WO2009073726A1

WO2009073726A1 - Systems and methods for rapid three-dimensional modeling with real façade texture

Info

Publication number: WO2009073726A1
Application number: PCT/US2008/085410
Authority: WO
Inventors: Joseph G. Freund; Ran Gal
Original assignee: Pictometry International Corporation; Idan Computers Engineering Ltd.
Priority date: 2007-12-03
Filing date: 2008-12-03
Publication date: 2009-06-11
Also published as: US10229532B2; US9520000B2; US20150170413A1; US9275496B2; US20160180583A1; US20130293542A1; US10896540B2; US10573069B2; US20140125664A1; US20180261006A1; US11263808B2; US8970615B2; US20170193694A1; US20200193702A1; US20210209843A1; US20180089891A1; US20190206125A1; US9836882B2; US8648872B2; US20090141020A1

Abstract

Methods are described for automatically generating a three-dimensional model of a structure including real façade textures obtained from geo-referenced oblique images. The geographical position of at least one real façade texture is located. Geo-referenced images containing the real façade texture are located. A base image having the real façade texture is selected from the geo-referenced images by analyzing image raster content. The real façade texture of the base image is applied to the three-dimensional model. Systems for generating the three-dimensional model are also described herein.

Description

SYSTEMS AND METHODS FOR RAPID THREE-DIMENSIONAL MODELING WITH REAL FAQADE TEXTURE

Cross-Reference to Related Applications

[0001] This application claims the benefit of U.S. Patent Application Serial Number 11/998,974, filed on December 3, 2007, the contents of which are incorporated herein by reference as if set forth in its entirety.

Background of the Invention

[0002] Technology advancements within the three-dimensional modeling industry are providing avenues for physical simulation of real-life and hypothetical situations on computer systems. These models can provide valuable information for strategic and tactical planning. For example, three-dimensional models of city streets can provide first responders information regarding current city developments including entryway locations, building recognition, and the like. This information is valuable in reducing response time during emergency conditions. Further, emergency personal can train for emergency situations through simulated scenarios provided by or with the three-dimensional models. [0003] Currently within the art, oblique images obtained from aerial photographs are used to provide close-to-accurate representations of each building's surface within the three- dimensional model. However, generally, it is difficult to obtain these oblique images with unobstructed views of the building's surface. For instance, most buildings in downtown regions of a major metropolitan city are close in proximity to one another. It becomes burdensome and time consuming to capture aerial images of each building without an obstruction, such as a neighboring building. Further, finding a single image without obstructions from the thousands of aerial images obtained, would be extremely time consuming and cost-prohibitive and may require three-dimensional modeling of all obstructing structures.

[0004] Some three-dimensional models edit the obstructed portion of the oblique image by approximating the building's surface using other portions of the same building. Although this method does provide a representation of the building within the three-dimensional model, the representation goes on the assumption that all portions of the building are created equal. However, this assumption is problematic as an obstructed area may have uniquely placed doorways and/or windows that may be ignored by the approximation. [0005] Alternatively, a user can manually manipulate two or more oblique images to form a single image having an unobstructed view of the faςade. This type of manual manipulation is slow and tedious, and requires experience and expertise in the modeling field. Brief Summary of the Invention

[0006] In one embodiment, the present invention is related to a method of automatically generating a three-dimensional model of a structure. The three-dimensional model of the structure preferably includes real fagade textures obtained from geo-referenced oblique images. The geographical position of the real fagade texture is provided using wire-frame data. A base oblique image is selected from geo-referenced oblique images using the geographical positions obtained from the wire-frame data. The base oblique image is selected from the geo-referenced oblique images based on analysis of the image raster content of the real fagade texture, and the real fagade texture of the base oblique image is then applied to the three-dimensional model.

[0007] In one version, the real fagade texture is analyzed to locate at least one occlusion. Occlusions can include obstructions in the view of the real fagade texture such as an overlapping building, image distortions within the base oblique images, and/or the like. Occlusions may be located using pattern recognition, contrast, and/or the like. Unoccluded image characteristics of the real fagade texture are provided by at least one geo-referenced oblique image. The unoccluded image characteristics of the occlusion are applied to the real fagade texture to form a mosaic image.

[0008] In another version, the real fagade texture is analyzed and corrected for misalignment. Misalignment may be corrected by shrinking and/or stretching the real fagade texture, extracting portions of the real fagade texture, and/or the like. For example, the outer boundaries of the real fagade texture may be extracted using discontinuities in depth, discontinuities in surface orientation, variations in scene illumination, and/or the like. In another version, the wire-frame data of the three-dimensional model is analyzed to locate the geographical position of a roof of the structure. Images containing the roof are provided and a base image is selected and applied to the three-dimensional model. The images may be nadir images or oblique images. Selection of the base image is based on the image raster content.

[0009] In another embodiment, the present invention is related to a method of automatically generating a three-dimensional model having structures with real fagade textures. The real fagade textures are obtained from geo-referenced aerial oblique images. Wire-frame data is analyzed to locate geographical positions of the real fagade textures of the structures within a geographical area. An oblique image showing the real fagade texture is selected. Where present, at least one incorrect area within at least a portion of the real fagade texture may be identified. The incorrect area within the portion of the real fagade texture is automatically corrected. Brief Description of the Several Views of the Drawing

[0010] So the above-recited features and advantages of the present invention may be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments thereof that are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting in scope, for the invention may admit to other equally-effective embodiments. [0011] Figure 1 is a block diagram of one embodiment of the rapid three-dimensional modeler system having real-facade textures obtained from geo-referenced images in accordance with the present invention.

[0012] Figure 2 is a pictorial representation of a three-dimensional model having real facade textures.

[0013] Figure 3 is a pictorial representation of wire-frame data providing a geographical position of a real facade texture of a structure.

[0014] Figure 4 is a pictorial representation of terrain data providing a geographical position of a real faςade texture of a structure.

[0015] Figure 5 is a pictorial representation of an exemplary oblique image obtained from an aerial camera, the oblique image displaying a real facade texture of a structure. [0016] Figure 6 is an exemplary function flow chart of a method for generating a three-dimensional model of a structure including real faςade textures obtained from geo-referenced images.

[0017] Figure 7 is a schematic block diagram of the formation of a mosaic image from a base image and a geo-referenced image including an unoccluded image characteristic. [0018] Figure 8 is a schematic block diagram of the formation of a mosaic image from a base image and a geo-referenced image including an unoccluded image characteristic, the base image and the geo-referenced image having different views.

[0019] Figure 9 is a schematic diagram of an exemplary embodiment of a system for providing three-dimensional models having real faςade textures in accordance with the present invention.

Detailed Description of Exemplary Embodiments

[0020] Embodiments of the invention are shown in the above-identified Figures and described in detail below. In describing the embodiments, like or identical reference numerals are used to identify common or similar elements. The Figures are not necessarily to scale and certain features and certain views of the Figures may be shown exaggerated in scale or in schematic in the interest of clarity and conciseness.

[0021] Referring now to the drawings and in particular to Figures 1 and 2, shown therein and designated by a reference numeral 10 is an exemplary system for rapidly creating a three- dimensional model 12 of a structure 14 including real facade textures 16. The system 10 provides real fagade textures 16 obtained from geo-referenced images 18. The three- dimensional model 12, including the real fagade textures 16, provides an accurate representation of a geographic area. An exemplary three-dimensional model 12 is shown in Figure 2 and preferably includes structures such as buildings, thoroughfares, and other associated fixtures. Such features provide real-life representation of the geographical area useful in strategic planning, tactical planning, tactical debriefing, simulation, real-time simulation, first response actions, engineering designs, and/or the like. Although the geographic area applies to physical characteristics of an area, it is not limited to the topographical features of the Earth's surface. For example, the system and methods described herein may apply to rapid three-dimensional modeling techniques for medical imaging.

[0022] Generally, the system 10 identifies a geographical position 22 of at least one real facade texture 16 of the structure 14. The system 10 locates geo-referenced images 18 containing the real faςade texture 16 and selects a base image 19 (See Figure 7) having optimal image characteristics of the real fagade texture 16. This base image 19 provides at least one real fagade texture applied to the structure 14 forming the three-dimensional model 12.

[0023] The system 10 identifies the geographical position 22 of the real fagade textures 16 within the geo-referenced images 18. In one embodiment, the system 10 uses wire-frame data 20 of the structure 14 to provide the geographical position 22 through the identification of boundaries (17a, 17b, 17c, 17d ...) of the real fagade texture 16, as best illustrated in Figure 3. Although the wire-frame data 20 of Figure 3 illustrates boundaries 17a-17d of the real fagade texture 16 as a rectangle, it will be appreciated by one skilled in the art, the boundaries 17a-17d may comprise a circle, square, triangle, or any fanciful shape. Alternatively, an edge-detection algorithm may be used to locate the geographical position 22 of the real fagade texture 16 within the wire-frame data. Preferably, wire-frame data 20 is obtained from publicly available information of buildings, structures, elevations and the like. For example, publicly available wire-frame data 20 commonly stored as ^*.shp files may be used. Alternatively, wire-frame data 20 of the structure 14 may be created based on the particular application of the present invention using techniques commonly known within the art.

[0024] The system 10 may also use terrain data 24 to provide the geographical position 22 of the real fagade texture 16. For example, the system may identify boundaries (17a, 17b, 17c, 17d ...) of the real fagade texture 16 as illustrated in Figure 4. Although Figure 4 illustrates boundaries 17a-17d of the real fagade texture 16 as a rectangle, it will be appreciated by one skilled in the art, the boundaries 17a-17d may comprise a circle, square, triangle, or any fanciful shape. Alternatively, an edge-detection algorithm may be used to locate the geographical position 22 of the real faςade texture 16 within the terrain data. Terrain data may include Digital Terrain Models (DTMs), Digital Elevation Models (DEMs), and/or the like. Generally, terrain data 24 is comprised of sets of universal terrain map coordinates identifying location and elevation of geographical regions. The coordinate system may be any coordinate system including latitude and longitude or any other geographical coordinate system suitable for use with the present invention. Preferably, the terrain data 24 is a publicly available Digital Terrain Model. Alternatively, terrain data 24 for a geographical area may be extracted or created from publicly available contour images, and/or may be created for the specific application intended.

[0025] The geo-referenced image 18 providing the real facade texture 16, as illustrated in Figure 5, may use any color space and be stored in any industry supported file format, such as TIFF, JPEG, GIF, BMP, ECW, and/or the like. Additionally, the geo-referenced images preferably contain or are referenced to or coordinated with information regarding the location and orientation of the camera, focal length, the physical size of the sensor, and/or the like. [0026] The geo-referenced images 18 are preferably nadir images and/or oblique images. Nadir images, as described herein, provide vertical or orthogonal views of the geographic area. Nadir images may be an image captured from an overhead position, a position above the structure 14, or at a right angle or an angle near to a right angle to the structure 14. For example, the image may be taken from an overhead position as related to the structure 14 at an eighty-seven degree angle or similar angle close to a ninety-degree angle. A nadir image may be rectified to fit an associated wire-frame, DTM, DEM, and/or the like. Rectifying a nadir image may entail mapping the pixels to the coordinates of the wire-frame, DTM, DEM, and/or the like.

[0027] As nadir images typically do not express the details and characteristics of structures and objects, in the currently preferred embodiment, oblique images will typically, but not exclusively, be used for the purposes of the present invention. Oblique images, as described herein, are images taken at an angle other than that of the nadir perspective or images derived from the nadir perspective. Oblique images provide a perspective line of sight that generally reveals information not visible from an orthogonal or orthophoto view. For example, an oblique image may have an angle to the structure 14 in the range of about zero degrees to about eighty-nine degrees.

[0028] Referring to Figures 1 and 6, the first step 50 in generating a three-dimensional model having real faςade textures 16 is to obtain a relevant geographic location of the structure 14. Once the geographic location is determined, a set of geo-referenced images 18 representing the location are obtained, which is the second step 52. The geo-referenced images 18 may be obtained as discussed above or in any other suitable fashion. The relevant geographical location 22 may be determined internally, calculating a position based upon a mathematical formula, or through an external entity, for example, by a database or user. Preferably, a user indicates an initial geographic location, for example, such that subsequent locations may be mathematically calculated using image boundaries. [0029] The third step 54 is to locate the geographical position 22 of at least one real fagade texture 16 of the structure 14. Wire-frame data 20 and/or terrain data may be used to locate the geographical position 22 coordinating to the structure 14 as previously discussed herein. [0030] The fourth step 56 is to select a base image 19 containing the real fagade texture 16 from the geo-referenced images 18. The base image 19 is preferably selected as one that accurately represents the real fagade texture 16 and may be automatically selected by the system 10 or selected or determined by the user from a limited selection of geo-referenced images 18 provided by the system 10.

[0031] The base image 19 containing the real fagade texture 16 is selected from the geo- referenced images 18 of the structure 14 based upon a factorial analysis of the image raster content of each geo-referenced image 18 of the structure 14. The factorial analysis may include a weighted determination based on the resolution of the image, colour depth of the image, proportional size of the real fagade texture, contrast, time of day the image was captured, time of year the image was captured and/or the like. Foliage may also play a role in the factorial analysis. For example, the base image 19 may be selected based on the contrast in lighting conditions. Contrast may be measured as a histogram and provide information regarding shadows within the image. Such information is relevant in constructing replications of large geographical areas, as the three-dimensional model 12 would generally include real fagade textures 16 representing the same approximate time of day and/or time of year.

[0032] The fifth step is to apply or relate the real fagade texture 16 of the base image 19 to the three-dimensional model 12. The three-dimensional model 12 may comprise base factors such as wire-frame data and/or terrain data. For example, the real fagade texture 16 of the base image 19 may be applied to the wire-frame of the structure 14. The system 10 may automatically rotate, stretch, or shrink the real fagade texture 16 to align with edges of the wire-frame data 20. Additionally, the system 10 may provide for user evaluation and/or manipulation of the real fagade texture 16.

[0033] Alternatively, wire-frame data 20 and/or terrain data 22 may provide a coordinate system for the three-dimensional model 12 such that the real fagade texture 16 is geographically positioned within the three-dimensional model according to the corresponding coordinates of the wire-frame data and/or terrain data. Additionally, the real fagade texture 16 may be applied to form the entire three-dimensional model 12 or, alternatively, at least a portion of the three-dimensional model 12. Multiple base images 19 with multiple real facade textures 16 may form the three-dimensional model 12.

[0034] In another embodiment, the real facade texture may be further processed to correct occlusions 70, as illustrated in Figures 7 and 8. Generally, it is difficult to provide the real facade texture 16 of the base image 19 of the structure 14 without occlusions 70 such as overlapping structures, visual distortions, and/or the like. For example, Figure 7 illustrates the real faςade texture 16 blocked by an overlapping structure 71. This blocked overlapping structure 71 causes an occlusion 70 within the real faςade texture 16. Multiple geo- referenced images 18 may not be able to provide the real faςade texture 16 without the blocked overlapping structure 71. Therefore, methods are provided to correct for the occlusion 70 within the real faςade texture 16.

[0035] To correct for the occlusion 70 within the real faςade texture 16, the base image 19 is analyzed to locate a geographical position 72 of the occlusion 70. For example, the geographical position 72 may be identified using boundaries (74a, 74b, 74c, 74d...). Although the boundaries 74a-74d are illustrated as a rectangle, it will be appreciated by one skilled in the art, the boundaries 74a-74d may comprise a circle, square, triangle, or any fanciful shape. Additionally, an edge-detection algorithm may be used to identify the geographical position of the occlusion 70.

[0036] Once the geographical position 72 of the occlusion 70 is located, the geo-referenced images 18 of the structure 14 are analyzed to locate at least one image 76 having an unoccluded image characteristic 78. As illustrated in Figure 7, the geo-referenced images 18 of the structure 14 correcting for the occlusion 70 are not required to be the same view as the base image 19.

[0037] The unoccluded image characteristic is applied to the real faςade texture 16 forming a mosaic image 80. For example, pixels of the unoccluded image characteristic 78 may replace pixels of the occlusion 70 of the real faςade texture 16. The mosaic image 80 may include several unoccluded image characteristics 78 in addition to the real faςade texture 16 of the base image 19.

[0038] Storage of the mosaic image 80 may be problematic if the file size of the mosaic image 80 is extremely large. Thus, it may be beneficial to use an algorithm for storing the mosaic image 80. For example, a rectangular area of the mosaic image 80 can be extracted and all information outside of the rectangular space cropped and removed. However, the geo-referencing information of the area would still be maintained within the mosaic image 80.

[0039] Occlusions 70 are not limited to blocked areas on the real faςade texture 16 but can also include misalignment of the real faςade texture 16 when applied to the three-dimensional model 12 and/or undesirable background images within the real faςade texture 16. Occlusions 70, misalignments, and background images may be identified using pattern recognition, contrast, and/or the like. For example, pattern recognition technically can be used to analyze each pixel or a group of pixels to determine whether the pixel belongs to the real fagade texture 16 or is part of the occlusion 70. The three-dimensional geometry of the surrounding buildings may also be a key factor.

[0040] Additionally, the real fagade texture 16 may be optimized for depth complexity. Methods for optimizing depth complexity involve the system 10 first determining whether overlapping elements within the real fagade texture 16 are occlusions 70 or desirable features of the real fagade texture 16. If the system 10 determines the overlapping elements are desirable features of the real fagade texture 16, the system 10 applies the real fagade texture 16 of the overlapping structures to the three-dimensional model. For example, the system 10 may detect a small roof structure that is overlapping a large roof structure. The system 10 will assume that the small roof structure is an element on top of the large roof structure and considered desirably within the real fagade texture 16. The system will apply the real fagade texture 16 of the overlapping structures to the three-dimensional model 12. Depth complexity analyses the geometry of overlapping 3D structures, such as a small room on a large roof. Usually, walls are represented by vertical rectangles, from the top of the roof to the terrain. In this case, walls of the small room are created only to the large roof and not to the terrain.

[0041] The formation of the three-dimensional model 12 with real fagade textures 16 may be a continuous or intermittent process. Preferably, the formation of the three-dimensional model 12 is automatically performed by the system 10 to facilitate rapid modeling of an area. Using the methods as described herein, the system 10 determines the geographic location of an area of interest and retrieves the appropriate wire-frame data 20 and/or terrain data 24 associated with the area. The system 10 automatically locates and identifies structures 14 within the area based upon the wire-frame data 20 and/or terrain data 24 provided. Geo- referenced images 18 of the area are located by the system 10 using the wire-frame data 20 and/or terrain data 24. The base oblique image 19 of the structure 14 is automatically selected by the system 10 based on the image raster content of each of the geo-referenced images 18. If the base oblique image 19 contains more than one structure 14, the system 10 may locate each real fagade texture 16 independently or together. Once the real fagade texture 16 is located, the system 10 may fix any occlusions if necessary and apply the real fagade texture to the three-dimensional model 12.

[0042] Generally, the system 10 is a computer system that is able to embody and/or execute the logic of the processes described herein. The logic embodied may be executed on any appropriate hardware such as, for example, a dedicated system or systems, personal computer system, distributed processing computer system, and/or the like. [0043] Referring now to Figure 9, the system 10 is preferably distributed, and includes a host system 112, communicating with one or more user devices 114 via a network 116. The network 116 can be the Internet or other network. In either case, the host system 112 typically includes one or more servers 118 configured to communicate with the network 116 via one or more gateways 120. When the network 116 is the Internet, the primary user interface of the system 10 is delivered through a series of web pages, but the primary user interface can be replaced by another type of interface, such as a Windows-based application. This method is also used when deploying the system 10 in a stand-alone environment, such as a kiosk.

[0044] The network 116 can be almost any type of network, although Internet and Internet 2 networks are preferred because of the wide support of their underlying technologies. One embodiment of the network 116 exists in an Internet environment, which means a TCP/IP- based network. It is conceivable that in the near future, the preferred or other embodiments, may wish to use more advanced networking topologies.

[0045] The servers 120 can be networked with a LAN 130. The gateway 120 is an entity responsible for providing access between the LAN 130 and the network 116. The gateway 120 can also be used as a security means to protect the LAN 130 from attack from external networks such as the network 116.

[0046] The LAN 130 network can be based on TCP/IP network such as the Internet, or it can based on another underlying network transport technology. The preferred embodiment uses an Ethernet network with TCP/IP because of the availability and acceptance of underlying technologies, but other embodiments may use other types of networks such as Fibre Channel, SCSI, Gigabit Ethernet, etc.

[0047] As discussed above, in one embodiment, the host system 112 includes the servers 118. The configuration of the server hardware will depend upon the requirements and needs of the particular embodiment of the system 10. Typical embodiments, including the preferred embodiment, will include multiple servers 118 with load balancing to increase stability and availability. It is envisioned that the servers 118 will include database servers and application/web servers. The database servers are preferably separated from the application/web servers to improve availability and to provide the database servers with improved hardware and storage.

[0048] The user devices 14 can be any number and type of device. Generally, user devices 14 involve a user 32, using a computer 34 with a display 36, keyboard 38, and mouse 40. It is contemplated user devices 14 may include a touch screen element on the display in addition to or in lieu of the keyboard 38 and/or mouse 40.

[0049] Typically, the user device 14 uses a type of software called a "browser" as indicated by reference numeral 42 to render HTML/XHTML content that is generated when requesting resources from a source, such as the host system 112. In the preferred embodiment, the system 10 is designed to be compatible with major Web Browser vendors (Microsoft Internet Explorer, Netscape Navigator, and Opera). Other embodiments may wish to focus on one particular browser depending upon the common user base using the system 10. It should be noted, user devices 14 may interact with the system 10 through any suitable functional software, such as a program specifically designed for the individual embodiments of the system 10.

[0050] The user devices 14 can also be implemented as a portable device such as a laptop computer 150 (or handheld computer); a cellular telephone 52 with a micro or embedded Web Browser; a Portable Digital Assistant 54 (PDA) capable of wireless network access; a pen-based or tablet computer 56, and/or the like. In another embodiment, the user device 14 may be a cable box 60 or other similar device for viewing through a display 62 or television. Current embodiments of the system 10 can also be modified to use any of these or similar future developed devices.

[0051] The system 10 is designed to provide flexibility in its deployment. Depending upon the requirements of the particular embodiment, the engine may be designed to work in almost any environment such as a desktop application, a web application, a series of web services designed to communicate with an external application, and/or the like. [0052] The hardware and system software are designed with two key concerns: flexibility and scalability. Although some specifics for software and hardware components are described herein, it will be understood that a wide array of different components may be substituted. For example, different database vendors may be used, SML-based document stores may be used, and/or the like.

[0053] When the system 10 is used to execute the logic of the processes described herein, such computer(s) and/or execution may be conducted at a same geographical location or multiple geographic locations. Furthermore, the execution of the logic may be conducted continuously or at multiple discrete times.

[0054] The system 10 includes one or more computer readable medium storing instructions for displaying a pixel representation of one or more of the images described herein including the geo-referenced images 18, the base oblique images 19, wire frame data 20, terrain data 24, the three-dimensional model 12, and the like. The computer readable medium may be part of the host system 112, the user devices 114, or combination thereof. Additionally, the system 10 uses one or more databases or servers 118 to store the images in an organized format. For example, the geo-referenced images 18 of the structure 14 may be sorted and stored by the direction showing the particular real faςade texture 16.

[0055] The system 10 may also include instructions for (1) displaying pixel representations of the images as described above; (2) selecting the geographical location of interest; (3) altering and/or editing images; and (4) other similar tasks. The instructions typically run on a combination of the user devices 114 and the host system 12.

[0056] Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be obvious to those skilled in the art that certain changes and modifications may be practiced without departing from the spirit and scope thereof, as described in this specification and as defined in the appended claims below.

Claims

What is claimed is:

1. A method for automatically generating a three-dimensional model of a structure including real fagade textures obtained from geo-referenced oblique images, comprising the steps of: analyzing wire-frame data of the three-dimensional model of the structure to locate geographical positions of at least one real facade texture of the structure; locating oblique images containing the real facade texture; selecting a base oblique image from the oblique images by analyzing image raster content of the real fagade texture of each of the oblique images; and, applying the real fagade texture of the base oblique image to the three-dimensional model.

2. The method of claim 1 , further comprising the steps of: analyzing the real fagade texture to locate geographical positions of at least one occlusion; locating at least one oblique image having an unoccluded image characteristic of the occlusion in the real fagade texture; and, applying the unoccluded image characteristic to the real fagade texture forming a mosaic image.

3. The method of claim 2, wherein the occlusion is located using pixel pattern recognition of the real fagade texture.

4. The method of claim 2, wherein the oblique images are geo-referenced solved oblique images.

5. The method of claim 2, further comprising the step of removing the occlusion from the real fagade texture.

6. The method of claim 1 , further comprising the step of extracting the real fagade texture from the base oblique image, the real fagade texture having geo-referenced information.

7. The method of claim 1 , further comprising the steps of: analyzing wire-frame data to locate the geographical position of a roof of the structure; locating images containing the roof; selecting a base image from the images containing the roof by analyzing image raster content; and, applying the roof to the three-dimensional model.

8. The method of claim 7, wherein the images containing the roof are nadir images.

9. The method of claim 7, wherein the images containing the roof are oblique images.

10. The method of claim 7, wherein the images containing the roof are nadir images and oblique images.

11. The method of claim 7, wherein the base image contains a greater proportion of pixels associated with the roof as compared with other oblique images.

12. The method of claim 1 , wherein the oblique images are geo-referenced solved oblique images.

13. The method of claim 1, further comprising the step of correcting misalignment of the real facade texture.

14. The method of claim 13, wherein the misalignment is corrected using an edge- detection algorithm.

15. The method of claim 1 , further comprising the step of aligning the real fagade texture to the wire-frame data of the three-dimensional model.

16. The method of claim 1 , further comprising the step of aligning the wire-frame data to the facade.

17. A method for automatically generating a three-dimensional model of structures including real facade textures obtained from geo-referenced aerial oblique images, comprising the steps of: analyzing wire-frame data to locate geographical positions of real fagade textures of the structures within a geographical area; choosing an oblique image showing the real fagade texture; identifying an incorrect area within the part of the real fagade texture; automatically correcting the incorrect area within the part of the real facade texture.

18. A computer system for automatically generating a three-dimensional model of structures including real facade textures obtained from geo-referenced aerial oblique images, the system comprising: one or more computer readable medium storing wire-frame data indicative of a three- dimensional model of one or more structures and further storing a plurality of geo-referenced aerial oblique images wherein the images substantially correspond to the geographical location of the structures; an input device receiving input from a user specifying a geographical location of the one or more structures in which to generate a three-dimensional model including real facade textures; a processor analyzing the wire-frame data of the three-dimensional model of the one or more structures to locate geographical positions of at least one real fagade texture of the one or more structures, locating oblique images containing the real faςade texture, selecting a base oblique image from the oblique images by analyzing image raster content of the real faςade texture of each of the oblique images, and then applying the real fagade texture of the base oblique image to the three-dimensional model of the one or more structures; and a display device displaying the three-dimensional model including real faςade textures to the user.

19. A system for automatically generating a three-dimensional model of structures including real faςade textures obtained from geo-referenced aerial oblique images, the system comprising: a user device comprising an input device, an output device, and a communications system, the input device receiving input from a user specifying a geographical location of one or more structures in which to generate a three-dimensional model including real faςade textures, the communications system transmitting the geographical location of the one or more structures and receiving the generated three-dimensional model including real faςade textures, the output device displaying the three-dimensional model including real faςade textures to the user; and a host system receiving the geographical location of the one or more structures from the user device, the host system comprising one or more computer readable medium, a processor, and a communications system, the one or more computer readable medium storing wire-frame data indicative of a three- dimensional model of the one or more structures and further storing a plurality of geo-referenced aerial oblique images wherein the oblique images substantially correspond to the geographical location of the structures, the processor analyzing the wire-frame data of the three-dimensional model of the one or more structures to locate geographical positions of at least one real facade texture of the one or more structures, locating oblique images containing the real facade texture, selecting a base oblique image from the oblique images by analyzing image raster content of the real fagade texture of each of the oblique images, and then applying the real fagade texture of the base oblique image to the three-dimensional model of the one or more structures, the communication system transmitting the generated three- dimensional model including real faςade textures to the user device.

20. A computer program product for automatically generating a three-dimensional model of structures including real fagade textures obtained from geo-referenced aerial oblique images, the computer program product comprising: code for analyzing wire-frame data of the three-dimensional model of the structure to locate geographical positions of at least one real facade texture of the structure; code for locating oblique images containing the real fagade texture; code for selecting a base oblique image from the oblique images by analyzing image raster content of the real fagade texture of each of the oblique images; and, code for applying the real fagade texture of the base oblique image to the three- dimensional model.