JP5769248B2 - Stereo matching processing device, stereo matching processing method, and program - Google Patents

Description

  The present invention relates to a stereo matching processing device, a stereo matching processing method, and a program for realizing these, for obtaining a corresponding point in each image capturing the same object for a stereo image.

  Conventionally, three-dimensional data [DSM (Digital Surface Model) data] indicating terrain is generated by performing stereo matching processing based on images obtained from artificial satellites or aircraft as an automatic method for generating three-dimensional data. How to do is widely known. Further, in this method, when there is a point that cannot be dealt with, correction is performed with an operator interposed.

  Here, the stereo matching process is to obtain a corresponding point in each image capturing the same point from two images taken from different viewpoints, that is, a so-called stereo image, and use that parallax to calculate triangulation. It is a process for obtaining the depth to the object and the shape of the object according to the principle.

  Various methods have already been proposed for this stereo matching process (see, for example, Patent Document 1). Specifically, in the method disclosed in Patent Document 1, first, each point in one photograph is specified for a set of two stereoscopic photographs obtained by photographing from different viewpoints. Next, a point representing the same ground point as each point specified in one photo is specified from the other photo. Then, based on the parallax between each point in one photo and each corresponding point in the other photo, the height information of these points is calculated, and the three-dimensional terrain data is obtained from the calculated height information. Generated.

JP 2004-046777 A

  By the way, in the stereo processing disclosed in Patent Document 1, points (corresponding points) corresponding to each other between two images including overlapping portions are searched for, and based on the corresponding points, corresponding to the principle of triangulation The depth of the point is required. For this reason, searching for corresponding points is important, and it may take a lot of calculation time to search for corresponding points.

  In particular, when an aerial photograph is used as an image, one aerial photograph is an image having about 10000 pixels in the vertical direction and about 10000 pixels in the horizontal direction, and it is very difficult to search for corresponding points between the two aerial photographs. Calculation time may be required.

  On the other hand, the following method can be considered as a method for shortening the calculation time required for searching for corresponding points. First, when the distance between the line connecting the two shooting positions and the corresponding point falls within a certain range, the corresponding amount of parallax also falls within the certain range. Therefore, first, a range (hereinafter referred to as “search range”) is set in advance for the amount of parallax, and the image is divided with a larger division size than the set search range. Next, stereo processing for each divided image is executed in parallel. According to such a method, it is considered that the processing time can be shortened.

  However, since it is difficult to know in advance the depth of the object in the image (the height of the feature in the case of aerial photographs) before performing stereo processing, the parallax of all the objects in the image It is also difficult to determine. Therefore, it is necessary to check whether the image is divided so that the division size is larger than the amount of parallax in all the captured objects, and it is necessary to try matching corresponding points for all the divided images. is there.

  Then, as a result of matching, in the two divided images to be matched, if the amount of parallax for a certain corresponding point is equal to or greater than the search range, it is necessary to perform image division setting again and perform matching again. is there. Conversely, if the amount of parallax is within the search range, a wasteful matching process has been performed. Thus, it is considered difficult to shorten the processing time by simply dividing the image.

  An example of an object of the present invention is a stereo matching process that solves the above-described problem and that can obtain a corresponding point that captures the same point from a stereo image, and that can speed up the process when calculating the parallax. An apparatus, a stereo matching processing method, and a program are provided.

In order to achieve the above object, a stereo matching processing device according to one aspect of the present invention is a device for performing stereo matching processing on two images obtained by photographing the same object from different angles,
A first parallax calculating unit that specifies a plurality of pairs of points corresponding to each other in the two images and calculates a parallax for each of the specified sets;
An image dividing unit that divides the two images so that the divided parts correspond to each other;
For each of the two divided parts corresponding to each other, the corresponding set of points existing in the part is specified, and based on the value of the parallax calculated for the specified set, the parallax maximum in the part A parallax search range setting unit that specifies a value and a minimum value and sets a parallax search range;
For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, the parallax is calculated for each newly specified pair, and the calculated value is A second parallax calculation unit that outputs the calculated parallax as the newly specified parallax in the part when it falls within the parallax search range set in the part;
It is characterized by having.

In order to achieve the above object, a stereo matching processing method according to one aspect of the present invention is a method for performing stereo matching processing on two images obtained by photographing the same object from different angles. And
(A) identifying a plurality of pairs of points corresponding to each other in the two images, and calculating a parallax for each identified group; and
(B) dividing the two images such that the respective divided portions correspond to each other;
(C) For each of the two divided parts corresponding to each other, the set of corresponding points specified in the step (b) existing in the part is specified, and the parallax calculated for the specified set is specified. Identifying the parallax maximum and minimum values in the part based on the value, and setting a parallax search range; and
(D) For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated and calculated for each newly specified pair Outputting the calculated parallax as the newly identified parallax in the part when the value falls within the parallax search range set in the part; and
It is characterized by having.

Furthermore, in order to achieve the above object, a program according to one aspect of the present invention is a program for performing stereo matching processing on two images obtained by photographing the same object from different angles by a computer. And
In the computer,
(A) identifying a plurality of pairs of points corresponding to each other in the two images, and calculating a parallax for each identified group; and
(B) dividing the two images such that the respective divided portions correspond to each other;
(C) For each of the two divided parts corresponding to each other, the set of corresponding points specified in the step (b) existing in the part is specified, and the parallax calculated for the specified set is specified. Identifying the parallax maximum and minimum values in the part based on the value, and setting a parallax search range; and
(D) For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated and calculated for each newly specified pair Outputting the calculated parallax as the newly identified parallax in the part when the value falls within the parallax search range set in the part; and
Is executed.

  As described above, according to the present invention, it is possible to increase the processing speed when calculating the parallax by obtaining the corresponding point capturing the same point from the stereo image.

FIG. 1 is a block diagram showing a configuration of a stereo matching processing device according to an embodiment of the present invention. FIG. 2 is a flowchart showing the operation of the stereo matching processing device according to the embodiment of the present invention. FIG. 3 is a diagram illustrating reduced image creation processing performed in the embodiment of the present invention. FIG. 4 is a diagram for explaining image division processing performed in stereo matching processing according to the embodiment of the present invention. FIG. 5 is a block diagram illustrating an example of a computer that implements the stereo matching processing device according to the embodiment of the present invention.

(Embodiment)
Hereinafter, a stereo matching processing device, a stereo matching processing method, and a program according to an embodiment of the present invention will be described with reference to FIGS.

[Device configuration]
First, the configuration of the stereo processing device according to the embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of a stereo matching processing device according to an embodiment of the present invention.

  The stereo matching processing device 100 according to the present embodiment shown in FIG. 1 performs stereo matching processing on two images (hereinafter referred to as “stereo images”) obtained by photographing the same object from different angles. It is an apparatus which implements.

  As illustrated in FIG. 1, the stereo matching processing device 100 includes a first parallax calculation unit 30, an image division unit 40, a parallax search range setting unit 50, and a second parallax calculation unit 60. Among these, the first parallax calculation unit 30 specifies a plurality of sets of points corresponding to each other in the stereo image, and calculates the parallax for each specified set.

  The image dividing unit 40 divides the stereo image so that the divided parts correspond to each other. First, the parallax search range setting unit 50 identifies a pair of corresponding points existing in each paired part (hereinafter referred to as “paired part”) corresponding to each other. Next, the parallax search range setting unit 50 identifies the parallax maximum value and the minimum value in the pair portion based on the parallax value previously calculated for the identified pair of corresponding points, and sets the pair Set the parallax search range in the part.

  First, the second parallax calculation unit 60 newly specifies a plurality of pairs of points corresponding to each other between the pair parts for each pair part, and calculates the parallax for each newly specified set. Then, the second parallax calculation unit 60 outputs the calculated parallax as the newly specified parallax in the pair portion when the calculated value falls within the parallax search range set in the pair portion. To do.

  As described above, in the present embodiment, since the minimum value and the maximum value of the parallax are obtained in advance in the divided image, the parallax between the divided images may be calculated within the obtained range. For this reason, it is not necessary to perform useless processing when calculating the parallax by obtaining corresponding points capturing the same point from the stereo image, and the processing speed can be increased.

  Here, the configuration of stereo matching processing apparatus 100 in the present embodiment will be described more specifically. In the present embodiment, stereo matching processing device 100 can be realized by a computer as will be described later. The stereo matching processing device 100 can also be realized by a device that can execute an application program, for example, a mobile phone, a smartphone, a tablet-type information device, or a PDA, in addition to the computer described later.

  As shown in FIG. 1, the stereo matching process 100 includes an image reception unit in addition to the first parallax calculation unit 30, the image division unit 40, the parallax search range setting unit 50, and the second parallax calculation unit 60. 10, a reduced image creation unit 20, an image composition unit 70, and an image output unit 80.

  The image receiving unit 10 receives a plurality of stereo images (typically paired image data) used for stereo matching processing in response to a user operation using a keyboard or a mouse (not shown in FIG. 1). ).

  A “stereo image” is an image obtained by photographing almost the same region from different positions using two cameras. Here, among the stereo images, an image taken with the camera arranged on the left side is hereinafter referred to as “left image”, and an image taken with the camera arranged on the right side is hereinafter referred to as “right image”. write. Moreover, in this Embodiment, as a stereo image, a pair of aerial photograph converted into the digital image and a pair of satellite photograph are mentioned, for example. Further, the image data of the stereo image includes not only the image data itself but also the position where the image was captured, the direction in which the image was captured, the angle of view, and the like.

  The image receiving unit 10 can also perform an orientation process for locating camera parameters at the time of shooting, a parallelization process for reprojecting the left image and the right image onto a common parallel plane as necessary. Here, the “orientation process” refers to a process for obtaining a predetermined value for an evaluation target. Furthermore, the “parallelization processing” is to re-project the left image and the right image on a common parallel plane so that the epipolar line connecting the epipole on the left image and the right image matches one of the plurality of scanning lines. Refers to processing.

  Then, the image receiving unit 10 inputs the received image data of the stereo image to the reduced image creating unit 20 and the image dividing unit 40. Note that the rating process and the parallelization process may be executed by different functional blocks.

  The reduced image creation unit 20 creates a reduced image of each stereo image represented by the image data received from the image receiving unit 10. As illustrated in FIG. 3 to be described later, the reduced image creation unit 20 reduces the left image and the right image of the stereo image at an arbitrary magnification, for example, thereby reducing the reduced image of the left image and the reduced image of the right image. Is generated.

  In the present embodiment, the stereo image is reduced in this way, and the processing described later is performed on the reduced image. For this reason, according to the present embodiment, the stereo matching process can be further accelerated as compared with the case where the reduced image is not created.

  In the present embodiment, the first parallax calculation unit 30 performs the above-described parallax calculation based on a reduced image of a stereo image (hereinafter referred to as “stereo reduced image”). Hereinafter, the first parallax calculation unit 30 is referred to as a “reduced image parallax calculation unit 30”.

  Specifically, first, the reduced image parallax calculation unit 30 detects a position in an image that captures the same point among the reduced images generated by the reduced image creation unit 20. That is, a set of corresponding points (feature points) corresponding to the same point between the reduced images is detected. A pair of corresponding points corresponding to the same point can be detected from a position where the correlation coefficient is maximized by taking the image correlation of the corresponding small area in the two images.

  Further, the reduced image parallax calculation unit 30 performs stereo matching processing, specifically DP (Dynamic Programming) matching processing, on the left image and the right image (parallelized image pair) in the target stereo reduced image. Execute, measure the positional deviation (parallax) between the left image and the right image, and generate parallax data.

  When the stereo matching process is executed by the reduced image parallax calculation unit 30, as a result, the parallax between the left and right images is obtained corresponding to each pixel represented by the image data of the stereo reduced image. Here, it is assumed that the image clipping process for calculating the parallax is performed with reference to the ground, and the parallax is zero with respect to the ground. Therefore, the parallax obtained here can be considered as parallax corresponding to a protrusion from the ground such as a building.

  Further, as specific examples of the stereo matching process, there are various processes such as a process for obtaining and associating a general feature amount and a process for obtaining a correlation between left and right images. In this embodiment, a specific stereo matching process is provided. The processing is not particularly limited. As a specific example of the stereo matching process, for example, a process disclosed in Japanese Patent Publication No. 8-16930 can be cited.

  In the present embodiment, the image dividing unit 40 divides the stereo image passed from the image receiving unit 10 into a plurality of block images (see FIG. 4 described later). The size of the block image may be set in advance or may be calculated by the image dividing unit 40 based on the parallax value calculated by the reduced image parallax calculating unit 30.

  In the present embodiment, the parallax search range setting unit 50 is an area corresponding to the pair portion of the parallax values obtained by the reduced image parallax calculation unit 30 for all the pair portions created by the image dividing unit 40. The maximum parallax value and the minimum parallax value are extracted. The maximum value and the minimum value are respectively set as an upper limit and a lower limit of the parallax search range for each pair part.

  The second parallax calculation unit 60 (hereinafter, referred to as “parallax calculation unit 60”) detects the position in the image that captures the same point of each part for the pair part generated by the image dividing unit 40. To do. In other words, in the present embodiment, the parallax calculation unit 60 newly detects a set of corresponding points (feature points) corresponding to the same point of the paired images. A set of corresponding points corresponding to the same point is usually detected from a position where the correlation coefficient of the corresponding small region in the two images is taken and the correlation coefficient is maximized.

  Further, the parallax calculation unit 60 executes the above-described stereo matching process similar to the reduced image parallax calculation unit 30. Then, the parallax calculation unit 60 determines whether or not the parallax calculated by the stereo matching process is within the parallax search range. If the parallax is within the parallax search range, the parallax is calculated using the parallax of the corresponding point of the pair part. Is output to the image composition unit 70.

  The image synthesis unit 70 synthesizes the left image and the right image constituting the stereo image based on the parallax of the corresponding points calculated for each pair part by the parallax calculation unit 60, and represents three-dimensional data ( DSM data) is generated. In addition, the image composition unit 70 outputs the generated three-dimensional data to the image output unit 80.

  When the three-dimensional data is output from the image synthesis unit 70, the image output unit 80 generates a DSM image representing the three-dimensional data and outputs this. Further, the image output unit 80 can output the result of the stereo matching process and the like together with the DSM image.

[Device operation]
Next, the operation of the stereo matching processing device 100 according to the embodiment of the present invention will be described with reference to FIG. FIG. 2 is a flowchart showing the operation of the stereo matching processing device according to the embodiment of the present invention. In this embodiment, the stereo matching processing method is implemented by operating the stereo matching processing device 100. Therefore, the description of the stereo matching processing method in the present embodiment is replaced with the following description of the operation of the stereo matching processing device 100.

  As shown in FIG. 2, first, the image receiving unit 10 receives an input of a stereo image to be processed (step S101). Then, the image receiving unit 10 outputs the received stereo image to the reduced image creating unit 20 and the image dividing unit 40.

  Note that, as described above, a stereo image is an image obtained by photographing almost the same region from different positions using two cameras. Further, as will be described later, a set of corresponding points corresponding to each position in the stereo image is detected, and stereo matching processing is performed.

  Next, the reduced image creating unit 20 creates a stereo reduced image by reducing the stereo image passed from the image receiving unit 10 (step S102). FIG. 3 is a diagram illustrating reduced image creation processing performed in the embodiment of the present invention. As shown in FIG. 3, the reduced image creation unit 20 creates a reduced image by reducing the left image and the right image constituting the stereo image.

  In step S102, the reduction magnification when reducing the image, the size of the reduced image, and the like are set as appropriate. The reduced image creation unit 20 creates a stereo reduced image, and then outputs the created stereo reduced image to the reduced image parallax calculation unit 30.

  Next, the reduced image parallax calculation unit 30 detects a set of corresponding points in the image showing the same point from the stereo reduced image output from the reduced image creation unit 20 (see FIG. 3), and sets the corresponding points. From this, the parallax is calculated (step S103). Then, the reduced image parallax calculation unit 30 outputs the calculated parallax to the parallax search range setting unit 50. In the left image and the right image in FIG. 3, the circled portions indicate detected corresponding points, and the portions connected by lines correspond to each other.

  Next, as shown in FIG. 3, the image dividing unit 40 converts the right image and the left image constituting the stereo image received by the image receiving unit 10 into a plurality of block images each having a preset size. (Step S104). Then, the image dividing unit 40 outputs the generated divided image to the parallax calculating unit 60. FIG. 4 is a diagram for explaining image division processing performed in stereo matching processing according to the embodiment of the present invention.

  Next, the parallax search range setting unit 50 calculates the maximum parallax calculated in step S <b> 103 in the region surrounded by each division position for each divided image (pair portion) created by the image dividing unit 40. A value and a minimum value are specified, and the range from the specified minimum value to the maximum value is set as a parallax search range in the divided image (pair portion) (step S105). Further, the parallax search range setting unit 50 notifies the parallax calculation unit 60 of the set parallax search range.

  Specifically, the disparity search range setting unit 50 identifies, for each pair part, the pair of corresponding points detected in step S103 in the pair part region (see FIG. 3), and further calculates the identified pair. Identify the parallax. Then, the parallax search range setting unit 50 extracts the maximum value and the minimum value from the specified parallax, and sets the extracted minimum value to the maximum value as the parallax search range.

  Next, the parallax search range setting unit 50 determines whether or not step S105 is executed for all divided images (pair portions) (step S106). If step S105 is not executed for all the divided images as a result of the determination in step S106, the parallax search range setting unit 50 executes step S105 again.

  On the other hand, if the result of determination in step S106 is that step S105 has been executed for all of the divided images, the parallax calculation unit 60 includes, for each pair part created by the image division unit 40, an image in the image that captures the same point. A pair of corresponding points is newly detected, and the parallax is calculated for the pair of corresponding points. Further, the parallax calculation unit 60 determines whether the calculated parallax is within the parallax search range set in step S105, and only when the calculated parallax is within the parallax search range, The parallax is set as a point (step S107). Further, the parallax calculation unit 60 outputs the obtained parallax of the corresponding point of the pair part to the image synthesis unit 70.

  Next, the image synthesis unit 70 synthesizes the left image and the right image of the stereo image received in step S101 based on the parallax of the corresponding points of each pair portion output from the parallax calculation unit 60, Three-dimensional data (DSM data) is generated (step S108). Then, the image composition unit 70 outputs the generated three-dimensional data to the image output unit 80.

  Next, the image output unit 80 generates a three-dimensional image (DSM image) based on the three-dimensional data passed from the image composition unit 70, and outputs this (step S109). Thus, the processing in the stereo matching processing device 100 ends.

[Effects of the embodiment]
As described above, according to the present embodiment, it is possible to easily obtain corresponding points in each image capturing the same point for a stereo image. Further, since the parallax between the divided images may be calculated based on the parallax search range, it is not necessary to perform useless processing when calculating the parallax by obtaining the corresponding points from the stereo image. High speed is achieved. In addition, since the process for obtaining the parallax search range is performed based on the reduced stereo image, the speed can be further increased.

[program]
Moreover, the program in this Embodiment should just be a program which makes a computer perform step S101-S109 shown in FIG. By installing and executing this program in a computer, the stereo matching processing device 100 and the stereo matching processing method in the present embodiment can be realized. In this case, the CPU (Central Processing Unit) of the computer includes an image receiving unit 10, a reduced image creating unit 20, a reduced image parallax calculating unit 30, an image dividing unit 40, a parallax search range setting unit 50, a parallax calculating unit 60, an image composition. Functions as the unit 70 and the image output unit 80 and performs processing. That is, the stereo matching processing device 100 according to the present embodiment can be realized using a normal computer without using a dedicated system.

  Further, the program in the present embodiment may be distributed in a state of being stored in a computer-readable recording medium (flexible disk, CD-ROM, DVD-ROM, etc.). Furthermore, when the program is stored in a storage device included in a server device on a communication network such as the Internet, a normal computer downloads the program and installs it, so that the stereo matching processing device 100 is realized.

  Further, it is conceivable that the function of the stereo matching processing device 100 is realized by sharing between the OS (operating system) of the computer and the application program, or realized by cooperation between the OS and the application program. In such a case, only the application program portion needs to be stored in the recording medium or the storage device included in the server device.

  Further, the program in the present embodiment may be distributed via a communication network by being superimposed on a carrier wave. For example, when the program is posted on a Web server on a communication network, the program is distributed via the network. When the program is activated, it is executed and processed in the same manner as other application programs under the control of the OS.

  Here, an example of a computer that realizes the stereo matching processing device 100 by executing the program according to the present embodiment will be described with reference to FIG. FIG. 5 is a block diagram illustrating an example of a computer that implements the stereo matching processing device according to the embodiment of the present invention.

  As shown in FIG. 5, the computer 110 includes a CPU 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These units are connected to each other via a bus 121 so that data communication is possible.

  The CPU 111 performs various calculations by developing the program (code) in the present embodiment stored in the storage device 113 in the main memory 112 and executing them in a predetermined order. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Further, the program in the present embodiment is provided in a state of being stored in a computer-readable recording medium 120. Note that the program in the present embodiment may be distributed on the Internet connected via the communication interface 117.

  Specific examples of the storage device 113 include a hard disk and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and a mouse. The display controller 115 is connected to the display device 119 and controls display on the display device 119. The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, and reads a program from the recording medium 120 and writes a processing result in the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and another computer.

  Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash) and SD (Secure Digital), magnetic storage media such as a flexible disk, or CD-ROM (Compact Disk). Optical storage media such as Read Only Memory).

  Part or all of the above-described embodiments can be expressed by (Appendix 1) to (Appendix 12) described below, but is not limited to the following description.

(Appendix 1)
A device for performing a stereo matching process on two images obtained by photographing the same object from different angles,
A first parallax calculating unit that specifies a plurality of pairs of points corresponding to each other in the two images and calculates a parallax for each of the specified sets;
An image dividing unit that divides the two images so that the divided parts correspond to each other;
For each of the two divided parts corresponding to each other, the corresponding set of points existing in the part is specified, and based on the value of the parallax calculated for the specified set, the parallax maximum in the part A parallax search range setting unit that specifies a value and a minimum value and sets a parallax search range;
For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, the parallax is calculated for each newly specified pair, and the calculated value is A second parallax calculation unit that outputs the calculated parallax as the newly specified parallax in the part when it falls within the parallax search range set in the part;
A stereo matching processing device characterized by comprising:

(Appendix 2)
A reduced image creation unit that reduces the two images to generate two reduced images;
The first parallax calculation unit specifies a plurality of pairs of points corresponding to each other for the two reduced images, and calculates a parallax for each specified set;
The stereo matching processing device according to attachment 1.

(Appendix 3)
An image compositing unit that creates three-dimensional data to be imaged of the two images based on the parallax output from the second parallax calculating unit;
The stereo matching processing device according to appendix 1 or 2.

(Appendix 4)
The stereo matching processing device according to any one of appendices 1 to 3, wherein the two images are aerial photographs or satellite photographs.

(Appendix 5)
A method for performing a stereo matching process on two images obtained by photographing the same object from different angles,
(A) identifying a plurality of pairs of points corresponding to each other in the two images, and calculating a parallax for each identified group; and
(B) dividing the two images such that the respective divided portions correspond to each other;
(C) For each of two divided parts corresponding to each other, the set of corresponding points specified in the step (b) existing in the part is specified, and the parallax calculated for the specified set A parallax search range is set by identifying the maximum and minimum values of the parallax in the part based on the value of
(D) For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated and calculated for each newly specified pair Outputting the calculated parallax as the newly identified parallax in the part when the value falls within the parallax search range set in the part; and
A stereo matching processing method characterized by comprising:

(Appendix 6)
(F) further reducing the two images to generate two reduced images;
In the step (a), a plurality of pairs of points corresponding to each other are specified for the two reduced images, and the parallax is calculated for each specified group.
The stereo matching processing method according to attachment 5.

(Appendix 7)
(G) based on the parallax output in the step (d), further creating a three-dimensional data of the two images to be imaged;
The stereo matching processing method according to appendix 5 or 6.

(Appendix 8)
The stereo matching processing method according to any one of appendices 5 to 7, wherein the two images are aerial photographs or satellite photographs.

(Appendix 9)
A program for performing stereo matching processing on two images obtained by photographing the same object from different angles by a computer,
In the computer,
(A) identifying a plurality of pairs of points corresponding to each other in the two images, and calculating a parallax for each identified group; and
(B) dividing the two images such that the respective divided portions correspond to each other;
(C) For each of two divided parts corresponding to each other, the set of corresponding points specified in the step (b) existing in the part is specified, and the parallax calculated for the specified set A parallax search range is set by identifying the maximum and minimum values of the parallax in the part based on the value of
(D) For each of the two divided parts corresponding to each other, a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated and calculated for each newly specified pair Outputting the calculated parallax as the newly identified parallax in the part when the value falls within the parallax search range set in the part; and
A program that executes

(Appendix 10)
(F) causing the computer to further execute a step of reducing the two images to generate two reduced images;
In the step (a), a plurality of pairs of points corresponding to each other are specified for the two reduced images, and the parallax is calculated for each specified group.
The program according to appendix 9.

(Appendix 11)
(G) Based on the parallax output in the step of (d), creating the three-dimensional data of the imaging target of the two images, further causing the computer to execute the step.
The program according to appendix 9 or 10.

(Appendix 12)
The program according to any one of appendices 9 to 11, wherein the two images are aerial photographs or satellite photographs.

  As mentioned above, the above-mentioned embodiment is for description, does not limit the scope of the present invention, and various modifications and applications are possible in the present invention. Therefore, those skilled in the art can employ embodiments in which each or all of these elements are replaced with equivalent ones, and these embodiments are also included in the scope of the present invention. In addition, the above-described hardware configuration and flowchart are merely examples, and can be arbitrarily changed and modified.

  As described above, according to the present invention, it is possible to increase the processing speed when calculating the parallax by obtaining the corresponding point capturing the same point from the stereo image. The present invention is useful in the field where stereo matching processing is required, particularly in the field of creating a three-dimensional image from an aerial photograph or a satellite photograph.

DESCRIPTION OF SYMBOLS 10 Stereo matching processing apparatus 10 Image reception part 20 Reduced image creation part 30 Reduced image parallax calculation part (1st parallax calculation part)
40 image division unit 50 parallax search range setting unit 60 parallax calculation unit (second parallax calculation unit)
70 Image Composition Unit 80 Image Output Unit 100 Stereo Matching Processing Device 110 Computer 111 CPU
112 Main Memory 113 Storage Device 114 Input Interface 115 Display Controller 116 Data Reader / Writer 117 Communication Interface 118 Input Device 119 Display Device 120 Recording Medium 121 Bus

Claims (5)

An apparatus for performing a stereo matching process on two original images obtained by photographing the same object from different angles,
A reduced image creating unit that reduces the two original images to generate two reduced images;
A first parallax calculating unit that specifies a plurality of pairs of points corresponding to each other in the two reduced images, and calculates a parallax for each of the specified sets;
An image dividing unit that divides the two original images so that the respective divided portions correspond to each other;
For each of two divided parts corresponding to each other, the corresponding point set in the reduced image that exists in the part is specified, and the part based on the parallax value calculated for the specified set A parallax search range setting unit that specifies a parallax search range by specifying the maximum value and the minimum value of the parallax in
For each of the two divided portions corresponding to each other in the original image , a plurality of pairs of points corresponding to each other between the two portions are newly specified, and the parallax is calculated for each newly specified set and calculated. A second parallax calculation unit that outputs the calculated parallax as the newly specified parallax in the part when the calculated value falls within the parallax search range set in the part;
A stereo matching processing device characterized by comprising: An image compositing unit that creates three-dimensional data to be imaged of the two images based on the parallax output from the second parallax calculating unit;
The stereo matching processing device according to claim 1 . The two images are aerial or satellite photographs, stereo matching apparatus according to claim 1 or 2. A method for performing stereo matching processing on two original images obtained by photographing the same object from different angles,
(A) reducing the two original images to generate two reduced images; and
( B ) identifying a plurality of pairs of points corresponding to each other in the two reduced images, and calculating a parallax for each identified group; and
(C) dividing the two original images so that the respective divided portions correspond to each other; and
(D) For each of two divided parts corresponding to each other, the set of corresponding points specified in the step (b) in the reduced image existing in the part is specified, and the specified set Based on the calculated parallax value, specifying the maximum and minimum parallax values in the part and setting a parallax search range; and
( E ) For each of the two corresponding divided parts in the original image , a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated for each newly specified pair And when the calculated value falls within the parallax search range set in the part, the calculated parallax is output as the newly specified parallax of the part; and
A stereo matching processing method characterized by comprising: A program for performing stereo matching processing on two original images obtained by photographing the same object from different angles by a computer,
In the computer,
(A) reducing the two original images to generate two reduced images; and
( B ) identifying a plurality of pairs of points corresponding to each other in the two reduced images, and calculating a parallax for each identified group; and
(C) dividing the two original images so that the respective divided portions correspond to each other; and
(D) For each of two divided parts corresponding to each other, the set of corresponding points specified in the step (b) in the reduced image existing in the part is specified, and the specified set Based on the calculated parallax value, specifying the maximum and minimum parallax values in the part and setting a parallax search range; and
( E ) For each of the two corresponding divided parts in the original image , a plurality of pairs of points corresponding to each other between the two parts are newly specified, and the parallax is calculated for each newly specified pair And when the calculated value falls within the parallax search range set in the part, the calculated parallax is output as the newly specified parallax of the part; and
A program that executes

Images