adjacent reading unit is larger on the side close to one reading unit, and the weighting coefficient for the gradation value of one reading unit is increased. However, on the side closer to the other reading unit, the image reading apparatus has a larger weighting coefficient for the gradation value of the other reading unit.

[0009] Further, according to the present invention, in the overlapping range in the main scanning direction of the reading unit, in the overlapping range closer to the reading unit, the weighting coefficient of the reading value of the CIS is set to the main of the adjacent reading unit. In the joint processing of the range that is larger than the weight of the reading gradation value at the same scanning position and conversely overlaps in the main scanning direction closer to the adjacent reading section, the weighting coefficient of the reading gradation value of the reading section Is made smaller than the weighting coefficient of the reading value at the same position in the main scanning of the adjacent reading section, so that the unnatural gradation of the range to be joined can be reduced.

Brief Description of Drawings

FIG. 1 is a configuration diagram of a reading apparatus according to the present invention.

FIG. 2 is a diagram showing a positional relationship between a staggered arrangement of a reading unit and a document according to the present invention.

FIG. 3 is a diagram showing a configuration of a sensor unit according to the present invention.

FIG. 4 is a diagram showing a configuration of a reading unit according to the present invention.

FIG. 5 is a block diagram showing an electrical configuration of a reading unit according to the present invention.

6 is a diagram showing a flow of the data processing unit shown in FIG.

FIG. 7 is a diagram showing the overlap of light receiving elements of adjacent reading units.

8 is a diagram showing specific locations of symbols indicating the positions of the respective reading units shown in FIG.

FIG. 9 is a diagram showing the correspondence between the position of the light receiving element and the weighting coefficient in the sensor joining process.

FIG. 10 is a diagram showing a weighting expression between CIS (A) and CIS (B) in the example.

FIG. 11 is a diagram showing a weighting calculation formula between CIS (B) and CIS (C) in the embodiment.

FIG. 12 is a diagram showing a weighting expression between CIS (C) and CIS (D) in the example.

FIG. 13 is a diagram showing a weighting expression between CIS (D) and CIS (E) in the example.

[FIG. 14] An operation for calculating read data in a range not included in FIG. 10, FIG. 11, FIG. 12, and FIG. It is a figure which shows a formula.

FIG. 15 is a diagram showing the arrangement of a conventional reading unit.

FIG. 16 is a diagram for explaining joining of conventional reading units.

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described with reference to FIGS.

As shown in FIG. 4, the reading unit called CIS 41 has 5000 light receiving elements 42 arranged in one row in the main running direction. The light receiving element 42 also has a combination force of an optical component such as a self-occ lens (not shown) and a photoelectric conversion element such as a phototransistor (not shown), and the reflected light from the original is coupled to the light receiving surface of the photoelectric conversion element by the self-occ lens. I have an image. Furthermore, three light guides are arranged in the CIS 41 at a position parallel to the arrangement direction of the light receiving elements. The light guides illuminate the original with red, blue, and green light, respectively, by LEDs that emit red, blue, and green light.

As shown in FIG. 3, the sensor unit fixes five reading units 21 to 25 called CIS on the plate 31 so as to form a staggered arrangement. Each CIS reads the document by transporting the document 26 in the document transport direction 27 on the sensor unit. Fig. 2 shows the positional relationship between each CIS and the manuscript from the perspective of Fig. 3 from above. As can be seen from Figs. 2 and 3, CIS (A) 21, CIS (C) 23, and CIS (E) 25 are arranged at the same position in the sub-scanning direction so that they do not overlap in the main scanning direction. CIS (A), (C), (CIS (B) 22 and CIS (D) 24 are located downstream of the three CISs in the sub-scanning direction at the same position in the sub-scanning direction. E) and arranged in a staggered arrangement. In CIS (A) 21 and CIS (B) 22, CIS (B) 22 and CIS (C) 23, CIS (C) 23 and CIS (D) 24, CIS (D) 24 and CIS (E) 25 In this embodiment, the light receiving elements are arranged so as to overlap each other in the main scanning direction.

[0014] Next, the flow of CIS output data on the sensor unit will be described with reference to the block diagram of FIG. The outputs of CIS (A) 21, CIS (B) 22, CIS (C) 23, CIS (D) 24, and CIS (E) 25 are input to AD converter 51. The output of the AD converter 51 is input to the data processing unit 52. An image memory 55 and a non-volatile memory 56 are connected to the data processing unit 52, and read / write can be performed by the data processing unit 52. The output of the data processing unit 52 is input to the image processing unit 57. The output of the image processing unit 57 is connected to the outside of the scanner. The CPU 53 is connected to the data processing unit 52, and the CPU 53 controls the data processing unit 52. The CPU 53 is connected to a program memory 54 and stores a program that defines the operation of the CPU 53.

[0015] For each CIS, the sensor unit starts from the trigger pulse input from the data processing unit 52, and outputs the output values from the light receiving elements in order from the first to the 5000th on each CIS. Is output as an analog signal in synchronization with the rising edge of the CIS reference clock not shown. The AD converter 51 samples in accordance with the rising edge of the CIS reference clock, performs A / D conversion, and outputs it as 8-bit digital data. The data processing unit 52 includes a gate array, a cell-based ASIC, an FPGA, and the like, receives digital data output from the AD converter, and performs shading correction. A specific method of shading correction will be described later. In addition, when the upstream CIS outputs read data for the p-th line, the downstream CIS outputs read data for the p-q line. Therefore, the data processing unit temporarily stores the data after the upstream CIS processing in the image memory 55. After conveying the document for q lines, the CIS on the downstream side reads the p line, and after AD conversion in the data processing unit, the data on the same line by the upstream CIS is read from the memory, and the inter-sensor connection processing is performed. Do. Through the stitching process, the data that has been processed by dividing the data into 5 pieces so far becomes one line of data and sent to the image processing unit.

[0016] The image processing unit is configured by a DSP (Digital Signal Processor) or the like, and performs image processing such as contrast adjustment and gamma correction, cut-out processing in the main scanning direction and sub-scanning direction, enlargement / reduction, 2 Processing such as digitization is performed as necessary, then sent to the outside of the reading device, and stored as image data, printed out, or remotely via LAN or WAN, depending on the application. It is transferred to the ground.

The flow of the data processing unit will be described below with reference to FIG. The data input from the AD converter 51 enters the FIF061 in order to absorb the difference between the transfer rate between the AD converter 51 and the data processing unit 52 and the transfer rate in the data processing unit. The in-read section processing unit 62 reads data from the FIF061 and does not require a CIS sensor as will be described later. The data for 100 pixels at both ends is discarded and sent to the shading correction unit. Shading correction is calculated by the following formula when the number of gradations is 256 to widen the dynamic range of the read data.

Data after shading correction = 255 X ((N (X) Bk (X)) Z (Wh (x) Bk (x))) N (x): The AD converter output at the position of x when the document is scanned Read data Bk (X): Read data of AD converter output at X position with CIS LED off

Wh (x): Data read from AD converter output at position X when white reference plate is read

Bk (x) used for shading correction stores the AD conversion value of the light receiving element output in the nonvolatile memory 56 with the CIS LED turned off before reading the document, and Wh (x) reads the same document. Is obtained by storing the AD conversion value of the output of the light receiving element when the white reference plate provided in the apparatus is read in the nonvolatile memory 56, and from the nonvolatile memory 56 at the time of shading correction, the Bk (x ), Wh (x) is read and shading correction is performed.

The data after the shading correction is written into the image memory 55 for delay processing by the memory controller 64. The delay processing unit 65 reads the data corresponding to the same line on the document via the memory controller 64 and puts it in the inter-sensor connection processing unit 66 in consideration of the shift of the transport time in the sub-scanning direction of the CIS arranged in a staggered manner. It is done. The operation of the inter-sensor connection processing unit 66 will be described later.

[0019] The data after the linking process is sent to the main scanning direction data cutout processing unit 67, and only the range in the main scanning direction specified by the CPU is extracted and sent to the output FIF068. The output FIF 068 reads out data from the output FIF 068 and transfers it to the outside of the data processing unit 52 so as to meet the transfer rate between the data processing unit 52 and the image processing unit, and the image processing unit 57 performs the above-described processing.

A detailed example of the inter-sensor connection process will be described below.

[0021] In each CIS, the 100 pixels at both ends are used for adjustment when the deviation in the main scanning direction between adjacent CISs is greater than the gap between the light receiving elements. When adjustment is not required, the output for 100 pixels at the end is invalid.

Therefore, from the 101st light receiving element, 4800 light receiving elements including the 4900th light receiving element are effective. If there are adjacent CISs, the light receiving elements for 64 pixels at the end of 4800s, that is, the 101st to 164th and 4737th to 4800th light receiving elements are combined in the CIS. In the stitching calculation, weighting calculation is performed on the 101st received light data of one CIS and the 4737th received light data of the other CIS, and the data of that pixel is calculated. Similarly, the weighting operation is performed on the 102nd and 4738th of one CIS, the 164th and 4800th of the other.

[0023] Therefore, in CIS (A) and CIS (E), 4800 light receiving elements are effective. Of these, 64 are used for connecting to adjacent CIS, and the light receiving elements that have not undergone the connecting process are There are 4736 pieces. In CIS (B), CIS (C), and CIS (D), which require joint processing at both ends, 4800 light receiving elements are effective, of which 128 are connected to adjacent CIS. There are 4672 light-receiving elements that are not connected and processed. Therefore, the number of pixels in one line after joining is 4736 + 64 +4672 + 64 +4672 + 64 +4672 + 64 +4736 = 23744.

[0024] In the weighting calculation in the stitching process, when the gradation of one light receiving element is Na and the gradation of the other light receiving element is Nb, the calculation is performed by the following equation.

Nout = ((ga / G) X Na) + (((G -ga) / G) X Nb)

Nout: Gradation after stitching processing

G: integer greater than or equal to l

ga: Weighting factor consisting of an integer between 1 and G

In this embodiment, G is 9 and the stitching range of 64 pixels is divided into 8 small areas, and the same weighting coefficient is used in each small area. Here, G is not limited to 9, and if the above subdivision areas are subdivided, the number of subdivision areas will be more than that. For example, if the stitching range of 64 pixels is divided into 64 small areas, the G will be a number of 64 or more, and if the stitching range is 64 pixels or more, it will be larger. Number. As the subdivision area and G are divided into subdivision areas, the larger the G, the more natural the seam of images can be realized. The amount of data that needs to be increased.

[0025] The weighting coefficient ga is an integer of 1 or more and G or less, and becomes G when added to the weighting coefficient gb of the other CIS at the same position in the main scanning direction.

In addition, the weighting coefficient ga (x) at the position X in the main scanning direction within one reading unit and the weighting coefficient ga (y) at the position y adjacent to the position X and close to the other reading unit within the same reading unit , Ga (x) ≥ga (y) is satisfied. In other words, as the light receiving element is closer to the end of the CIS, the weighting factor becomes smaller and a more natural read image can be obtained.

[0026] FIG. 7 (a) shows the positional relationship between the joints of CIS (A) and CIS (B). In the figure, alO to al8 indicate the positions of the light receiving elements of CIS (A). bl to b9 indicate the position of the light receiving element of CIS (B). Similarly, Fig. 7 (b) shows CIS (B) and CIS (C), Fig. 7 (c) shows CIS (C) and CIS (D), and Fig. 7 (d) shows CIS (D) and CIS (D). Show me about CIS (E)!

[0027] Symbols in Fig. 7 (al0 to al8, bl to b9, bl0 to bl8, cl to c9, cl0 to cl8, dl to d9, dl 0 to dl8, el to e9) indicate which light receiving element on each CIS. Figure 8 shows whether they are specified.

[0028] Further, FIG. 9 shows the weighting coefficient for each small area that requires the weighting calculation shown in FIG. 7 and FIG.

[0029] FIGS. 10 to 14 show calculation formulas for pixel data of one line after stitching using the weighting coefficients of FIG.

[0030] By these joining processes, the peripheral pixels between the reading units include both reading data at the same position in the main scanning direction, so CIS manufacturing variations and CIS sensor unit mounting accuracy It is possible to eliminate the unnaturalness of the scanned image that occurs due to the stitching of the scanning unit, which occurs due to variations in the image quality.

Example 1

FIG. 1 shows a configuration of a reading apparatus according to an embodiment of the present invention.

In FIG. 1, reference numeral 1 denotes a sensor unit that irradiates a document with reading light and reads the reflected light. Reference numeral 3 denotes a white platen placed at a position facing the sensor unit 1 with the document table and the document table 3 interposed therebetween. Fix reference plate 2. A document transport path 4 is between the document table 3 and the white reference plate 2. Reference numerals 5a and 5b are transmission sensors that detect the insertion of an original when an external force original is inserted into a document carry-in locuser (not shown). For transporting documents to the platen 3 when document insertion is detected There are a driving roller 6a and a driven roller 6b for loading. 8a and 8b are sensors for detecting the conveyance of the document to the document table 3, 7a is a driving roller for conveying the document, and 7b is a driven roller for conveying the document. Reference numeral 9a denotes a discharge driving roller for discharging the original after reading out of the apparatus, and 9b denotes a discharge driven roller.

Next, the operation of the embodiment shown in FIG. 1 will be described. Before reading the document, read the black reference value with sensor unit 1 turned off, turn on the LED on sensor unit 1, read white reference plate 2, and store the parameters for shading correction. A document inserted from a document loading locuser (not shown) is detected by the transmission sensors 5a and 5b, and until the transmission sensors 8a and 8b detect the leading edge of the document by the loading drive roller 6a and the driven driven roller 6b. Sent to 3. The document is read by the sensor unit 1 while the document is being conveyed by the document feeding drive roller 7a and the document feeding driven roller 7b. When reading of the predetermined line is completed, or when the transmission sensors 5a and 5b detect the trailing edge of the document, the sensor unit 1 finishes reading, and is driven by the discharge drive roller 9a and the discharge driven roller 9b. , Discharged outside the device.

Industrial applicability

[0034] As described above, the reading device of the present invention can provide an image reading device that can eliminate the unnaturalness of the read image that occurs when the reading unit is joined.

Claims

The scope of the claims

[1] By irradiating the original with reading light and photoelectrically converting the reflected light with the light receiving element, the reading part that reads the image of the original is used as a part of the light receiving element in the adjacent reading part. At least two or more are arranged at different positions in the main scanning direction so as to overlap with each other, and the reading unit moves the sensor units or the originals, which are the reading unit forces, in the sub-scanning direction using moving means. By scanning the original, the image data read by the reading unit positioned upstream and the image data read by the reading unit downstream are combined using a combining unit, and one line is read. In the image reading apparatus that generates image data, the synthesizing unit obtains a gradation value in a range in which the light receiving elements overlap in the main scanning direction of the adjacent reading unit from the reading unit located upstream. A gradation image and a reading unit located downstream are provided with joint image processing means for calculating by weighted averaging processing of the obtained gradation values, and the weighting coefficient used for the averaging processing calculation is the main scanning. An image reading apparatus, wherein the overlapping range differs depending on a direction position.

2. The image reading apparatus according to claim 1, wherein the image reading apparatus fixes the sensor unit and moves the document using the document moving means.

3. The image reading apparatus according to claim 1, wherein the image reading apparatus fixes a document and moves the sensor unit using a sensor unit moving means.

[4] The image according to any one of [1] to [3], wherein the reading unit is arranged at the same position in each of the odd-numbered and even-numbered main scanning directions in the sub-scanning direction. Reading device.

[5] The image reading apparatus according to any one of claims 1 to 4, wherein the weighted averaging processing calculation is calculated by the following equation.

Nout (x) = Na (x) X (ga (x) / G) + Nb (x) X (gb (x) / G)

G = ga (x) + gb (x)

x: Pixel position in the main scanning direction of the joined images

Na (X): The gradation value of the pixel of one reading unit at the position in the main scanning direction X. Nb (x): The gradation value of the pixel of the other reading unit at the position in the main scanning direction x.

ga (X): A weighting coefficient of the pixel of one reading unit at the position in the main scanning direction X. gb (X): Weighting coefficient of the pixel of the other reading unit at the position in the main scanning direction X.

G: integer greater than or equal to l

The weighting coefficient used for the averaging is the weighting coefficient ga (x) of the pixel located at the position X close to the center of one reading section and the other adjacent to X in the adjacent pixels in the main scanning direction in one reading section. 6. The image reading device according to claim 1, wherein a weighting coefficient ga (y) of a pixel located at a position y close to the reading unit satisfies ga (x) ≥ga (y). .