A method and system for geometric matching of Fourier Merlin transform images with decoupling in the pole logarithmic domain

Abstract

This invention provides a Fourier-Melin transform image geometric matching method and system with decoupled logarithmic domain, comprising: estimating the rotation angle using Fourier-Melin transform; rotating the image to be matched in the current frame relative to the template image in the previous frame to obtain an axially aligned image search region; transforming the template image and the image to be matched from Cartesian coordinates to logarithmic polar coordinates to obtain the image search region corresponding to the logarithmic polar coordinates; sequentially segmenting the image search regions corresponding to the image to be matched and the template image along the θ-axis according to a set ratio; stitching the horizontally correlated and vertically correlated image regions into two twin image search regions; calculating the cross power spectrum through phase correlation to obtain the scale factors in the horizontal and vertical directions; and aligning the images to be matched according to the scale factors in the two directions. This invention has the ability to estimate scale and rotation proportionally, and is scale adaptive, and is robust to out-of-plane flipping and rotation of the target object.

Description

Technical Field

[0001] This invention relates to the field of image registration technology, and in particular to a geometric matching method and system for Fourier Merlin transform images based on pole-log domain decoupling. Background Technology

[0002] The Fourier-Merlin transform (FMT) image registration method is an image pair matching algorithm based on Fourier shift theory and phase correlation methods. It can be used to solve image detection and image registration problems. Compared with traditional methods such as SIFT and SURF, it can effectively achieve accurate geometric matching when there are large-angle rotations between the images and is less time-consuming. Traditional Fourier-Merlin transform and improved image matching methods have a default premise: the template and the two images to be registered can only have the same aspect ratio, such as a large round coin and another round coin, or a large square and a small square. The scale and rotation estimation are proportional. Therefore, this method can only handle similar transformations with two degrees of freedom. However, when the target is rotated out of plane, the aspect ratio will inevitably change, and the scale changes in the width and length directions are not consistent. This makes the single scale factor estimated by the traditional Fourier-Merlin transform unable to represent the scale changes in both width and height directions.

[0003] Current extensions and applications based on Fourier-Melin transform, such as patent CN202010794306.4 entitled "An Extended Fourier-Melin Localization Algorithm Applied to Multi-Depth Scenes," utilize interval sampling to obtain a scale vector and combine it with Fourier-Melin transform to achieve robot localization and navigation in multi-depth scenes, thus solving the problem that the original Fourier-Melin algorithm can only be applied to planar scenes and the scene needs to be parallel to the camera's imaging plane. However, it relies on multiple matching, resulting in high time complexity. Another example is patent CN201910345255.4 entitled "A Cargo Localization Method Based on Fourier-Melin Transform," which uses traditional Fourier-Melin transform image registration to calculate the pose relationship between cargo and the retrieval mechanism to achieve high-precision localization of various shelves and cargo. However, when the target undergoes out-of-plane flipping within the camera, this method cannot achieve accurate localization.

[0004] Therefore, for fast image geometric matching in high dynamic and large-angle scenes, traditional image matching methods and various matching methods based on Fourier-Merlin transform cannot simultaneously achieve matching accuracy, computational efficiency, and adaptability to size changes. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a Fourier-Merlin transform image geometric matching method and system with decoupled pole-log domain. It aims to solve the problems that traditional matching methods such as SIFT cannot accurately match large-angle rotations and that the size estimation of Fourier-Merlin transform matching methods cannot adapt to changes in image aspect ratio without increasing computational complexity.

[0006] The technical solution adopted in this invention is as follows:

[0007] This invention provides a geometric matching method for Fourier Merlin transform images with decoupled pole-log domain, comprising the following steps:

[0008] Estimating rotation angle using Fourier-Merlin transform , will the current frame i The image to be matched relative to the previous frame i -1 template image rotation Obtain the axially aligned image search region ;

[0009] The template image and the image to be matched are transformed from Cartesian coordinates to logarithmic polar coordinates to obtain the image search region. Image search region corresponding to logarithmic polar coordinates :

[0010]

[0011] in, and The first i -1 frame, the i The image search area corresponding to the frame image; Indicates along Cartesian coordinates polar axis, Represents pixels and The included angle, express and exist An increment in the axial direction, express and exist An increment in the axial direction;

[0012] Based on the different phases of the image regions, the image search region is... According to the set ratio The axis is divided into five parts in sequence. , , , , Image region , , It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0013] Will be in the image search area Image search region on the template image for matching Along the axis according to the set proportion The direction sequence is divided into five parts. , , , , Image region , and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0014] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0015] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0016] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0017] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0018] Registration and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction ;

[0019] Using scale factors in two directions , The images to be matched are aligned by affine transformation to obtain the final matching result.

[0020] The set ratio is 1:2:2:2:1.

[0021] The registration and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction ,include:

[0022] Calculate the cross power spectrum of the search regions of the two images to be registered in the frequency domain:

[0023]

[0024] In the formula, Represents the cross power spectrum in the Fourier domain. These represent the horizontal and vertical coordinates of the Fourier domain after frequency shift, respectively, with subscripts indicating the coordinates. These represent the horizontal and vertical directions, respectively. Indicates the reorganization of the region or , express The discrete Fourier transform form; express The complex conjugate, express The discrete Fourier transform form, Indicates the reorganization of the region or ;

[0025] Calculate the Dirac function:

[0026]

[0027] in, Represents the horizontal and vertical coordinates within the spatial domain The Dirac function value, Indicates the inverse Fourier transform;

[0028] Find the Dirac function The peak value is obtained, and the displacement is obtained by taking the coordinates of the peak value. According to displacement Calculate the rotation angle and scaling factor .

[0029] Calculate the rotation angle using the following formula. and scaling factor :

[0030]

[0031] In the formula, Horizontal scaling factor Or the vertical scaling factor ; for The length of the shaft, Indicates the limiting factor; It is the maximum value of the response. , , This represents the undecoupled cross power spectrum. This indicates the undecoupled displacement.

[0032] The Cartesian coordinates polar axis Pixels and The included angle Calculated using the following formula:

[0033]

[0034] In the formula, Let be the ordinate of the Cartesian coordinate system. Indicates the pivot point.

[0035] The present invention also provides a Fourier Merlin transform image geometric matching system with decoupled pole-log domain, comprising:

[0036] The matching image rotation alignment module uses Fourier-Merlin transform to estimate the rotation angle. , will the current frame i The image to be matched relative to the previous frame i -1 template image rotation Obtain the axially aligned image search region ;

[0037] The image logarithmic polar coordinate transformation module transforms the template image and the image to be matched from the Cartesian coordinate system to the logarithmic polar coordinate system to obtain the image search region. Image search region corresponding to logarithmic polar coordinates :

[0038]

[0039] in, and The first i -1 frame, the i The image search area corresponding to the frame image; Indicates along Cartesian coordinates polar axis, Represents pixels and The included angle, express and exist An increment in the axial direction, express and exist An increment in the axial direction;

[0040] The region segmentation module in logarithmic polar coordinates segments the image search region according to the phase of the image region. According to the set ratio The axis is divided into five parts in sequence. , , , , Image region , , It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0041] Will be in the image search area Image search region on the template image for matching Along the axis according to the set proportion The direction sequence is divided into five parts. , , , , Image region , and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0042] The image patch filtering and stitching module is used for:

[0043] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0044] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0045] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0046] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0047] The phase correlation calculation dual-scale factor module, which is based on registration... and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction ;

[0048] The matching image scale alignment module utilizes... , The images to be matched are aligned by affine transformation to obtain the final matching result.

[0049] The set ratio is 1:2:2:2:1.

[0050] The registration and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction ,include:

[0051] Calculate the cross power spectrum of the search regions of the two images to be registered in the frequency domain:

[0052]

[0053] In the formula, Represents the cross power spectrum in the Fourier domain. These represent the horizontal and vertical coordinates of the Fourier domain after frequency shift, respectively, with subscripts indicating the coordinates. These represent the horizontal and vertical directions, respectively. Indicates the reorganization of the region or , express The discrete Fourier transform form; express The complex conjugate, express The discrete Fourier transform form, Indicates the reorganization of the region or ;

[0054] Calculate the Dirac function:

[0055]

[0056] in, Represents the horizontal and vertical coordinates within the spatial domain The Dirac function value, Indicates the inverse Fourier transform;

[0057] Find the Dirac function The peak value is obtained, and the displacement is obtained by taking the coordinates of the peak value. According to displacement Calculate the rotation angle and scaling factor .

[0058] Calculate the rotation angle using the following formula. and scaling factor :

[0059]

[0060] In the formula, Horizontal scaling factor Or the vertical scaling factor ; for The length of the shaft, Indicates the limiting factor; It is the maximum value of the response. , , This represents the undecoupled cross power spectrum. This indicates the undecoupled displacement.

[0061] The Cartesian coordinates polar axis Pixels and The included angle Calculated using the following formula:

[0062]

[0063] In the formula, Let be the ordinate of the Cartesian coordinate system. Indicates the pivot point.

[0064] The beneficial effects of this invention are as follows:

[0065] This invention extends the degrees of freedom of the traditional Fourier-Merlin transform, enabling it to adapt to changes in scale and aspect ratio without significantly increasing computational complexity. It addresses the limitations of traditional matching methods like SIFT in accurately matching large-angle rotations, and the inability of Fourier-Merlin transform matching methods to adapt to changes in image aspect ratio. This invention achieves scale-independent Fourier-Merlin transform by decoupling the image, while also possessing similarity transformation capabilities (proportional scale and rotation estimation) and scale adaptability. It can be used for scale-rotation-adaptive visual recognition and localization, and is robust to out-of-plane flipping and rotation of objects.

[0066] The decoupled Fourier-Merlin transform proposed in this invention utilizes logarithmic polar domain transformation and Fourier shift target, combining the Fourier-Merlin transform matching method with the independence of image scale changes in logarithmic polar coordinates, solving the scale factor in two directions separately, extending the equal width and height scale estimation to a more general arbitrary width and height scale estimation, realizing 3-DOF rotation and scale estimation, and significantly improving the accuracy and efficiency of image geometric matching.

[0067] Other features and advantages of the invention will be set forth in the following description or may be learned by practicing the invention. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating the method of an embodiment of the present invention.

[0069] Figure 2 This is a schematic diagram illustrating the principle of the method in an embodiment of the present invention, using a standard diagram as an example.

[0070] Figure 3 This is a schematic diagram illustrating the principle of independence of scale changes in the logarithmic polar coordinate system of the method in this embodiment of the invention.

[0071] Figure 4 This is a schematic diagram of the result obtained during the processing of a specific image to be matched by the method of this embodiment of the invention. Detailed Implementation

[0072] The specific embodiments of the present invention are described below with reference to the accompanying drawings.

[0073] Example 1

[0074] See Figure 1 This embodiment of a method for geometric matching of Fourier Merlin transform images with decoupled pole-log domain includes the following steps:

[0075] S1. Estimate the rotation angle using Fourier-Merlin transform. , will the current frame i The image to be matched relative to the previous frame i -1 template image rotation Obtain the axially aligned image search region .

[0076] Among them, the image search area It is the region on the image to be matched that is larger than the smallest bounding box that contains the target to be tracked.

[0077] S2. Based on the properties of logarithmic polar coordinate mapping, transform the template image and the image to be matched from the Cartesian coordinate system to the logarithmic polar coordinate system to obtain the image search region. Image search region corresponding to logarithmic polar coordinates :

[0078]

[0079] Among them, the logarithmic polar coordinate transformation has scale invariance and rotation invariance, as shown in the above equation. and The first i -1 frame, the i The image search area corresponding to the frame image; Indicates along Cartesian coordinates polar axis, Represents pixels and The included angle, express and exist An increment in the axial direction, express and exist An increment in the axial direction;

[0080] Wherein, along Cartesian coordinates polar axis Pixels and The included angle Calculated using the following formula:

[0081]

[0082] In the formula, Let be the ordinate of the Cartesian coordinate system. Indicates the pivot point.

[0083] S3. Based on the independence of scale changes in logarithmic polar coordinates, the image search region is divided according to the different phases of the image region. According to the set ratio The axis is divided into five parts in sequence. , , , , Image region , , It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0084] Will be in the image search area Image search region on the template image for matching Along the axis according to the set proportion The direction sequence is divided into five parts. , , , , Image region , and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction.

[0085] The set ratio is 1:2:2:2:1.

[0086] S4. After dividing the area, assemble it sequentially. The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0087] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0088] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0089] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. .

[0090] S5, Registration and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction Specifically, this includes:

[0091] Calculate the cross power spectrum of the search regions of the two images to be registered in the frequency domain:

[0092]

[0093] In the formula, To represent the cross power spectrum in the Fourier domain by utilizing the time-shifting property of the Fourier transform, These represent the horizontal and vertical coordinates of the Fourier domain after frequency shift, respectively, with subscripts indicating the coordinates. These represent the horizontal and vertical directions, respectively. Indicates the reorganization of the region or , express The discrete Fourier transform form; express The complex conjugate, express The discrete Fourier transform form, Indicates the reorganization of the region or ;

[0094] Calculate the Dirac function:

[0095]

[0096] in, Represents the horizontal and vertical coordinates within the spatial domain The Dirac function value, Indicates the inverse Fourier transform;

[0097] Find the Dirac function The peak value is obtained, and the displacement is obtained by taking the coordinates of the peak value. According to displacement Calculate the rotation angle according to the following formula. and scaling factor :

[0098]

[0099] In the formula, Horizontal scaling factor Or the vertical scaling factor ; for The length of the shaft, This represents a constraint factor used to control the strength of the decoupling effect; It is the maximum value of the response. , , This represents the undecoupled cross power spectrum. This indicates the undecoupled displacement.

[0100] S6. Utilizing scale factors in two directions , The images to be matched are aligned by affine transformation to obtain the final matching result.

[0101] See Figure 2 The document uses standard graphic examples to explain the conceptual design of the graphic transformation method in this embodiment, namely, the operation of scale and rotation estimation based on decoupled Fourier-Merlin transform. Figure 2 middle , , , , This refers to the embodiment of the present invention. , , , , , and , and This refers to the present embodiment. and , and As shown in the figure, the image regions represent different colors, and the image regions... , , It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction; similarly, image regions , and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction. Therefore, based on the different phases of the image regions, two twin pixel regions are decoupled in logarithmic polar coordinates, with each region containing scale information in different directions. The decoupled images are then combined and stitched together according to the scale information they contain to obtain two new pixel regions, and phase correlation is performed with the template to obtain the final scale factor.

[0102] Figure 3 This is a schematic diagram illustrating the independence of scale changes in logarithmic polar coordinates in this embodiment. For example... Figure 3As shown, sample image (a) is considered as the template image, and the other three sample images (b), (c), and (d) are magnified images of this template. (a'), (b'), (c'), and (d') are the logarithmic polar coordinate transformation results of (a), (b), (c), and (d), respectively. To highlight the areas with redundant pixels, the pixel values ​​of (a') are added to the corresponding pixel values ​​of (b'), (c'), and (d'), resulting in the new image shown in the second row: (a') + (b'), (a') + (c'), (a') + (d'). It is clear from the new image that the horizontal scale has expanded, represented by two consecutive regions of the same color in the image; the vertical scale has also expanded, represented by three consecutive regions of the same color in the image.

[0103] Figure 4 This is a specific example diagram illustrating scale estimation using decoupled Fourier-Merlin transform in this embodiment. To highlight the scale change, Figure 4 Two frames with significantly different aspect ratios were selected. The two input images in the first column were captured by UAV from different angles of the same target; therefore, the target's scale and aspect ratio are inconsistent in these two frames. It is evident that performing a logarithmic polar coordinate transformation on the template and the image to be matched, based on the target's change from the first column to the second, clearly shows the effect of the logarithmic polar coordinate transformation on the image pixels. That is, in the Cartesian coordinate system, pixel values ​​in the same direction from the target center to the surrounding area are mapped to the same phase in the logarithmic polar coordinate system. (Axis). Then, the horizontal and vertical directions of the target image search region are segmented and recombined to obtain twin image search regions, which are represented as two inverted triangular regions in Cartesian coordinates. Through decoupling and phase correlation operations in logarithmic polar coordinates, two different phase correlation responses in the fourth column can be obtained. In the horizontal direction, the phase correlation response of the twin image search region yields a scaling factor of 1.3611; similarly, in the vertical direction, a scaling factor of 1.0539 is obtained, which is basically consistent with the intuitive scale of the target in the two input images. However, when using the traditional Fourier-Mellin transform, the scaling factor calculated by phase correlation is 1.0653. A single scaling factor cannot characterize two directions, and therefore cannot adapt to aspect ratio changes. Although phase correlation processing is used twice to solve the aspect ratio adaptation problem, only half of the image search region is input each time, so the running time does not increase directly. Compared with other scale estimation methods, this embodiment is more effective and robust for out-of-plane rotation.

[0104] Example 2

[0105] This embodiment provides a Fourier Merlin transform image geometric matching system with pole-log domain decoupling corresponding to the method of Embodiment 1. The system includes:

[0106] The matching image rotation alignment module uses Fourier-Merlin transform to estimate the rotation angle. , will the current frame i The image to be matched relative to the previous frame i -1 template image rotation Obtain the axially aligned image search region ;

[0107] The image logarithmic polar coordinate transformation module transforms the template image and the image to be matched from the Cartesian coordinate system to the logarithmic polar coordinate system to obtain the image search region. Image search region corresponding to logarithmic polar coordinates :

[0108]

[0109] in, and The first i -1 frame, the i The image search area corresponding to the frame image; Indicates along Cartesian coordinates polar axis, Represents pixels and The included angle, express and exist An increment in the axial direction, express and exist An increment in the axial direction;

[0110] The region segmentation module in logarithmic polar coordinates segments the image search region according to the phase of the image region. According to the set ratio The axis is divided into five parts in sequence. , , , , Image region , , It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0111] Will be in the image search area Image search region on the template image for matching Along the axis according to the set proportion The direction sequence is divided into five parts. , , , , Image region , and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction;

[0112] The image patch filtering and stitching module is used for:

[0113] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0114] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0115] Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. ;

[0116] Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. ;

[0117] The phase correlation calculation dual-scale factor module, which is based on registration... and Obtain the scale factor in the horizontal direction Registration and Obtain the scale factor in the vertical direction ;

[0118] The matching image scale alignment module utilizes... , The images to be matched are aligned by affine transformation to obtain the final matching result.

[0119] In summary, this invention takes into account the difficulty of decomposing the scale changes of the target in the Cartesian coordinate system and the independence of the scale changes in the horizontal and vertical directions in the logarithmic polar coordinate system. It combines the Fourier-Merlin transform matching method with the independence of the scale changes of the image in the logarithmic polar coordinate system, solves the scale factors in the two directions separately, extends the equal width and height scale estimation to a more general arbitrary width and height scale estimation, realizes 3-degree-of-freedom rotation and scale estimation, and significantly improves the accuracy and efficiency of image geometric matching.

[0120] It will be understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A geometric matching method for Fourier-Merlin transform images decoupled in the pole-log domain, characterized in that, Includes the following steps: Estimating rotation angle using Fourier-Merlin transform Rotate the image to be matched in the current frame i relative to the template image in the previous frame i-1. Obtain the axially aligned image search region p i ; The template image and the image to be matched are transformed from Cartesian coordinates to logarithmic polar coordinates to obtain the image search region p. i Image search region corresponding to logarithmic polar coordinates in, and ρ represents the image search region corresponding to the (i-1)th frame and the i-th frame, respectively; ρ represents the polar axis along the Cartesian coordinate x, θ represents the angle between the pixel and ρ, and θ0 represents... and An increment in the θ-axis direction, logs represents and An increment in the ρ-axis direction; Based on the different phases of the image regions, the image search region is... Divide the parts sequentially along the θ axis according to a set ratio into five parts. Image region It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction; Will be in the image search area Image search region on the template image for matching Divide the parts sequentially along axis θ according to a set ratio into five parts. Image region and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction; Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. Registration and The horizontal scaling factor s is obtained. h Registration and Obtain the vertical scaling factor s v ; Using the scale factor s in two directions h s v The images to be matched are aligned by affine transformation to obtain the final matching result.

2. The method according to claim 1, characterized in that, The set ratio is 1:2:2:2:

1.

3. The method according to claim 1, characterized in that, The registration and The horizontal scaling factor s is obtained. h Registration and Obtain the vertical scaling factor s v ,include: Calculate the cross power spectrum of the search regions of the two images to be registered in the frequency domain: In the formula, Represents the cross-power spectrum in the Fourier domain, ω h,v ,ξ h,v Let h and v represent the horizontal and vertical coordinates of the Fourier domain after frequency shift, respectively, with the subscripts h and v indicating the horizontal and vertical directions, respectively. Indicates the reorganization of the region or express The discrete Fourier transform form; express The complex conjugate, express The discrete Fourier transform form, Indicates the reorganization of the region or Calculate the Dirac function: Where, δ(x) dfm ,y dfm ) represents the x and y coordinates within the spatial domain. dfm ,y dfm The Dirac function value, Indicates the inverse Fourier transform; Find the Dirac function δ(x) dfm ,y dfm) The peak value is obtained, and the displacement is obtained by taking the coordinates of the peak value. <x dfm ,y dfm >, according to displacement <x dfm ,y dfm Calculate the rotation angle θ and the scaling factor s.

4. The method according to claim 3, characterized in that, The rotation angle θ and the scaling factor s are calculated using the following formula: In the formula, s is the horizontal scaling factor. h Or the vertical scaling factor s v L is the length of the θ-axis, and κ represents the constraint factor. It is the maximum value of the response. This represents the undecoupled cross power spectrum. <x dfm ,y dfm > indicates an undecoupled displacement.

5. The method according to claim 1, characterized in that, The angle θ between the pixel and the polar axis ρ along the Cartesian coordinate x is calculated by the following formula: In the formula, y is the Cartesian coordinate, and (x0, y0) represents the center point.

6. A Fourier-Merlin transform image geometric matching system decoupled in the pole-log domain, characterized in that, include: The matching image rotation alignment module uses Fourier-Merlin transform to estimate the rotation angle. Rotate the image to be matched in the current frame i relative to the template image in the previous frame i-1. Obtain the axially aligned image search region p i ; The image logarithmic polar coordinate transformation module transforms the template image and the image to be matched from the Cartesian coordinate system to the logarithmic polar coordinate system to obtain the image search region p. i Image search region corresponding to logarithmic polar coordinates in, and ρ represents the image search region corresponding to the (i-1)th frame and the i-th frame, respectively; ρ represents the polar axis along the Cartesian coordinate x, θ represents the angle between the pixel and ρ, and θ0 represents... and An increment in the θ-axis direction, logs represents and An increment in the ρ-axis direction; The region segmentation module in logarithmic polar coordinates segments the image search region according to the phase of the image region. Divide the parts sequentially along the θ axis according to a set ratio into five parts. Image region It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction; Will be in the image search area Image search region on the template image for matching Divide the parts sequentially along axis θ according to a set ratio into five parts. Image region and It is only related to the scale change in the horizontal direction, image region and It is only related to scale changes in the vertical direction; The image patch filtering and stitching module is used for: Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. Sequential splicing The image region related to the horizontal scale change is used to obtain a new image search region. Sequential splicing The image region related to the vertical scale change is used to obtain a new image search region. The phase correlation calculation dual-scale factor module, which is based on registration... and The horizontal scaling factor s is obtained. h Registration and Obtain the vertical scaling factor s v ; The matching image scale alignment module utilizes s h s v The images to be matched are aligned by affine transformation to obtain the final matching result.

7. The system according to claim 6, characterized in that, The set ratio is 1:2:2:2:

1.

8. The system according to claim 6, characterized in that, The registration and The horizontal scaling factor s is obtained. h Registration and Obtain the vertical scaling factor s v ,include: Calculate the cross power spectrum of the search regions of the two images to be registered in the frequency domain: In the formula, Represents the cross-power spectrum in the Fourier domain, ω h,v ,ξ h,v Let h and v represent the horizontal and vertical coordinates of the Fourier domain after frequency shift, respectively, with the subscripts h and v indicating the horizontal and vertical directions, respectively. Indicates the reorganization of the region or express The discrete Fourier transform form; express The complex conjugate, express The discrete Fourier transform form, Indicates the reorganization of the region or Calculate the Dirac function: Where, δ(x) dfm ,y dfm ) represents the x and y coordinates within the spatial domain. dfm ,y dfm The Dirac function value, Indicates the inverse Fourier transform; Find the Dirac function δ(x) dfm ,y dfm The peak value is obtained, and the displacement is obtained by taking the coordinates of the peak value. <x dfm ,y dfm >, according to displacement <x dfm ,y dfm Calculate the rotation angle θ and the scaling factor s.

9. The system according to claim 8, characterized in that, The rotation angle θ and the scaling factor s are calculated using the following formula: In the formula, s is the horizontal scaling factor. h Or the vertical scaling factor s v L is the length of the θ-axis, and κ represents the constraint factor. It is the maximum value of the response. This represents the undecoupled cross power spectrum. <x dfm ,y dfm > indicates an undecoupled displacement.

10. The system according to claim 6, characterized in that, The angle θ between the pixel and the polar axis ρ along the Cartesian coordinate x is calculated by the following formula: In the formula, y is the Cartesian coordinate, and (x0, y0) represents the center point.

Images