KR20200064222A - Method and system for image registering using regional warping - Google Patents

Abstract

Disclosed are an image matching method using regional warping and a system thereof. The image matching method using regional warping comprises the steps of: specifying, by an image matching system, a warping area on which regional warping is to be performed from a matched image in which a plurality of images are matched; separating, by the image matching system, the specified warping area from the matched image to warp separated warping area images in a predetermined manner; and combining, by the image matching system, the warped warping area images to the matched image. Therefore, natural matching can be achieved while reducing conversion errors of the warping area.

Description

Method and system for image registering using regional warping}

The present invention relates to an image matching method and a system thereof, and to a method and system capable of enhancing matching results by warping only a portion of a matching image in which the images to be matched are matched.

Technical ideas for matching two different images are well known.

Image registration is commonly used when generating an image that covers a wide viewing angle, such as a panoramic image or a 360-degree image, and is provided by the mobile device according to many popular trends of mobile devices (eg, smart phones, etc.). These image registrations are becoming more widely used for a variety of image solutions.

An important factor in image registration is to allow visually natural matching when two different images are registered. In general, in order to match such an image, when two different images overlap regions, the two images are commonly matched by using a transformation relationship between feature points that are common to the overlapping regions, that is, they match each other.

On the other hand, when two images are matched using these feature points, natural matching is generally performed, but particularly when a plurality of images are matched (for example, to generate a 360-degree image), all of the images to be matched are matched. As they affect each other, the overall quality of registration degradation may be severe.

In the related art, a method of manually re-calibrating or calibrating by adding a control point or manually calibrating while changing a seam line has been used to solve the matching quality.

However, since the method of adding a control point is a method of generating new transformation information (transformation matrix) based on the control point and performing re-alignment, when re-aligning through the new transformation information, the well-matched part is misaligned. There is this.

In addition, in the case of calibration through the adjustment of the seam line, if there are other objects with a high visual weight in addition to an area or an object for improving the matching performance on the seam line to be adjusted, there is a problem that the matching of the other objects is rather bad. In addition, the calibration through the seam line adjustment has a problem in that it is not appropriate to perform the calibration except mainly the flat area of the association.

Korean Registered Patent Registration No. 10-1697229 "A lane information-based automatic correction device and method for matching images for vehicles"

The technical problem to be achieved by the present invention is a method for performing a natural matching with very low conversion error of a corresponding warping region by performing a warping of only a partial region of the entire registration image separately from the entire region through regional warping, and It is to provide a system.

An image matching method using regional warping according to an embodiment of the present invention for achieving the above technical problem includes: an image matching system specifying a warping region to perform regional warping from a matched image in which a plurality of images are matched; Separating the warping region specified by the image registration system from the registration image, and warping the separated warping region image in a predetermined manner, and combining the warping region image warped by the image matching system to the registration image Steps.

The step of specifying the warping region may include specifying a feature point pair in which a matching result error exists in the registration image and a specified feature point pair, and specifying the warping region having a size of a region proportional to the distance between the feature point pairs. It may include steps.

The step of warping the separated warping area image in a predetermined method includes setting a fixed number of fixed points or more on a periphery of the separated warping area image, and a warping algorithm using a pair of the fixed point and the feature point as inputs. And through the step of warping the separated warping region image.

The step of warping the separated warping region image in a predetermined manner includes separating the warping region image into a first partial warping region image and a second partial warping region image based on the midpoint of the feature point pair, and the first. The method may include obtaining a warping result of the warping area image by combining the warping results after warping the partial warping area image and the second partial warping area image, respectively.

The step of obtaining the warping result of the warping region image by combining the warping results after warping the first partial warping region image and the second partial warping region image respectively, among the feature point pairs included in the first partial warping region image Warping the first partial warping region image through a warping algorithm using the first feature point and the midpoint included in the first partial warping region image and the pair of feature points included in the second partial warping region image And warping the second partial warping region image through the warping algorithm using the second feature point and the midpoint included in the second partial warping region image as input.

The warping of the first partial warping area image and the warping of the second partial warping area image may be performed in opposite directions.

The warping algorithm may be a thin plate spline warping algorithm.

The above method can be implemented by a computer program.

An image matching system using a local warping for solving the technical problem includes a processor and a storage medium in which a program executed by the processor is recorded, and the program performs regional warping from a matched image in which a plurality of images are matched. The warping region to be performed is specified, the specified warping region is separated from the registration image, the separated warping region image is warped in a predetermined manner, and the warped region image is combined with the registration image.

The program may specify a pair of feature points in which a matching result error exists in the matched image, and may specify the warping region having a size of a region that includes the specified feature point pairs and is proportional to the distance between the feature point pairs.

The program may set a fixed number of fixed points or more on a periphery of the separated warping area image, and warp the separated warping area image through a warping algorithm using the set pair of fixed points and the feature points as inputs.

The program divides the warping region image into a first partial warping region image and a second partial warping region image based on the midpoint of the feature point pair, and warps the first partial warping region image and the second partial warping region image, respectively. After that, the warping results of the warping region image may be obtained by combining the warping results.

The program warps the first partial warping region image through a warping algorithm that inputs the first feature point and the midpoint included in the first partial warping region image among the feature point pairs included in the first partial warping region image. And warping the second partial warping region image through the warping algorithm using the second feature point and the midpoint included in the second partial warping region image among the feature point pairs included in the second partial warping region image. can do.

Since the conversion of each image for the conventional image matching is made of one conversion formula, it is difficult to fundamentally improve the matching quality by recalculating or modifying the conversion formula, whereas according to the technical idea of the present invention, By converting only a partial region in a manner independent of the image transformation, that is, by performing regional warping, at least some of the regions have an effect of obtaining a higher matching quality than the conventional method.

In particular, the improvement of the matching quality can be very high for the area by performing regional warping (local warping) independent of the entire image conversion so that there is no error or very low error between the two points with large errors. , There is an effect that can solve the problems of the conventional method.

BRIEF DESCRIPTION OF THE DRAWINGS In order to better understand the drawings cited in the detailed description of the present invention, a brief description of each drawing is provided.
1 shows a schematic configuration of an image matching system for implementing an image matching method using regional warping according to an embodiment of the present invention.
2 schematically shows a logical configuration of an image matching system using regional warping according to an embodiment of the present invention.
3 is a flowchart illustrating an image matching method using regional warping according to an embodiment of the present invention.
4 to 6 are diagrams showing examples of images to which an image matching method using regional warping according to an embodiment of the present invention is applied.

In order to fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings and the contents described in the accompanying drawings, which illustrate preferred embodiments of the present invention.

In addition, in the present specification, when one component'transmits' data to another component, the component may directly transmit the data to the other component, or through at least one other component It means that the data may be transmitted to the other components. Conversely, when one component'directly transmits' data to another component, it means that the data is transmitted from the component to the other component without passing through the other component.

Hereinafter, the present invention will be described in detail by explaining preferred embodiments of the present invention with reference to the accompanying drawings. The same reference numerals in each drawing denote the same members.

1 shows a schematic configuration of an image matching system for implementing an image matching method using regional warping according to an embodiment of the present invention. In addition, FIG. 2 schematically shows a logical configuration of an image matching system using regional warping according to an embodiment of the present invention.

First, referring to FIG. 1, an image matching system using regional warping according to an embodiment of the present invention (hereinafter referred to as an'image matching system') may be implemented as a predetermined data processing apparatus.

The image matching system 100 includes a processor 110 and a storage medium 120 for implementing functions defined herein as shown in FIG. 1. The processor 110 may mean a computing device capable of executing a predetermined program (software code), an example of implementation of the data processing device, or a vendor mobile processor, microprocessor, CPU, single processor, multiprocessor, etc. It can be named by various names.

The storage medium 120 may mean a device in which a program for implementing the technical idea of the present invention is stored/installed. Depending on the implementation example, the storage medium 120 may be divided into a plurality of different physical devices, and depending on the implementation, a part of the storage medium 120 may exist inside the processor 110. The storage medium 120 may be implemented as a hard disk, a solid state disk (SSD), an optical disk, a random access memory (RAM), and/or various other types of storage media, depending on implementation. The image matching system 100 may be implemented detachably.

The image matching system 100 may be, for example, a mobile terminal (eg, a mobile phone, a laptop, a tablet, etc.), but is not limited to any data processing device (eg, computer, server) capable of processing data to execute the program. Device, etc.).

In addition, the image matching system 100 includes the processor 110, the storage medium 120, and various peripheral devices (eg, input/output devices, display devices, audio devices, etc.) provided in the image matching system 100. , 141, 142) and a communication interface (eg, communication bus, 130, etc.) for connecting these devices may be easily inferred by the average expert in the technical field of the present invention.

Meanwhile, the technical idea of the present invention may be implemented by organically combining the program stored in the storage medium 120 and the processor 110, and the functional configuration unit executed by the image matching system 100 is It may be as shown in Figure 2.

That is, the image matching system 100 may include a control module 210, a feature point processing module 220, a reference warping module 230, and a conversion information calculation module 240.

In the present specification, the module means hardware for performing the technical idea of the present invention (eg, the processor 110 and/or storage medium 120) and software for driving the hardware (eg, the technical of the present invention. It may mean a functional and structural combination of the above program) for realizing the idea. For example, each of the above components may mean a logical unit of a predetermined code and a hardware resource for performing the predetermined code, and must mean a physically connected code, or one type or a specific number of hardware. It does not mean that can be easily inferred to the average expert in the technical field of the present invention. Therefore, each of the above configurations refers to a combination of hardware and software performing functions defined in this specification, and does not mean a specific physical configuration.

The control module 210 is a function of other components (eg, the feature point processing module 120, the warping module 230, and/or the conversion information calculation module 240) included in the image matching system 100. And/or resources.

According to an example, the control module 210 communicates with a predetermined sensor or control device (for example, a processor, etc.) installed in a data processing device (for example, a mobile phone) in which the image matching system 100 is implemented. It is also possible to transmit and receive signals or information for implementing the technical idea of the present invention.

For example, the control module 210 may receive information (eg, camera parameters) about an action performed between each of the images to be matched from the data processing device (eg, a mobile phone). For example, the action may be an action in which the field of view (position or direction) is changed. Alternatively, the action may be, for example, an action in which a scale is changed when an image is taken, and various actions may be possible. Information on such an action may be used, for example, to specify a matching area or matching relationship of each image to be matched.

In addition, the control module 210 may combine the warping image after the regional warping or the local warping is performed to the original image, as will be described later.

The control module 210, after combining the warping image to the original position, a predetermined interpolation (interpolation) to alleviate the unnaturalness near the boundary of the warping region, that is, the portion where warping is performed and the portion where warping is not performed. ). When such interpolation is performed, the boundary of the warping region, that is, the portion near the outline of the warped image, is warped by setting a fixed point on the periphery when warping. It has the effect of being able to combine images. Interpolation techniques for naturalness are well known, so detailed descriptions thereof will be omitted herein.

The feature point processing module 220 may extract feature points of objects included in each matching region of the matching target images.

The method for extracting the feature points may be various. For example, the SIFT algorithm may be used, but is not limited to, any method or algorithm may be used as long as it can extract the characteristic points of the object independent of brightness, scale, and/or rotation, that is, the feature points.

In addition, the feature point extraction module 220 may determine a feature point that matches each other among a plurality of feature points extracted from a matching region of each of the matching target images. That is, among the second feature points extracted from the second matching region of the second matching target image, the matching point can be determined, matching each of the first feature points extracted from the first matching region of the first matching target image. .

According to an example, the feature point extraction module 220 may extract a plurality of first feature points from the first registration region. In addition, the feature point extraction module 220 may extract a plurality of second feature points from the second registration region, and determine a feature point that matches each of the first feature points among the extracted second feature points.

The first registration region and/or the second registration region may be the first registration target image and/or the second registration target image itself. Depending on the implementation example, the first matching area and/or the second matching area may be part of the image to be matched. Then, the feature point extraction module 220 may extract a feature point from each matching region of the images to be matched and determine matching feature points.

In addition, the feature point processing module 220 may specify a pair of control points for specifying a warping area.

The pair of control points may be feature points that match each other. For example, a specific feature point in a first image and a feature point that must match the specific feature point in a second image to be matched with the first image may be a pair of control points.

Ideally, these pairs of control points should be transformed to be points at exactly the same position on the registration image, but there may be more than a certain level of error, and thus may not be completely matched.

The pair of control points may be automatically identified as feature points matching each other having an error of a predetermined level or more after image matching is performed. Depending on the embodiment, a pair of control points having the largest error within a predetermined region may be automatically extracted.

The predetermined area may be automatically selected by a preset criterion (eg, an area having high image complexity or an area in which a specific object exists), or may be arbitrarily selected by a user. For example, the user selects an area of interest to increase matching quality, and at least one of matching feature points existing in the selected area of interest may be specified as a pair of control points.

When a pair of control points is specified in various ways, the warping module 230 may specify a warping area based on the specified pair of control points.

The warping area can be sized by a specified pair of control points. The size of the warping area may need to be set larger than the area containing the pair of control points as a minimum size because the intensity of the overall distortion varies after performing the warping depending on the size of the area to be warped. That is, when the size of the warping region is set too small, the combination of the warping region and the surrounding image after warping may be very unnatural.

Accordingly, according to the technical idea of the present invention, the warping area may be set to be proportional to the distance between pairs of control points, that is, the size of the warping distance. For example, a x-axis component of a line segment connecting a pair of control points may be scaled by a predetermined ratio, and a y-axis component may be scaled by a predetermined ratio to specify a warping region. Of course, in order to specify the warping region, a condition that each component of the x-axis and the y-axis may have a minimum size may be further provided. Therefore, even if the x-axis component or y-axis component of the line segment connecting the control point is none or very small, the warping region may be set to have a certain length for each component. In any case, the size of the warping region may be variously set to be proportional to the size of the warping distance.

When the warping region is specified as described above, the warping module 230 separates (eg, crops) the warping region from the original registration image, and independently of the transformation of each of the images included in the registration image. You can do the transformation, i.e., warping.

The warping module 230 may perform warping to move one control point included in a pair of control points to another control point.

At this time, in order to prevent the boundary line of the warping area from being distorted through warping, the warping module 230 sets a predetermined number of fixed points on the boundary line of the warping area, and the fixed point does not move, and any one control point Warping can be performed so that can be moved to another control point.

Then, the warping module 230 may warp the image corresponding to the warping region through a predetermined warping algorithm using the set pair of fixed points and the feature points, that is, a pair of control points.

According to an example, the warping module 230 may perform warping using a thin plate spline algorithm. The thin-line spline warping algorithm can find a smooth surface with minimal deformation that passes all of the given coordinate values on the plane. Accordingly, the thinning spline warping algorithm may perform the thinning spline warping algorithm using the coordinates of the control point pair and the fixed points as input values, thereby allowing the warping module 230 to warp the image of the warping region.

The thin-line spline warping algorithm used by the warping module 230 to perform warping is exemplary, and in addition, various warping methods may be used to allow control point pairs to be matched in various ways while not moving the boundary lines of the warping area. Will be easily deduced by an average expert in the technical field of the present invention.

Meanwhile, as described above, the warping module 230 may simply perform one warping to match a pair of control points. Depending on the embodiment, split warping may be performed.

That is, a pair of control points having a high error value must have a high warping distance, and when image warping is performed, one image is warped in one direction, and thus the image of the image may be distorted a lot. In this case, the error may be lowered to zero or close to zero, but it may appear visually awkward. In order to prevent excessive distortion of the image, the warping distance must be reduced. To this end, the warping module 230 may divide and perform warping.

For example, the warping module 230 divides an image based on a midpoint of a pair of control points (eg, a midpoint of a line connecting a pair of control points) (eg, passing through a midpoint of a line segment connecting the pair of control points, and the line segment Orthogonal straight line).

When warping is performed for each of the divided images, the warping distance at the time of each warping may be halved.

For example, after setting each pair of control points as a source point in a corresponding image, and setting a center point of the control point pair as a destination point, warping may be performed on each of the divided images. In this case, since each segmented image is warped in different directions, image distortion may also be reduced.

In this case, the fixed point included in the corresponding area among the fixed points set in the entire warping area may be input as an input value to the warping algorithm, and regional warping may be performed for each of the divided images.

Also, in this case, since the warping algorithm has the characteristic that the execution time increases exponentially when the warping distance is increased, it can be more effective to reduce the warping distance in half and double the number of executions, and each video is reduced because the warping distance is reduced. The degree of distortion is also reduced.

Of course, after warping is performed for each of the divided images, the warping images for the entire warping area may be obtained by combining partial images on which warping is performed.

Then, the control module 210 may combine the obtained warping image back into the registration image. At this time, the control module 210 may interpolate a predetermined area near the boundary line of the warping area so that a natural combination of the warping image and other parts can be achieved.

Of course, if the regional warping is performed on one pair of control points and the combined warping image is combined with the matched image, additional regional warping may be performed on another pair of control points.

Meanwhile, the conversion information calculation module 240 may calculate conversion information (eg, a conversion matrix) for a plurality of images input to the image matching system 100. The control module 210 may generate a matched image based on the calculated conversion information.

The reference feature points may need to be selected in order for the conversion information calculation module 240 to calculate the conversion information for each image. The reference feature points may be feature point pairs that serve as a reference for calculating transformation information among feature points extracted by the feature point processing module 220. Generally, conversion information can be calculated by selecting three specific point pairs as reference feature points.

The conversion information calculation module 240 may select reference feature points in a predetermined manner. An example for such a method is the RANSAC (Random Sample Consensus) algorithm.

3 is a flowchart illustrating an image matching method using regional warping according to an embodiment of the present invention.

Referring to FIG. 3, when the image matching system 100 receives a plurality of images to be matched, a matched image may be generated (S100 ),

Then, the image matching system 100 may specify a warping area in which regional warping is to be performed (S110).

The warping area may be automatically selected by the image registration system 100 (ie, a control point pair may be automatically selected) by a predetermined criterion as described above, or may be specified by user selection.

To this end, the image matching system 100 may provide the user with a first generated matching image, and may select a user's region of interest based on the provided image. That is, the user may select an area in which a matching is desired particularly well. Then, a pair of control points can be specified in the selected region. When a pair of control points is specified, a warping region may be specified. The user's region of interest and the warping region may of course be different.

When the warping region is specified, the image matching system 100 may perform regional warping on the warping region image as described above (S120).

In addition, the combined warping image may be combined with the registration image to complete a regional warping process corresponding to a pair of control points (S130 ).

Of course, the regional warping process may be repeated multiple times as necessary.

4 to 6 are diagrams illustrating an example of applying an image matching method using regional warping according to an embodiment of the present invention.

First, FIG. 4A exemplarily shows a pair of control points 10 and 11 in a given image and a warping area (eg, an area within a green square) in the image, and FIG. 4B shows another pair of control points 12 and 12 in the image. 13) and a corresponding warping region (eg, an area within a green square) are exemplarily shown.

4A and 4B, it can be seen that the size of the warping region may be set to be proportional to the distance of the pair of control points.

If the warping region is thus specified to correspond to a pair of control points, the warping region can be separated from the registration image. Then, as shown in FIG. 4C, a warping region image may be obtained.

When the warping area image is obtained, as described above, the image registration system 100 may set a predetermined number of fixed points on the boundary line of the warping area (or the appearance line of the warping area image).

5A shows a case where five fixed points (for example, green points) are set for each boundary line of the warping area, and FIG. 5B shows a case where fixed points (for example, green points) are set for only four vertices.

And the results of warping using these fixed points and the red point as the source point and the blue point as the destination point are shown on the right side of FIGS. 5A and 5B, respectively.

It can be seen from FIG. 5B that the number of fixed points is not sufficient, and thus the movement to the boundary of the warping region image is performed. If the number of fixed points is too large, the warping efficiency may be lowered, and if the number of fixed points is small, a boundary line may be moved as shown in FIG. 5B, so it is necessary to set the number of appropriate fixed points. The setting of the number of fixed points may be selected through an experiment, the number of optimal fixed points may be calculated while repeatedly performing warping, and various embodiments may be possible.

Also, by setting the fixed point, the warping intensity is weakened so that the intensity of warping is strong and the distance is far away from the source point and the destination point, so that the warped warping area image is naturally applied to the original image, that is, the registration image. It also has the effect of allowing it to be combined.

FIG. 6A shows a result (right image) of calibration when a method of manually adding a conventional control point based on a distant object with respect to the image in the left square is performed. In this case, it can be seen that there may be a problem in that the matching quality is lowered in the close region (the original circle on the right).

FIG. 6B shows the result (right image) of the calibration performed by the conventional seam adjustment method based on the portion close to the image in the left square. In this case, it can be seen that there may be a problem that the matching quality is lowered in a distant area (the original circle on the right).

And Figure 6c shows the results of performing the calibration using the image matching method using the regional warping according to the technical idea of the present invention, each of the areas in which the matching quality is lowered in Figures 6a and 6b is relatively excellent matching quality You can see that you can have.

After all, according to the technical idea of the present invention, by allowing intensive calibration to be performed on a specific area independently of transformation of each of the images to be matched, other areas, that is, relative to increase the matching quality of one area As a result, it is possible to increase the problem that the matching quality of the region where the matching quality is good is lowered.

The image matching method using regional warping according to an embodiment of the present invention can be implemented as computer readable codes on a computer readable recording medium. The computer-readable recording medium includes all types of recording devices in which data readable by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tapes, hard disks, floppy disks, and optical data storage devices. In addition, the computer-readable recording medium can be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the present invention can be easily inferred by programmers in the technical field to which the present invention pertains.

The present invention has been described with reference to one embodiment shown in the drawings, but this is merely exemplary, and those skilled in the art will understand that various modifications and other equivalent embodiments are possible therefrom. Therefore, the true technical protection scope of the present invention should be determined by the technical spirit of the appended claims.

Claims (14)

Specifying, by the image matching system, a warping area to perform regional warping from a matched image in which a plurality of images are matched;
Separating the warping region specified by the image registration system from the registration image, and warping the separated warping region image in a predetermined manner; And
And combining the warped region image in which the image matching system has been warped into the matched image.
According to claim 1, The step of specifying the warping region,
Specifying a pair of feature points in which a matching result error exists in the matching image; And
And specifying the warping region having a size of an area proportional to the distance between the feature point pairs.
The method of claim 2, wherein the step of warping the separated warping region image in a predetermined manner,
Setting a fixed number of fixed points or more on a periphery of the separated warping region image;
And warping the separated warping area image through a set of warping algorithms using the fixed point and the feature point pair as inputs.
The method of claim 2, wherein the step of warping the separated warping region image in a predetermined manner,
Separating the warping region image into a first partial warping region image and a second partial warping region image based on the midpoint of the feature point pair; And
And warping the first partial warping region image and the second partial warping region image, respectively, and combining warping results to obtain warping results of the warping region image.
The method of claim 4, wherein obtaining the warping result of the warping region image by combining the warping results after warping the first partial warping region image and the second partial warping region image, respectively,
Warping the first partial warping region image through a warping algorithm using the first feature point and the midpoint included in the first partial warping region image among pairs of the feature points included in the first partial warping region image; And
Warping the second partial warping region image through the warping algorithm using the second feature point and the midpoint included in the second partial warping region image among the pair of feature points included in the second partial warping region image. Image registration method using the regional warping including a.
The method of claim 5, wherein the warping of the first partial warping region image and the second partial warping region image are performed in opposite directions.
The method of claim 1, wherein the image matching system combines the warped region images warped with the matched images,
And interpolating a predetermined region based on an outline of the warped region image in which the image matching system is warped.
The method of claim 3, wherein the warping algorithm,
Image matching method using regional warping, characterized in that it is a thin plate spline warping (thin plate spline warping) algorithm.
A computer program installed in a data processing apparatus and stored in a computer-readable recording medium for performing the method according to any one of claims 1 to 8.
Processor; And
It includes a storage medium in which the program executed by the processor is recorded,
The above program,
Specifying a warping area to perform regional warping from a matched matched image of a plurality of images, separating the specified warping area from the matched image, warping the separated warping area image in a predetermined manner, and warping the An image registration system using regional warping that combines a warping region image with the registration image.
The method of claim 10, wherein the program,
An image matching system using regional warping which specifies a pair of feature points in which a matching result error exists in the matched image, and includes the specified feature point pairs and specifies the warping region having a size of a region proportional to the distance between the feature point pairs.
The method of claim 11, wherein the program,
An image using regional warping that sets a fixed number of fixed points or more on an outer periphery of the separated warping area image, and warps the separated warping area image through a warping algorithm using the set pair of fixed points and the feature points as input. Matching system.
The method of claim 11, wherein the program,
The warping region image is divided into a first partial warping region image and a second partial warping region image based on the midpoint of the feature point pair, and the warping is performed after warping the first partial warping region image and the second partial warping region image, respectively. An image matching system using regional warping that combines results to obtain a warping result of the warping region image.
The method of claim 13, wherein the program,
Warping the first partial warping region image through a warping algorithm using the first feature point and the midpoint included in the first partial warping region image among the feature point pairs included in the first partial warping region image,
Among the pair of feature points included in the second partial warping region image, the second feature warping region image included in the second partial warping region image and the midpoint are input to warp the second partial warping region image. Image matching system using journal warping.

Images