JP2000242810A - Method and device for texture mapping, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically map optimum texture to respective polygons of a three-dimensional structure model by selecting a video image having the optimum texture, cutting the texture out of the video image, and automatically pasting the texture on the corresponding polygon. SOLUTION: The three-dimensional structure model is read in (SO3) and one object polygon is selected from the read-in three-dimensional structure model (SO4). Further, a video image, etc., which does not have the texture corresponding to the selected object polygon is excluded from search images, which are narrowed down (SO5). The video image having the best texture is selected out of the narrowed-down search images (SO6) and read in (SO7), and the texturecorresponding to the corresponding polygon is cut out of the read-in video image (SO8) and pasted on the object polygon (SO9).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for constructing a three-dimensional virtual space, and more particularly to a computer-executable method and apparatus for texture-mapping a real image to a three-dimensional structural model and a processing procedure of the texture mapping method. The present invention relates to a recording medium on which various programs are recorded.

[0002]

2. Description of the Related Art In recent years, studies for generating a three-dimensional virtual environment using a real image have been actively conducted. We are conducting research aiming at three-dimensional reproduction of urban landscapes. When reproducing cityscapes, it is necessary to reproduce a large number of complex and numerous buildings over a wide range, and these must be reproduced efficiently and accurately.

Conventionally, a target object is photographed with a camera, and the object is viewed using a computer graphic editor.
A skilled worker has pasted a texture cut out from a captured image one by one to each polygon of the three-dimensional structure model. Furthermore, if necessary, a photographed image is digitized, and a photographed image is shaped, modified, and edited by a skilled person.

[0004]

SUMMARY OF THE INVENTION A task of photographing a real image of a texture point with a camera, a task of selecting an image showing an optimal texture from a photographed image, and cutting and editing a texture from the selected image. There is a problem that the work of pasting to each polygon of the three-dimensional structure model requires skill and requires an enormous amount of time.

[0005] In addition, since images are taken one by one, it is extremely difficult to obtain a real image from all camera orientations and positions, and there is a problem that many textures cannot be obtained.

An object of the present invention is to provide a technique capable of automatically mapping an optimal texture to each polygon of a three-dimensional structure model.

Another object of the present invention is to perform a conventional operation of photographing a real image of a texture point with a camera, an operation of selecting an image on which an optimal texture appears from a photographed image, and a method of extracting a texture from the selected image. Cut, edit, 3
An object of the present invention is to provide a technique capable of greatly reducing and reducing the amount of work to be attached to each polygon of a dimensional structure model and the work time.

Another object of the present invention is to provide a technique capable of easily obtaining a uniquely optimal texture from a large number of images. The above and other objects and novel features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

[0009]

SUMMARY OF THE INVENTION Among the inventions disclosed in the present application, the outline of a representative one will be briefly described.
It is as follows. (1) In a method of automatically texture-mapping a photographed image taken by a camera with respect to a three-dimensional structure model having position information, position and orientation data of a video camera at the time of photographing is read, and the read data is read. Estimate the position and orientation of the video camera at the time of live-action shooting from
A target polygon is selected from the three-dimensional structure model, a range of a video image to be searched is narrowed based on a relative positional relationship between the selected target polygon and a video camera, and an optimal texture is reflected from the narrowed video image. In this method, a selected video image is selected, a texture is cut from the selected video image, and the cut texture is automatically pasted to a corresponding polygon.

(2) In the texture mapping method of (1), the processing procedure for estimating the position and orientation of the video camera at the time of actual photographing includes reading the highly accurate camera position and orientation data estimated at certain intervals, The number of frames of the video camera image existing between the read camera position / posture data is calculated, and the position / posture information (X, Y, Z, κ) of the video camera that has captured the video image corresponding to the relevant frame is calculated. , Φ, ω), store the calculated value in the video image information, and repeat until the processing procedure is completed for all of the camera position / posture data estimated at the constant interval. is there.

(3) In the texture mapping method of (1) or (2), the processing procedure for narrowing down the range of the video image to be searched based on the relative positional relationship between the three-dimensional structure model and the video camera is as follows. One data is read from the video image information storing the posture data,
One vertex of the selected target polygon is selected, a distance between the selected vertex and the photographing point is calculated, vertex coordinates of a texture surface corresponding to the selected vertex are calculated, and an image number, a vertex number, The coordinates, the distance to the photographing point are stored, and the above-described processing procedure is performed on all vertices, and all or only a part of the texture and the image may be displayed in the video image. Distinguishes from the image, removes the image that shows all or only part of the texture from the texture information, and maintains the center of gravity of the target polygon for the image that may have the whole texture in the video image 1 included in the distance
One polygon is selected, one vertex in the selected polygon is selected, the coordinates of a point on the video image plane corresponding to the selected vertex are calculated, and the corresponding point is included in the target texture plane. If the corresponding point is within the texture plane, the distance L1 between the selected vertex and the shooting point is calculated, and the distance L2 between the center of gravity of the target polygon and the shooting point is calculated. By comparing the two, the distance L1
Is shorter than the texture information, the information relating to the current video image is deleted from the texture information, the above processing is performed on all vertices, and the processing is performed on all polygons within a certain distance from the center of gravity of the target polygon. This is a method for processing.

(4) In the texture mapping method of any one of (1) to (3), the video image selection processing procedure reads one piece of data from texture information and calculates an area of a texture surface. Calculate the distance between the center of gravity of the textured surface and the principal point, compare the calculated area of the textured surface and the distance between the center of gravity of the textured surface and the principal point, and obtain the video image having the largest area, or the area concerned. In the case where there are a plurality of video images having the maximum distance, the video image having the minimum distance is uniquely selected.

(5) A recording medium for recording a program for causing a computer to execute all the procedures in the texture mapping method of (1). (6) A recording medium storing a program for causing a computer to execute all the procedures in the texture mapping method of (2).

(7) A recording medium for recording a program for causing a computer to execute all procedures in the texture mapping method of (3). (8) A recording medium storing a program for causing a computer to execute all the procedures in the texture mapping method of (4).

(9) A central processing unit having a temporary storage means, and an external storage device in which a real image taken by a video camera is stored. A texture mapping apparatus for automatically texture-mapping an image, a position / posture data reading means for reading position / posture data from a photographed image taken by a video camera from the external storage device, and a position of the video camera at the time of photographing the photographed image. Position / posture estimating means for estimating the position and orientation, and a search video image range narrowing means for narrowing the range of the search video image from the relative positional relationship between the three-dimensional structure model and the video camera;
A video image selecting means for selecting a video image on which an optimal texture is displayed from the video images narrowed down by the search video image range narrowing means, and a texture cutout for cutting a texture from the video image selected by the video image selecting means And a pasting means for automatically pasting a texture cut by the texture cutting means to a corresponding polygon.

(10) In the texture mapping apparatus of (9), a real image taken by a video camera provided with a high-resolution camera is stored in an external storage device, and the position and orientation of the video camera are determined based on the position and orientation of the high-resolution camera. It is a device that acquires the posture. (11) The texture mapping device according to (9) or (10) is mounted on an automobile or a helicopter, and a photographed image captured by a video camera is stored in an external storage device.
This is a device that acquires the position and orientation of the video camera from the position and orientation of the high-resolution camera.

Hereinafter, the present invention will be described in detail together with embodiments (examples) with reference to the drawings. In all the drawings for describing the embodiments, components having the same function are denoted by the same reference numerals, and their repeated description will be omitted.

[0018]

(Embodiment 1) FIG. 1 is a block diagram showing a schematic configuration of a texture mapping apparatus according to Embodiment 1 of the present invention. The texture mapping device according to the first embodiment is a device that automatically texture-maps a photographed image captured by a video camera with respect to a three-dimensional structure model having position information, as shown in FIG.
It comprises an external storage device 1, a central processing unit (CPU) 2, and an internal storage device 3 having a temporary storage area.

The external storage device 1 includes a camera position / posture data storage area 101, a three-dimensional structure model storage area 10
2, a video image storage area 103 and a textured three-dimensional structure model storage area 104.

The central processing unit 2 is provided with the aforementioned area 1
Data reading means 201 for reading various data stored in the external storage device 1,
Video camera position / posture estimating means 202 for estimating the position and orientation at the time of video camera shooting based on the camera position / posture data read from the camera position / posture data storage area 101, and 3 read out from the three-dimensional structure model storage area 102. Search image narrowing means 203 for narrowing the range of a video image for performing a texture search using the three-dimensional structure model and the video camera position and attitude data estimated by the video camera position and attitude estimation means 202, and this search image narrowing means 203 Selecting texture 204 for selecting an optimal texture from the video images narrowed down by the method, texture cutting means 205 for cutting the texture from the video image read from the video image storage area 103, and reading from the three-dimensional structure model storage area 102 Three-dimensional structure Texture mapping means pasting textures cut the model each polygon 2
06 and a data writing unit 207 for storing the three-dimensional structure model to which the texture is attached in the textured three-dimensional structure model storage area 104 of the external storage device 1.

The internal storage device 3 has a video camera position / posture data storage area 301 for storing data obtained from the video camera position / posture estimation means 202 and a texture information storage area for storing data obtained from the search image narrowing means 203. 302.

FIG. 2 is a diagram for explaining the principle of the estimating method in the position / posture estimating means 202 of the video camera according to the first embodiment. In FIG. 3, A and B are camera position / posture data used for estimation. The A data and the B data can be regarded as sampling data of the video camera position / posture data. In the present invention, the video camera position / posture is estimated based on the data. Specifically, it is assumed that the video camera moves linearly between the A data and the B data, and the position and orientation of each video camera are estimated using the value obtained by dividing the difference between the two by the number of frames N as the amount of change.

K (K = 1, 2, 3,... N) between the A data (A frame) and the B data (B frame)
The position / posture information of the third frame is as shown in Expression 1.

[0024]

(Equation 1)

FIG. 3 is a diagram for explaining the principle of a method of reading a video image in the search image narrowing means 203 according to the first embodiment. In the present invention, a video image in which all or a part of the target texture is shown (FIG. 4A) and a video image in which the target texture is hidden by other textures (FIG. 4C) The search range of the optimal texture is narrowed by excluding it from the video image for acquiring the texture.

FIG. 4 is a diagram for explaining the principle of the method of selecting a video image in the texture selection method 204 according to the first embodiment. Reference numeral 51 denotes a projection center or a shooting point of the video camera, 52 denotes a video image plane which is an image plane reflected on the video camera, and 53 denotes a plane which is reflected on a video image plane 52 obtained by taking a three-dimensional structural model from the shooting point 51. A certain texture surface, 54 is the center of gravity of the texture surface 53, and 55 is the photographing point 51.
The principal point, which is the intersection of the perpendicular from
6 is a target polygon which is a surface in the three-dimensional structure model;
Is a three-dimensional structural model. Further, the target polygon 56
Each vertex of the texture surface 53 corresponding to each vertex is referred to as a corresponding point.

In the first embodiment, an image in which the area of the texture surface 53 is maximized is defined as a video image having an optimal texture. Further, the texture surface 53 having the same area is used.
Is present in multiple video images, the texture surface 5
The video image in which the distance between the area 54 of No. 3 and the principal point 55 is minimized is the video image having the optimal texture. This is a direction in which the center of gravity of the target polygon 56 coincides with the principal point 55, and the photographing point 51 where the area of the texture surface 53 is maximized.
Means that the image taken from has the optimal texture.

FIG. 5 is a flowchart for explaining the operation of the entire texture mapping apparatus according to the first embodiment. As shown in FIG. 5, the texture mapping apparatus according to the first embodiment uses a camera position / position as a basis of position / posture data.
The posture data is read (S01), and the position / posture of the video camera is estimated from the read camera position / posture data (S02). Next, a three-dimensional structural model is read (S03), and one target polygon is selected from the read three-dimensional structural model (S04). Furthermore, a video image not showing the texture corresponding to the selected target polygon, a video image showing only a part of the texture, and a video image not showing all the textures hidden by other textures are excluded from the search image. (S05), a video image having an optimal texture is selected from the narrowed search images (S06), the selected video image is read (S07), and the read video image is read. , A texture corresponding to the corresponding polygon is cut out (S08), and the cut out texture is pasted on the target polygon (S09).

The processing of steps S01 to S09 is performed on all polygons in the three-dimensional structural model (S
10), the textured three-dimensional structure model created by the above processing is stored (S09).

FIG. 6 shows the position of the video camera according to the first embodiment.
It is a flowchart which shows the processing procedure of attitude | position estimation. As shown in FIG. 7, the processing procedure for estimating the position and orientation of the video camera according to the first embodiment includes, as shown in FIG. 7, image numbers sequentially assigned to consecutive video images and position information ( X, Y, Z) and orientation information (κ, φ, ω) are initialized (S0201), and the high-precision camera position / posture data estimated at certain intervals is initialized. After reading (S0202), the number of frames N of the video camera image existing between the read camera position / posture data is calculated (S0203).

Next, a counter is initialized to count the number of frames calculated (S020).
4) position and orientation information (X, Y, Z, κ,
(φ, ω) is calculated (S0205), and the calculated value is stored in the video image information (S0206). The processing of steps S0205 and S0206 is counted N times by a counter and repeated N times (S0207). Further, the processing is repeated until the processing of steps S0202 to S0207 is completed for all the camera position / posture data estimated at regular intervals.

FIGS. 7 and 8 are flowcharts showing the procedure of the search image narrowing down process of the first embodiment. As shown in FIG. 7, the search image narrowing process according to the first embodiment is performed by sequentially assigning image numbers to successive video images, and vertices for distinguishing each vertex of a texture appearing in the video image of the image number. The texture information stored in the temporary storage area for storing the number, the coordinates of the vertex corresponding to each vertex number on the video image plane, and the distance between each vertex and the shooting point is initialized (S0501), and the video is initialized. One piece of data is read from the video image information in which the position / orientation data of the camera is stored (S0502), and step S of FIG.
One vertex of the target polygon selected in 04 is selected (S0503).

Next, the distance between the selected vertex and the photographing point is calculated (S0504), the vertex coordinates of the texture surface corresponding to the selected vertex are calculated (S0505), and the image number, vertex number, The coordinates and the distance to the photographing point are stored (S0506). Step S0503-
The process of S0506 is performed on all vertices (S050
7) As a result, when the coordinates of all the corresponding points are within the video image plane, the process proceeds to FIG. 8B, and when even one is not included in the video image plane, the process proceeds to C of FIG. Go on (S05
08). This process distinguishes between an image that may have all of the textures in the video image and an image that has all or only some of the textures.

Further, one polygon included within a certain distance from the center of gravity of the target polygon is selected from the image in which all of the texture may be reflected in the video image (S0509), and the selected polygon is selected. One vertex in the polygon is selected (S0510), and the coordinates of a point on the video image plane corresponding to the selected vertex are calculated (S051).
1) Investigate whether the corresponding point is included in the target texture plane (S0512). If the corresponding point is in the texture plane, there is a possibility that the corresponding texture is hidden by the target texture. Therefore, the distance L1 between the vertex selected in step S0510 and the photographing point is calculated (S051).
3) The distance L2 between the center of gravity of the target polygon and the photographing point is calculated (S0514), and the two are compared (S0515).
If L1 is short, the information on the current target polygon is deleted from the texture information (S05).
17). The processing of steps S0510 to S0515 is performed on all vertices (S0516), and the processing of steps S0509 to S0516 is performed on all polygons within a predetermined distance from the center of gravity of the target polygon (S016).
518), the processing of steps S0502 to S0518 is performed on all data of the video image information (S
0519).

FIG. 9 is a flowchart showing the procedure of the texture selection processing according to the first embodiment. As shown in FIG. 10, the texture selection processing procedure initializes variables storing the maximum value of the area of the texture surface and the minimum value of the distance between the center of gravity and the principal point in the texture surface (S0601). One data is read (S0602), the area of the texture surface is calculated (S0603), and the distance between the center of gravity of the texture surface and the principal point is calculated (S0604).

The calculated area of the texture surface and the distance between the center of gravity of the texture surface and the principal point are compared (S060).
5 to S0607) When the area is the maximum, the area, the distance, and the image number at that time are stored when the area is equal to the maximum and the distance is the minimum (S060).
8). The processing of steps S0602 to S0608 is performed on all data (S0609).

FIG. 10 shows storage items of video camera information which is a temporary storage area. In the video camera information used in the present invention, video camera position information bx, by, bz and video camera attitude information κ, ψ, ω are stored for image numbers sequentially allocated to continuous video images.

FIG. 11 shows storage items of texture information which is a temporary storage area. The texture information used in the present invention includes an image number corresponding to the image number of the video camera information shown in FIG. 10, a vertex number indicating each vertex of a texture surface in the video image, coordinates of the vertex, and each vertex. Stores the distance from the shooting point.

As described above, according to the first embodiment, since the three-dimensional structure model having the position information and the position information and the posture information of the video camera at the time of the actual photographing are used,
An optimal texture can be automatically mapped to each polygon of the three-dimensional structure model.

Further, since the video image is used for acquiring the texture, the amount of work and the work time can be greatly reduced and shortened as compared with the conventional camera shooting.

Further, since a large number of images can be obtained, it is possible to easily obtain a texture which has not been obtained from a small number of images.

Further, an image on which an optimal texture is reflected can be automatically selected. Also, it can be automatically pasted on the corresponding polygon. As a result, the amount of work can be significantly reduced as compared with the conventional texture mapping work.

(Embodiment 2) FIG. 12 is a diagram showing a schematic configuration of a video camera apparatus for simultaneously acquiring camera position / posture data and video images according to Embodiment 2 of the present invention.

As shown in FIG. 12, the high-resolution camera 61
GPS from GPS (global positioning system)
Acquisition of camera position / orientation information necessary for the present invention using signal 63 (for details, refer to Japanese Patent Application No. 10-255338).
Then, in order to estimate the video camera position / posture information using the acquired camera position / posture, a video camera 62 is provided in addition to the high-resolution camera 61 as shown in FIG. This makes it possible to acquire high-precision camera position / posture information using the high-resolution camera 61, and to acquire a video image to which the information is added from the video camera 62.

(Embodiment 3) FIG. 13 is a block diagram showing a schematic configuration of a texture mapping apparatus according to Embodiment 3 of the present invention. Reference numeral 63 denotes a GPS signal from GPS. 101A is a camera position / posture data storage area of the high resolution camera 61.

In the third embodiment, the texture mapping device shown in FIG. 13 is mounted on, for example, a helicopter, an automobile, or the like, and a high-precision camera 61 is used by using a GPS signal 63 and a high-resolution image from the high-resolution camera 61. Of the camera position and orientation data (for details, refer to Japanese Patent Application No. 10-255338).
No.). The three-dimensional city structure model having the estimated data and the position information, the aerial video image, and the like are stored in an external storage device, and the texture mapping device of the third embodiment is used for the textured three-dimensional city construction model using the data. Can be created.

In particular, in the present invention, it is not necessary to use aerial video data, and a similar textured three-dimensional structural model can be obtained by mounting the apparatus shown in FIG. Can be created. Further, as the performance of the video camera 62 and the GPS signal 63 becomes higher, smaller, and lighter, texture can be pasted on a three-dimensional structure model having positional information by a portable device as shown in FIG. Will be possible.

As described above, the present invention has been specifically described based on the above-described embodiments. However, the present invention is not limited to the above-described embodiments, and may be variously modified without departing from the gist thereof. Of course.

[0049]

The effects obtained by typical aspects of the invention disclosed in the present application will be briefly described as follows. (1) Since the three-dimensional structure model having the position information and the position information / posture information of the video camera at the time of actual shooting are used, the optimal texture can be automatically mapped to each polygon of the three-dimensional structure model. . (2) Since video images are used for texture acquisition,
The work amount and work time can be greatly reduced and shortened as compared with the conventional camera shooting. (3) Since a large number of images can be obtained, a texture that has not been obtained from a small number of images can be easily obtained. (4) Since the image showing the optimal texture can be automatically selected and automatically pasted to the corresponding polygon, the amount of work is greatly reduced as compared with the conventional texture mapping work. Can be reduced. .

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a texture mapping apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram for explaining the principle of a method for estimating the position and orientation of the video camera according to the first embodiment.

FIG. 3 is a diagram for explaining the principle of a video image reading method according to the first embodiment.

FIG. 4 is a diagram for explaining the principle of a video image selection method according to the first embodiment.

FIG. 5 is a flowchart for explaining the overall processing operation of the texture mapping apparatus according to the first embodiment;

FIG. 6 is a flowchart illustrating a processing procedure for estimating the position and orientation of the video camera according to the first embodiment.

FIG. 7 is a flowchart illustrating a procedure of a search image narrowing process according to the first embodiment.

FIG. 8 is a flowchart illustrating a procedure of a search image narrowing process according to the first embodiment.

FIG. 9 is a flowchart illustrating a texture selection processing procedure according to the first embodiment;

FIG. 10 is a diagram illustrating storage items of video camera information that is a temporary storage area according to the first embodiment.

FIG. 11 shows storage items of texture information which is a temporary storage area according to the first embodiment.

FIG. 12 is a diagram showing a schematic configuration of a video camera device for simultaneously acquiring camera position / posture data and a video image according to a second embodiment of the present invention.

FIG. 13 is a block diagram showing a schematic configuration of a texture mapping apparatus according to a third embodiment of the present invention.

[Explanation of symbols]

1 external storage device 2 central processing unit (CPU)
3. Internal storage device, 101, 101A: Camera position / posture data storage area, 102: 3D structure model storage area, 103: Video image storage area, 104: Textured 3D structure model storage area 104, 201: Data readout Means 202 video camera position / posture estimating means 203 search image narrowing means 204 texture selecting means 205 texture cutting means 206
Texture mapping means, 207: Data writing means, 301: Video camera position and orientation data storage area, 3
02: texture information storage area, 51: video camera projection center (shooting point), 52: video image plane, 53: texture plane, 54: center of gravity of texture plane, 55: principal point, 5
6: target polygon, 57: three-dimensional structure model, 61: high-resolution camera, 62: video camera, 63: GPS signal.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Shigeru Nagai 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo Nippon Telegraph and Telephone Corporation (72) Inventor Isao Miyagawa 3- 19-2, Nishi-Shinjuku, Shinjuku-ku, Tokyo No. Nippon Telegraph and Telephone Corporation (72) Inventor Toshiaki Sugimura 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo F-Term within Nippon Telegraph and Telephone Corporation 5B050 BA06 EA19 5B080 AA13 DA06 GA22 5C054 AA01 FD03 FF03 GB01 GB11 GB15 GB16 HA05

Claims (11)

[Claims] 1. A method for automatically texture-mapping a photographed image taken by a camera with respect to a three-dimensional structure model having position information, wherein the position and orientation data of the video camera at the time of photographing is read. The position and orientation of the video camera at the time of actual shooting are estimated from the obtained data, the target polygon is selected from the three-dimensional structure model, and the range of the video image to be searched from the relative positional relationship between the selected target polygon and the video camera. Narrow down,
From the narrowed down video image, select the video image showing the optimal texture, cut the texture from this selected video image, and automatically paste this cut texture to the corresponding polygon Texture mapping method. 2. The texture mapping method according to claim 1, wherein the processing procedure for estimating the position and orientation of the video camera at the time of actual shooting includes reading high-precision camera position and orientation data estimated at certain intervals, The number of frames of the video camera image existing between the read camera position / posture data is calculated, and the position / posture information (X, Y, Z, κ) of the video camera that has captured the video image corresponding to the relevant frame is calculated. , Φ, ω), store the calculated value in the video image information, and repeat until the processing procedure is completed for all of the camera position / posture data estimated at the fixed interval. Texture mapping method to be featured. 3. The texture mapping method according to claim 1, wherein the processing procedure for narrowing down a range of a video image to be searched based on a relative positional relationship between the three-dimensional structure model and the video camera is performed using position / posture data of the video camera. Is read from the video image information in which is stored, one vertex of the selected target polygon is selected, the distance between the selected vertex and the photographing point is calculated, and the vertex coordinates of the texture surface corresponding to the vertex are calculated. , And store the image number, vertex number, coordinates, and distance from the photographing point as texture information, and perform the above-described processing procedure on all vertices. Distinguish an image from an image that shows all or only part of the texture. Remove from broadcast, with respect to possible image in which all of the texture is reflected in the video image,
One polygon included within a certain distance from the center of gravity of the target polygon is selected, one vertex in the selected polygon is selected, and coordinates of a point on the video image plane corresponding to the selected vertex are calculated. Then, it is checked whether the corresponding point is included in the target texture plane. If the corresponding point is in the target texture plane, the distance L1 between the selected vertex and the photographing point is calculated, and the target polygon is calculated. The distance L2 between the center of gravity and the photographing point is calculated, the two are compared, and the distance L1 is calculated.
Is shorter than the texture information, the information relating to the current video image is deleted from the texture information, the above processing is performed on all vertices, and the processing is performed on all polygons within a certain distance from the center of gravity of the target polygon. A texture mapping method, wherein 4. The texture mapping method according to claim 1, wherein the video image selection processing procedure reads one data from texture information, calculates an area of a texture surface, The distance between the center of gravity of the textured surface and the principal point is calculated, and the calculated area of the textured surface and the distance between the center of gravity of the textured surface and the principal point are compared. A texture mapping method, wherein, when there are a plurality of maximum video images, the video image having the minimum distance is uniquely selected. 5. A method for automatically texture-mapping a photographed image taken by a camera with respect to a three-dimensional structure model having position information, wherein the position and orientation data of the video camera at the time of photographing the photographed image are read and read. The position and orientation of the video camera at the time of actual shooting are estimated from the obtained data, the target polygon is selected from the three-dimensional structure model, and the range of the video image to be searched from the relative positional relationship between the selected target polygon and the video camera. Narrow down,
The computer selects the video image with the best texture from the narrowed video images, cuts the texture from the selected video image, and automatically pastes the cut texture to the corresponding polygon. Recording medium for recording a program to be executed by a computer. 6. A texture mapping method according to claim 1, wherein the processing procedure for estimating the position and orientation of the video camera at the time of actual shooting includes reading high-precision camera position and orientation data estimated at certain intervals, The number of frames of the video camera image existing between the read camera position / posture data is calculated, and the position / posture information (X, Y, Z, κ) of the video camera that has captured the video image corresponding to the relevant frame is calculated. , Φ, ω), store the calculated value in the video image information, and repeat until the processing procedure is completed for all of the camera position / posture data estimated at the constant interval. Recording a program for causing a computer to execute the program. 7. A processing procedure for narrowing a range of a video image to be searched from a relative positional relationship between a three-dimensional structure model and a video camera according to the texture mapping method according to claim 1 or 2, comprising: Is read from the video image information in which is stored, one vertex of the selected target polygon is selected, the distance between the selected vertex and the photographing point is calculated, and the vertex coordinates of the texture surface corresponding to the vertex are calculated. , And store the image number, vertex number, coordinates, and distance from the photographing point as texture information, and perform the above-described processing procedure on all vertices. Distinguish an image from an image that shows all or only part of the texture. Remove from broadcast, with respect to possible image in which all of the texture is reflected in the video image,
One polygon included within a certain distance from the center of gravity of the target polygon is selected, one vertex in the selected polygon is selected, and coordinates of a point on the video image plane corresponding to the selected vertex are calculated. Then, it is checked whether the corresponding point is included in the target texture plane. If the corresponding point is in the target texture plane, the distance L1 between the selected vertex and the photographing point is calculated, and the target polygon is calculated. The distance L2 between the center of gravity and the photographing point is calculated, the two are compared, and the distance L1 is calculated.
Is shorter than the texture information, the information relating to the current video image is deleted from the texture information, the above processing is performed on all vertices, and the processing is performed on all polygons within a certain distance from the center of gravity of the target polygon. A recording medium on which a program for causing a computer to execute a processing procedure for processing is recorded. 8. The texture mapping method according to claim 1, wherein said texture selection processing step reads one piece of data from texture information, calculates an area of a texture surface, The distance between the center of gravity of the surface and the principal point is calculated, and the calculated area of the textured surface and the distance between the center of gravity of the textured surface and the principal point are compared. A recording medium storing a program for causing a computer to execute a processing procedure for uniquely selecting a video image having the smallest distance when there are a plurality of video images. 9. A central processing unit having a temporary storage means, and an external storage device in which a real image taken by a video camera is stored, wherein a real image taken by a video camera for a three-dimensional structural model is provided. A texture mapping device that automatically texture-maps the position and orientation data readout means for reading out position and orientation data from a photographed image taken by a video camera from the external storage device; Position / posture estimating means for estimating posture, search video image range narrowing means for narrowing the range of the search video image from the relative positional relationship between the three-dimensional structure model and the video camera, and video narrowed down by the search video image range narrowing means Select a video image with the best texture from among the images Selecting means, texture cutting means for cutting a texture from the video image selected by the video image selecting means, and pasting means for automatically pasting the texture cut by the texture cutting means to a corresponding polygon. Texture mapping device. 10. The texture mapping apparatus according to claim 9, wherein a real image taken by a video camera provided with a high-resolution camera is stored in an external storage device, and the position and orientation of the video camera are determined based on the position and orientation of the high-resolution camera. Texture mapping device characterized by acquiring posture. 11. A texture mapping device according to claim 9 or 10 mounted on an automobile or a helicopter, a real image taken by a video camera is stored in an external storage device, and a video camera is obtained from the position and orientation of the high-resolution camera. Texture mapping device characterized by acquiring the position and orientation of the object.