A method and device for calculating the posture of a non-cooperative target in space

Abstract

The present invention provides a method and device for solving the pose of a non-cooperative target in space. The method includes: obtaining an image to be processed of the non-cooperative target; determining the feature points of the same name between adjacent images to be processed, and calculating a homography matrix; obtaining an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points; using a plane induced parallax bundle method, determining the target core point from the core points corresponding to the initial pose solution; determining the basic matrix corresponding to the target core point, and decomposing it to obtain the target pose solution. The pose solving method for a non-cooperative target in space provided by the present solution effectively solves the problem of inaccurate and unstable pose inversion information of existing non-cooperative targets in space.

Description

Technical Field

[0001] The present invention relates to the technical field of space image processing, and in particular to a method and device for solving the posture of a space non-cooperative target. Background Art

[0002] With the development of space technology, topics such as on-orbit capture and repair of spacecraft and cleaning of space debris have gradually become hot research directions, and space targets have become the core of major strategic concerns such as resource grabbing and national defense. Among them, situational awareness of important space non-cooperative targets is an indispensable step to gain the initiative of space control, but the existing methods for solving the attitude information of space non-cooperative targets in planar degradation scenarios still have the problem of inaccurate and unstable attitude information. Summary of the invention

[0003] The embodiment of the present invention provides a method and device for solving the posture of a space non-cooperative target, which effectively solves the problem of inaccurate and unstable posture inversion information of the existing space non-cooperative target.

[0004] In a first aspect, an embodiment of the present invention provides a method for solving a posture of a spatial non-cooperative target, comprising:

[0005] Obtaining images of non-cooperative targets to be processed;

[0006] Determine the feature points with the same name between adjacent images to be processed, and calculate the homography matrix;

[0007] Obtaining an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points;

[0008] Using a plane induced parallax bundle method, a target core point corresponding to the initial pose solution is obtained;

[0009] The basic matrix corresponding to the target core point is determined and decomposed to obtain the target pose solution.

[0010] Optionally, determining feature points with the same name between adjacent images to be processed and calculating a homography matrix includes:

[0011] Extracting features from the image to be processed to obtain feature points;

[0012] Matching feature points of the same name between adjacent images to be processed to obtain feature points of the same name between adjacent images to be processed;

[0013] The homography matrix is ​​calculated based on the AC-RANSAC algorithm according to the feature points with the same name.

[0014] Optionally, obtaining an initial pose solution according to the homography matrix includes:

[0015] Obtaining a first pose solution according to the homography matrix;

[0016] Based on plane ipsilaterality and visibility constraints, the first pose solutions are screened to obtain the initial pose solutions; wherein the number of the first pose solutions is greater than the number of the initial pose solutions.

[0017] Optionally, the obtaining of the target core point corresponding to the initial pose solution by using a plane induced parallax bundle method includes:

[0018] Obtaining a core point corresponding to the initial pose solution;

[0019] Draw the plane induced parallax bundle corresponding to the feature point with the same name according to the preset point position error;

[0020] Determine the number of times each of the core points falls into the plane induced parallax bundle, and select the first core point with the largest number of corresponding times; wherein the number is not greater than the number of the feature points with the same name;

[0021] Determine whether the number of times the first core point corresponds is greater than a preset threshold;

[0022] If so, the first core point is determined to be a target core point.

[0023] Optionally, determining a basic matrix corresponding to the target core point and decomposing it to obtain a target pose solution includes:

[0024] Determining the basic matrix according to the target core points and the homography matrix;

[0025] Determine the camera matrix of the camera that acquires the image to be processed;

[0026] Obtaining an essential matrix according to the camera matrix and the basic matrix;

[0027] The essential matrix is ​​decomposed to obtain the target pose solution.

[0028] Optionally, decomposing the essential matrix to obtain the target pose solution includes:

[0029] Decomposing the essential matrix to obtain a second posture solution;

[0030] Based on the plane ipsilaterality constraint, the second posture solution is screened to obtain the target posture solution.

[0031] In a second aspect, an embodiment of the present invention further provides a posture solving device for a spatial non-cooperative target, comprising:

[0032] An acquisition module, used for acquiring the image to be processed of the non-cooperative target;

[0033] A processing module, used for determining feature points with the same name between adjacent images to be processed, and calculating a homography matrix;

[0034] A first pose extraction module, used to obtain an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points;

[0035] A core point determination module, used for obtaining a target core point corresponding to the initial pose solution by using a plane induced parallax bundle method;

[0036] The second posture extraction module is used to determine the basic matrix corresponding to the target core point and decompose it to obtain the target posture solution.

[0037] Optionally, the core point determination module is further configured to perform the following operations:

[0038] Obtaining a core point corresponding to the initial pose solution;

[0039] Draw the plane induced parallax bundle corresponding to the feature point with the same name according to the preset point position error;

[0040] Determine the number of times each of the core points falls into the plane induced parallax bundle, and select the first core point with the largest number of corresponding times; wherein the number is not greater than the number of the feature points with the same name;

[0041] Determine whether the number of times the first core point corresponds is greater than a preset threshold;

[0042] If so, the first core point is determined to be a target core point.

[0043] In a third aspect, an embodiment of the present invention further provides a computing device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the method for solving the posture of a spatial non-cooperative target as described in any one of the above items.

[0044] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed in a computer, the computer is enabled to execute any of the above-mentioned methods for solving the posture of a non-cooperative target in space.

[0045] The embodiment of the present invention provides a method and device for solving the pose of a spatial non-cooperative target. The method uses the feature points of the same name between adjacent images to be processed to calculate a homography matrix, uses the homography matrix as a non-degenerate mapping model, and solves the initial pose solution. Since each initial pose solution corresponds to a core point, the core point position is constrained by a plane-induced parallax bundle method to obtain a target core point, and then the target core point is decomposed based on the basic matrix corresponding to the target core point to obtain the target pose solution. In this way, the present invention adopts a plane-induced parallax bundle method based on the core point position constraint, which can fully utilize the parallax information to obtain a non-degenerate mapping model in the case of plane degradation, and ensures the robustness of the non-degenerate mapping model solution through the position constraint of the core point, thereby improving the accuracy of the spatial non-cooperative target pose inversion and enhancing the overall reliability of the spatial non-cooperative target pose inversion system. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 It is a flow chart of a method for solving the posture of a spatial non-cooperative target provided by one embodiment of the present invention;

[0048] Figure 2 is a schematic diagram of determining a target core point by using a plane-induced parallax bundle provided by an embodiment of the present invention;

[0049] Figure 3 is a hardware architecture diagram of a computing device provided by an embodiment of the present invention;

[0050] Figure 4 It is a structural diagram of a posture solving device for a spatial non-cooperative target provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0052] At present, there are three methods for solving the pose information of spatial targets in plane-degraded scenes. The first one, Geometric Robust Information Criterion (GRIC), can distinguish whether to use the basic matrix or the homography matrix, but for quasi-plane degradation problems, GRIC cannot correctly handle the degraded data and always tends to only obtain the homography matrix. The second one, proposed by Rebert et al. in 2019, uses the "plane-induced parallax" to calculate the correct basic matrix. This method uses the centroid of the polygon in the maximum overlapping area of ​​many parallax bundles as the core point of the image, but does not take into account the situation that all spatial points are located inside the plane, and the method of using the centroid of the polygon as the core point does not use the geometric prior properties of the homography matrix decomposition result as a constraint condition, which may result in incorrect results. The third one is a method based on the improvement of RANSAC, referred to as "DEGENSAC". The algorithm first solves the basic matrix, and then verifies by testing whether there are 5 pairs of matching points related to the homography matrix consistent with the basic matrix. However, in a single plane-degraded scene, it may solve an erroneous basic matrix that is consistent with the homography matrix.

[0053] Therefore, in order to improve the accuracy of posture inversion of spatial non-cooperative targets in planar degraded scenes, the present invention proposes a plane induced parallax bundle method based on kernel point position constraints for posture solution of spatial non-cooperative targets.

[0054] The following is the concept of the present invention: Figure 1 As shown, an embodiment of the present invention provides a method for solving the pose of a spatial non-cooperative target, the method comprising:

[0055] Step 100, obtaining an image to be processed of a non-cooperative target;

[0056] Step 102, determining feature points with the same name between adjacent images to be processed, and calculating a homography matrix;

[0057] Step 104, obtaining an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points;

[0058] Step 106, using the plane induced parallax bundle method, determining the target core point from the core points corresponding to the initial pose solution;

[0059] Step 108, determine the basic matrix corresponding to the target core point, and decompose it to obtain the target pose solution.

[0060] In the embodiment of the present invention, the homography matrix is ​​calculated by using the feature points of the same name between adjacent images to be processed, and the homography matrix is ​​used as a non-degenerate mapping model to solve the initial pose solution. Since each initial pose solution corresponds to a core point, the core point position is constrained by the plane induced parallax bundle method to obtain the target core point, and then the target core point is decomposed based on the basic matrix corresponding to the target core point to obtain the target pose solution. In this way, the robustness of the non-degenerate mapping model solution is ensured by the position constraint of the core point, the accuracy of the spatial non-cooperative target pose inversion is improved, and the overall reliability of the spatial non-cooperative target pose inversion system is enhanced.

[0061] Described below Figure 1 How the various steps are performed.

[0062] First, with respect to step 100 , a camera is used to acquire at least two frames of images to be processed of a non-cooperative target.

[0063] For step 102, the feature points with the same name between adjacent images to be processed are determined, and the homography matrix is ​​calculated, including:

[0064] Perform feature extraction on the image to be processed to obtain feature points;

[0065] Matching feature points of the same name between adjacent images to be processed to obtain feature points of the same name between adjacent images to be processed;

[0066] According to the feature points with the same name, the homography matrix is ​​calculated based on the AC-RANSAC algorithm.

[0067] It should be noted that AC-RANSAC aims to find a consistent set containing a controllable number of false alarms (NFA). It is based on actual data and avoids relying on experience to set the inlier threshold. It is a RANSAC method that does not require manual setting of parameters.

[0068] Specifically, first construct the formula for solving NFA:

[0069]

[0070] Where n is the number of all pairs of feature points with the same name; k is the number of assumed internal points; ε k is the kth smallest residual value of the homography matrix H among all feature points with the same name; α0 is the proportion of the area of ​​a circle with a radius of 1 pixel in the image to be processed, indicating the probability of having a residual of less than 1 pixel under uniform distribution, where the solution formula of α0 is

[0071]

[0072] Where A represents the area of ​​the image to be processed;

[0073] When the homography matrix H meets the following conditions, it is considered valid, otherwise it is considered invalid:

[0074]

[0075] In order to save computing time, after finding a valid homography matrix H, the iteration is stopped and the inliers of the homography matrix H are resampled 10 times, and finally the correct homography matrix H' (i.e., as a non-degenerate mapping model), the number of inliers k', the inlier set P corresponding to H', and the maximum residual ε in the inlier set are obtained.

[0076] It should be noted that ε k It is the kth residual value after arranging all residual values ​​from small to large.

[0077] In the present invention, a feature point extraction algorithm and a matching algorithm are used to obtain feature points with the same name in adjacent images to be processed, and then the AC-RANSAC strategy is used to calculate the homography matrix reflecting the geometric relationship between the binocular image and the spatial plane as a degenerate mapping model reflecting the spatial plane posture.

[0078] In step 104, an initial pose solution is obtained according to the homography matrix, including:

[0079] Get the first pose solution according to the homography matrix;

[0080] Based on the plane homolaterality and visibility constraints, the first pose solutions are screened to obtain initial pose solutions; wherein the number of the first pose solutions is greater than the number of the initial pose solutions.

[0081] Specifically, the camera matrix and calibration matrix of the camera in step 100 are obtained. By decomposing the homography matrix, a total of 8 sets of first-position solutions for position and posture can be obtained. However, not all of these 8 sets of solutions are valid solutions, and some of them do not meet the actual situation. First, the camera positions corresponding to the two images to be processed can only be located on the same side of the scene plane. This plane same-side constraint can eliminate 4 sets of meaningless solutions; secondly, all scene points can only be located in front of the camera. After this visibility constraint, only 2 sets of solutions are left that meet the actual situation, that is, the initial posture solution is obtained;

[0082] According to the results of homography matrix decomposition, two sets of pose solutions can be obtained, including two translation vectors, so two possible core point positions can be calculated:

[0083] e′=Kt (4)

[0084] Where K is the camera calibration matrix; t is the fourth column of the camera matrix P', which is used to represent the relative displacement.

[0085] In step 106, the target core point is determined from the core points corresponding to the initial pose solution using the plane induced parallax bundle method, including:

[0086] Get the core point corresponding to the initial pose solution;

[0087] Draw the plane induced parallax bundle corresponding to the feature points with the same name according to the preset point position error;

[0088] Determine the number of times each kernel point falls into the plane induced parallax bundle, and select the first kernel point with the largest number of corresponding times; wherein the number is not greater than the number of feature points with the same name;

[0089] Determine whether the number of times the first core point corresponds is greater than a preset threshold;

[0090] If so, the first core point is determined to be the target core point.

[0091] It should be noted that if the judgment result is no, or the number of correspondences of each core point is the same, it is considered that the homography matrix used as the degenerate mapping model reflecting the spatial plane posture in step 102 is still a degenerate plane homography matrix.

[0092] It should be noted that the "plane induced parallax" method proposed by Rebert et al. in 2019 is used to obtain the preset point error of each feature point with the same name, and the corresponding plane induced parallax bundles are drawn respectively, and then the centroid of the polygon in the maximum overlapping area of ​​many parallax bundles is used as the core point of the image, but the core point is constrained and screened by the above method to further determine the correct target core point. Among them, each feature point with the same name corresponds to a preset point error.

[0093] In the present invention, due to the uncertainty of the decomposition results of the homography matrix, the initial pose solution usually has erroneous solutions. The plane induced parallax bundle method can be used to screen the core points corresponding to the initial pose solution obtained in step 104, so as to obtain the correct target core points, and use the basic matrix corresponding to the target core points as the final non-degenerate mapping model to solve the final pose information. In this way, the degenerate mapping model is converted into a non-degenerate mapping model for solution, which effectively solves the problem of inaccurate and unstable inversion information of existing spatial non-cooperative target postures in plane degenerate scenes.

[0094] Specifically, for example, Figure 4 As shown in , the core points corresponding to the initial pose solution are e1' and e2' respectively, and there are feature points x1 and x2 with the same name in the image to be processed. First, obtain the preset point position error of the feature point x1 with the same name and the preset point position error of the feature point x2 with the same name, and draw the plane induced parallax bundles corresponding to x1 and x2 respectively, as shown in Figure 4It can be seen that the core point e1' only falls into the plane induced disparity bundle corresponding to x2, the number of times is 1; the core point e2' falls into the plane induced disparity bundle corresponding to x1 and x2, the number of times is 2. When the preset threshold is 1, the core point e2' is the target core point.

[0095] In step 108, the basic matrix corresponding to the target core point is determined and decomposed to obtain the target pose solution, including:

[0096] Determine the basic matrix according to the target core points and the homography matrix;

[0097] Determine the camera matrix of the camera that acquires the image to be processed;

[0098] According to the camera matrix and the basic matrix, the essential matrix is ​​obtained;

[0099] Decompose the essential matrix to obtain the target pose solution.

[0100] In a preferred embodiment, in step 108, the essential matrix is ​​decomposed to obtain a target pose solution, including:

[0101] Decompose the essential matrix to obtain the second pose solution;

[0102] Based on the plane ipsilaterality constraint, the second pose solution is screened to obtain the target pose solution.

[0103] Specifically, the basic matrix is ​​determined by the following formula:

[0104] F=[e'] x H' (5)

[0105] Among them, H' is used to represent the homography matrix; e' is used to represent the target core point; F is used to represent the basic matrix;

[0106] The essential matrix is ​​determined by the following formula:

[0107] E=K -T FK (6)

[0108] Among them, K is used to represent the camera matrix; F is used to represent the basic matrix; E is used to represent the essential matrix;

[0109] Then the essential matrix is ​​decomposed to obtain four groups of pose solutions, and then the only group of pose solutions that meets the plane homolaterality constraint is taken as the final solution of the spatial target pose (i.e., the target pose solution).

[0110] like Figure 3 , Figure 4As shown, an embodiment of the present invention provides a posture solving device for a non-cooperative target in space. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 3 As shown, a hardware architecture diagram of a computing device where a posture solving device for a space non-cooperative target provided by an embodiment of the present invention is located, except Figure 3 In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the computing device in which the device is located in the embodiment may also generally include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 4 As shown, as a device in a logical sense, the CPU of the computing device in which it is located reads the corresponding computer program in the non-volatile memory into the memory and runs it. This embodiment provides a posture solving device for a spatial non-cooperative target, including: an acquisition module 400, a processing module 402, a first posture extraction module 404, a core point determination module 406 and a second posture extraction module;

[0111] An acquisition module 400 is used to acquire an image to be processed of a non-cooperative target;

[0112] The processing module 402 is used to determine the feature points with the same name between adjacent images to be processed and calculate the homography matrix;

[0113] The first pose extraction module 404 is used to obtain an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points;

[0114] A core point determination module 406 is used to obtain a target core point corresponding to an initial pose solution by using a plane induced parallax bundle method;

[0115] The second pose extraction module 408 is used to determine the basic matrix corresponding to the target core point and decompose it to obtain the target pose solution.

[0116] In some specific implementations, the acquisition module 400 can be used to execute the above step 100, the processing module 402 can be used to execute the above step 102, the first pose extraction module 404 can be used to execute the above step 104, the core point determination module 406 can be used to execute the above step 106, and the second pose extraction module 408 can be used to execute the above step 108.

[0117] In some specific implementations, the processing module 402 is further configured to perform the following operations:

[0118] Perform feature extraction on the image to be processed to obtain feature points;

[0119] Matching feature points of the same name between adjacent images to be processed to obtain feature points of the same name between adjacent images to be processed;

[0120] According to the feature points with the same name, the homography matrix is ​​calculated based on the AC-RANSAC algorithm.

[0121] In some specific implementations, the first pose extraction module 404 is further configured to perform the following operations:

[0122] Get the first pose solution according to the homography matrix;

[0123] Based on the plane homolaterality and visibility constraints, the first pose solutions are screened to obtain initial pose solutions; wherein the number of the first pose solutions is greater than the number of the initial pose solutions.

[0124] In some specific implementations, the core point determination module 406 is further configured to perform the following operations:

[0125] Get the core point corresponding to the initial pose solution;

[0126] Draw the plane induced parallax bundle corresponding to the feature points with the same name according to the preset point position error;

[0127] Determine the number of times each kernel point falls into the plane induced parallax bundle, and select the first kernel point with the largest number of corresponding times; wherein the number is not greater than the number of feature points with the same name;

[0128] Determine whether the number of times the first core point corresponds is greater than a preset threshold;

[0129] If so, the first core point is determined to be the target core point.

[0130] In some specific implementations, the second posture extraction module 408 is further configured to perform the following operations:

[0131] Determine the basic matrix according to the target core points and the homography matrix;

[0132] Determine the camera matrix of the camera that acquires the image to be processed;

[0133] According to the camera matrix and the basic matrix, the essential matrix is ​​obtained;

[0134] The essential matrix is ​​decomposed to obtain the second pose solution, and based on the plane homolaterality constraint, the second pose solution is screened to obtain the target pose solution.

[0135] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a posture solving device for a space non-cooperative target. In other embodiments of the present invention, a posture solving device for a space non-cooperative target may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0136] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0137] An embodiment of the present invention further provides a computing device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a method for solving the posture of a spatial non-cooperative target in any embodiment of the present invention is implemented.

[0138] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a method for solving the posture of a spatial non-cooperative target in any embodiment of the present invention.

[0139] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0140] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0141] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0142] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0143] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0144] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical factors in the process, method, article or device including the elements.

[0145] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for solving the pose of a non-cooperative target in space, characterized in that: include: Obtaining images of non-cooperative targets to be processed; Determining feature points with the same name between adjacent images to be processed and calculating a homography matrix, including: extracting features from the images to be processed to obtain feature points; matching feature points with the same name between adjacent images to be processed to obtain feature points with the same name between adjacent images to be processed; and calculating the homography matrix based on the AC-RANSAC algorithm according to the feature points with the same name; Obtaining an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points; Determine the target core point from the core points corresponding to the initial pose solution by using the plane induced parallax bundle method, including: obtaining the core point corresponding to the initial pose solution; drawing the plane induced parallax bundle corresponding to the feature point with the same name according to the preset point position error; determining the number of times each core point falls into the plane induced parallax bundle, and screening out the first core point with the largest number of corresponding times; wherein the number is not higher than the number of the feature points with the same name; judging whether the number of times the first core point corresponds is greater than a preset threshold; if so, determining the first core point as the target core point; The basic matrix corresponding to the target core point is determined and decomposed to obtain the target pose solution.

2. The method according to claim 1, characterized in that The obtaining an initial pose solution according to the homography matrix comprises: Obtaining a first pose solution according to the homography matrix; Based on plane ipsilaterality and visibility constraints, the first pose solutions are screened to obtain the initial pose solutions; wherein the number of the first pose solutions is greater than the number of the initial pose solutions.

3. The method according to any one of claims 1 to 2, characterized in that: The determining of the basic matrix corresponding to the target core point and decomposing the basic matrix to obtain the target pose solution includes: Determining the basic matrix according to the target core points and the homography matrix; Determine the camera matrix of the camera that acquires the image to be processed; Obtaining an essential matrix according to the camera matrix and the basic matrix; The essential matrix is ​​decomposed to obtain the target pose solution.

4. The method according to claim 3, characterized in that: Decomposing the essential matrix to obtain the target pose solution includes: Decomposing the essential matrix to obtain a second posture solution; Based on the plane ipsilaterality constraint, the second posture solution is screened to obtain the target posture solution.

5. A posture solving device for a space non-cooperative target, characterized in that: include: An acquisition module, used for acquiring the image to be processed of the non-cooperative target; A processing module, used for determining feature points with the same name between adjacent images to be processed, and calculating a homography matrix; A first pose extraction module, used to obtain an initial pose solution according to the homography matrix; wherein different initial pose solutions correspond to different core points; A core point determination module, used to determine a target core point from the core points corresponding to the initial pose solution by using a plane induced parallax bundle method; A second posture extraction module is used to determine the basic matrix corresponding to the target core point and decompose it to obtain a target posture solution; The processing module is also used to perform the following operations: Extracting features from the image to be processed to obtain feature points; matching feature points of the same name between adjacent images to be processed to obtain feature points of the same name between adjacent images to be processed; and calculating the homography matrix based on the AC-RANSAC algorithm according to the feature points of the same name; The core point determination module is also used to perform the following operations: Obtaining a core point corresponding to the initial pose solution; Draw the plane induced parallax bundle corresponding to the feature point with the same name according to the preset point position error; Determine the number of times each of the core points falls into the plane induced parallax bundle, and select the first core point with the largest number of corresponding times; wherein the number is not greater than the number of the feature points with the same name; Determine whether the number of times the first core point corresponds is greater than a preset threshold; If so, the first core point is determined to be a target core point.

6. A computing device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method according to any one of claims 1 to 4.

Images