Automatic Registration Method, Device and System for Remote Sensing Image Data

Abstract

The present invention relates to the technical field of image data processing, and particularly relates to an automatic registration method, device and system for remote sensing image data. The method includes: acquiring remote sensing images of a to-be-acquired area at different times; analyzing any two to-be-registered remote sensing images to obtain key points in the two to-be-registered remote sensing images; obtaining reliable matching point pairs in the two to-be-registered remote sensing images according to the local network structure corresponding to each key point in the two to-be-registered remote sensing images and the brightness change of the key point corresponding to the local network structure; matching all the reliable matching point pairs in the two to-be-registered remote sensing images to obtain a target remote sensing image after registration. The present disclosure can improve the accuracy of remote sensing image registration.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to an automatic registration method, device and system for remote sensing image data. Background Art

[0002] Image registration is a process of matching and superimposing two or more images acquired at different times, by different sensors (imaging devices), or under different conditions (such as illumination, camera position and angle, etc.).

[0003] Remote sensing images are crucial for environmental monitoring such as river seepage. By comparing remote sensing images acquired at different times, changes in river seepage can be identified, and natural disasters can be monitored, etc.

[0004] However, when automatically registering remote sensing images, due to the existence of self-similar structures in remote sensing images, for example, there are large areas of farmland in the images, and a certain feature in the large area of farmland appears similar patterns multiple times in the remote sensing image. When the matching algorithm selects features in the self-similar area, it may mistakenly match similar features at different positions as the same feature, resulting in deviations in the registration of remote sensing images, which is not conducive to subsequent environmental monitoring based on the registered remote sensing images. Summary of the Invention

[0005] In order to solve the technical problem of deviation in the registration of remote sensing images, the purpose of the present invention is to provide an automatic registration method, device and system for remote sensing image data, and the specific technical solutions adopted are as follows:

[0006] In the first aspect of the present disclosure, an automatic registration method for remote sensing image data is provided, and the method includes:

[0007] Obtain remote sensing images of a to-be-acquired area at different times;

[0008] Analyze any two to-be-registered remote sensing images, and obtain key points in the two to-be-registered remote sensing images;

[0009] According to the local network structure corresponding to each key point in the two to-be-registered remote sensing images and the brightness change of the key point corresponding to the local network structure, obtain reliable matching point pairs in the two to-be-registered remote sensing images;

[0010] Match all the reliable matching point pairs in the two to-be-registered remote sensing images to obtain a target remote sensing image after registration.

[0011] In one embodiment, obtaining reliable matching point pairs in the two remotely sensed images to be registered according to the local network structure corresponding to each key point in the two remotely sensed images to be registered and the brightness change of the key point corresponding to the local network structure includes:

[0012] Obtaining matching point pairs among the key points of the source image and the key points of the target image; the source image and the target image are the two remotely sensed images to be registered;

[0013] Obtaining a first stable matching point pair possibility value for each pair of the matching point pairs;

[0014] Obtaining first target matching point pairs for which the first stable matching point pair possibility value is greater than a first preset threshold;

[0015] Obtaining a first local network structure corresponding to each remaining key point of the source image; the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pairs in the source image;

[0016] Obtaining a second local network structure corresponding to each remaining key point of the target image; the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pairs in the target image;

[0017] Obtaining reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure;

[0018] Obtaining the reliable matching point pairs according to the first target matching point pairs and second target matching point pairs, where the second target matching point pairs are composed of the remaining key points in the source image and the corresponding reliable matching points.

[0019] In one embodiment, obtaining matching point pairs among the key points of the source image and the key points of the target image includes:

[0020] Obtaining a first descriptor for each key point of the source image;

[0021] Obtaining a second descriptor for each key point in the target image;

[0022] Performing the following steps on the first descriptor of each key point of the source image:

[0023] Obtaining the distance between the current first descriptor and each of the second descriptors;

[0024] Obtain the target second descriptor corresponding to the minimum distance;

[0025] The key point of the source image corresponding to the current first descriptor and the key point in the target image corresponding to the target second descriptor form a pair of the matching point pairs.

[0026] In one embodiment, the obtaining the first stable matching point pair possibility value for each pair of the matching point pairs includes:

[0027] Perform the following steps for each pair of the matching point pairs:

[0028] Obtain the key point distribution density within the first preset neighborhood corresponding to the current key point;

[0029] Obtain the key point distribution density within the second preset neighborhood corresponding to the initial nearest neighbor key point; the initial nearest neighbor key point and the current key point form a pair of the matching point pairs; the current key point is the key point in the source image, and the initial nearest neighbor key point is the key point in the target image;

[0030] Obtain the feature performance credibility of the current key point and the initial nearest neighbor key point according to the key point distribution density within the first preset neighborhood and the key point distribution density within the second preset neighborhood;

[0031] Obtain the nearest neighbor key point and the second nearest neighbor key point of the current key point in the target image;

[0032] Obtain the relative distances between the current key point and the initial nearest neighbor key point, between the current key point and the nearest neighbor key point, and between the current key point and the second nearest neighbor key point respectively;

[0033] Obtain the angle values corresponding to the initial nearest neighbor key point and the nearest neighbor key point in the target image respectively;

[0034] Obtain the first stable matching point pair possibility value of the current key point and the initial nearest neighbor key point according to the feature performance credibility, the relative distance, and the angle value.

[0035] In one embodiment,

[0036] The obtaining the first local network structure corresponding to each remaining key point of the source image includes:

[0037] Perform the following steps for each of the remaining key points of the source image:

[0038] In all the remaining key points of the source image, obtain a preset number of first remaining key points that are closest to the current remaining key point in the source image;

[0039] Connect the preset number of first remaining key points in sequence to obtain the first local network structure corresponding to the current remaining key points in the source image, where the current remaining key points in the source image are the source key points of the first local network structure, and the preset number of first remaining key points are the edge key points of the first local network structure;

[0040] The obtaining the second local network structure corresponding to each remaining key point of the target image includes:

[0041] Perform the following steps for each of the remaining key points of the target image:

[0042] Among all the remaining key points of the target image, obtain the preset number of second remaining key points that are closest to the current remaining key point in the target image;

[0043] Connect the preset number of second remaining key points in sequence to obtain the second local network structure corresponding to the current remaining key point in the target image, where the current remaining key point in the target image is the source key point of the second local network structure, and the preset number of second remaining key points are the edge key points of the second local network structure.

[0044] In one embodiment, the obtaining the reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure includes:

[0045] Perform the following steps for each of the first local network structures:

[0046] Calculate the initial similarity between the current first local network structure and each of the second local network structures;

[0047] Obtain the target second local network structure corresponding to the maximum initial similarity;

[0048] Obtain the final structural similarity between the current first local network structure and the target second local network structure;

[0049] According to the brightness value of the current first local network structure and the brightness value of the target second local network structure, obtain the ground vegetation color similarity between the current first local network structure and the target second local network structure;

[0050] Obtain the likelihood value of the second stable matching point pair between the source key points corresponding to the current first local network structure and the source key points in the target second local network structure according to the final structure similarity and the ground vegetation color similarity;

[0051] When the likelihood value of the second stable matching point pair is greater than the second preset threshold, determine the source key points in the target second local network structure as the reliable matching points of the source key points corresponding to the current first local network structure.

[0052] In one embodiment, the obtaining the final structure similarity between the current first local network structure and the target second local network structure includes:

[0053] Obtain the first distance between the source key points of the current first local network structure and each of the edge key points of the current first local network structure;

[0054] Obtain the second distance between the source key points of the target second local network structure and each of the edge key points of the target second local network structure;

[0055] Obtain the first angle between the line connecting the source key points of the current first local network structure and each of the edge key points of the current first local network structure and the horizontal line;

[0056] Obtain the second angle between the line connecting the source key points of the target second local network structure and each of the edge key points of the target second local network structure and the horizontal line;

[0057] Obtain the final structure similarity between the current first local network structure and the target second local network structure according to the first distance, the second distance, the first angle, and the second angle.

[0058] In one embodiment, the obtaining the ground vegetation color similarity between the current first local network structure and the target second local network structure according to the brightness value of the current first local network structure and the brightness value of the target second local network structure includes:

[0059] Obtain the first gray value of each of the edge key points of the current first local network structure;

[0060] Obtain the second gray value of each of the edge key points of the target second local network structure;

[0061] Obtain the ground vegetation color similarity according to all the first gray values and all the second gray values.

[0062] In a second aspect of the present disclosure, there is provided an automatic registration device for remote sensing image data, the device comprising:

[0063] A first acquisition module for acquiring remote sensing images of a to-be-acquired area at different times;

[0064] A second acquisition module for analyzing any two to-be-registered remote sensing images and acquiring key points in the two to-be-registered remote sensing images;

[0065] A third acquisition module for acquiring reliable matching point pairs in the two to-be-registered remote sensing images according to the local network structure corresponding to each key point in the two to-be-registered remote sensing images and the brightness change of the key point corresponding to the local network structure;

[0066] A matching module for matching all the reliable matching point pairs in the two to-be-registered remote sensing images to obtain a target remote sensing image after registration.

[0067] In a third aspect of the present disclosure, there is provided an automatic registration system for remote sensing image data, the system comprising a drone and the automatic registration device for remote sensing image data as described in the second aspect;

[0068] The drone is configured to capture remote sensing images of a to-be-acquired area at different times and send the remote sensing images of the to-be-acquired area at different times to the first acquisition module;

[0069] The first acquisition module is configured to acquire remote sensing images of the to-be-acquired area at different times;

[0070] The second acquisition module is configured to analyze any two to-be-registered remote sensing images and acquire key points in the two to-be-registered remote sensing images;

[0071] The third acquisition module is configured to acquire reliable matching point pairs in the two to-be-registered remote sensing images according to the local network structure corresponding to each key point in the two to-be-registered remote sensing images and the brightness change of the key point corresponding to the local network structure;

[0072] The matching module is configured to match all the reliable matching point pairs in the two to-be-registered remote sensing images to obtain a target remote sensing image after registration.

[0073] The present invention has the following beneficial effects:

[0074] In the present disclosure, remote sensing images at different times of the area to be collected are obtained; any two remotely sensed images to be registered are analyzed to obtain key points in the two remotely sensed images to be registered; according to the local network structure corresponding to each key point in the two remotely sensed images to be registered and the brightness change of the key point corresponding to the local network structure, reliable matching point pairs in the two remotely sensed images to be registered are obtained; all reliable matching point pairs in the two remotely sensed images to be registered are matched to obtain the target remotely sensed image after registration. By obtaining reliable matching point pairs in the two remotely sensed images to be registered according to the local network structure corresponding to each key point in the two remotely sensed images to be registered and the brightness change of the key point corresponding to the local network structure, the problem of false matching caused by self-similar structures can be effectively reduced, and the accuracy of remote sensing image registration can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0076] Figure 1 Flow chart of an automatic registration method for remote sensing image data provided by an embodiment of the present invention Figure 1 ;

[0077] Figure 2 An embodiment of the present invention provides a remote sensing image of an area to be collected at a certain time by means of aerial photography;

[0078] Figure 3 Schematic diagram of key points in the source image provided by an embodiment of the present invention;

[0079] Figure 4 Schematic diagram of key points in the target image provided by an embodiment of the present invention;

[0080] Figure 5 Schematic diagram of the target remote sensing image after registration provided by an embodiment of the present invention;

[0081] Figure 6 Flow chart of an automatic registration method for remote sensing image data provided by an embodiment of the present invention Figure 2 ;

[0082] Figure 7 Structural block diagram of an automatic registration device for remote sensing image data provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following specifically describes, in conjunction with the accompanying drawings and preferred embodiments, a method, device, and system for automatic registration of remote sensing image data according to the present invention, including its specific implementation manners, structures, features, and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0084] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0085] The following specifically describes the specific solutions of a method, device, and system for automatic registration of remote sensing image data provided by the present invention in conjunction with the accompanying drawings.

[0086] Please refer to Figure 1 , which shows a flowchart of a method for automatic registration of remote sensing image data provided by an embodiment of the present invention. As Figure 1 shown, the method includes the following steps S101 - S104:

[0087] S101. Obtain remote sensing images of the area to be collected at different times.

[0088] Multiple remote sensing images of the area to be collected can be obtained by means of aerial photography.

[0089] Determine the area and scope of data acquisition, and use a drone equipped with an optical camera or a multispectral sensor to collect the area to be collected, and then obtain a remote sensing image of the area to be collected at a certain time through digital scanning. By scanning the area to be collected multiple times at different times, multiple remote sensing images of the area to be collected are obtained.

[0090] Among them, multiple remote sensing images are taken from approximately the same viewpoint during the flight of the drone, and it is recommended to choose a time with good weather conditions for shooting.

[0091] If the area and scope of data acquisition are large, the method of fixed-altitude cruise flight can be adopted to collect multiple times to obtain all remote sensing images of the area to be collected. Figure 2 This is a remote sensing image of the area to be collected at a certain time obtained by aerial photography according to an embodiment of the present invention. As Figure 2 shown, the remote sensing image of the area to be collected mainly includes a large area of farmland and a river.

[0092] S102. Analyze any two remotely sensed images to be registered, and obtain key points in the two remotely sensed images to be registered.

[0093] Perform key point detection on the remote sensing image to obtain all possible key points in the remote sensing image.

[0094] The key point detection method in the SIFT algorithm can be used to perform key point detection on each remote sensing image to obtain all possible key points in the remote sensing image.

[0095] S103. Obtain reliable matching point pairs in two remotely sensed images to be registered according to the local network structure corresponding to each key point in the two remotely sensed images to be registered and the brightness change of the key point corresponding to the local network structure.

[0096] In this step, the two remotely sensed images to be registered will be analyzed to determine all reliable matching point pairs with relatively high reliability in the two remotely sensed images to be registered.

[0097] In the following embodiments, in order to clearly illustrate the solution in the present application, the source image and the target image are used as the two remotely sensed images to be registered for description. That is, all reliable matching point pairs with relatively high reliability in the source image and the target image are obtained in the following embodiments.

[0098] In one embodiment, as Figure 6 shown, step S103 includes the following sub-steps S1031-S1037:

[0099] S1031. Obtain the matching point pairs among the key points of the source image and the key points of the target image.

[0100] Remote sensing image registration is the process of matching and superimposing two or more images to align the pixel accuracy between the images. In the process of image registration, image matching is the most core step. Therefore, the main purpose of this step is to obtain as many possible key points and all possible matching point pairs in the two remotely sensed images to be registered, so as to screen stable and reliable matching point pairs in the two remotely sensed images to be registered subsequently, and then complete the automatic registration of the two remotely sensed images to be registered.

[0101] All possible key points in the two remotely sensed images to be registered can be obtained by performing key point detection on the two remotely sensed images to be registered, and then all possible matching point pairs between the source image and the target image can be obtained by matching the descriptors of the key points in any two remotely sensed images to be registered. Specifically, step S1031 includes the following sub-steps S10311-S10315:

[0102] S10311. Obtain the first descriptor of each key point of the source image.

[0103] The first descriptor of the key points in each source image can be obtained through the SIFT algorithm, where the first descriptor represents the descriptor in the source image.

[0104] S10312. Obtain the second descriptor of the key points in each target image.

[0105] The second descriptor of the key points in the target image can be obtained through the SIFT algorithm, where the second descriptor represents the descriptor in the target image.

[0106] Perform the following steps S10312 - S10314 on the first descriptor of the key points in each source image:

[0107] S10313. Obtain the distances between the current first descriptor and each of the second descriptors.

[0108] Among them, the Euclidean distance can be used as the distance.

[0109] S10314. Obtain the target second descriptor corresponding to the minimum distance.

[0110] S10315. The key point of the source image corresponding to the current first descriptor and the key point of the target image corresponding to the target second descriptor form a pair of matching point pairs.

[0111] Take the two key points with the closest distance between the descriptors in the source image and the target image as a pair of possible matching point pairs between the source image and the target image.

[0112] Figure 3 is a schematic diagram of the key points in the source image. Figure 4 is a schematic diagram of the key points in the target image. The key points detected are mainly the edge points and corner points of the farmland in the remote sensing image, as well as the bright points in the dark areas or dark points in the bright areas of the river. However, not all the detected key points are stable and reliable key points. This step is mainly to obtain as many possible key points and possible matching point pairs in the remote sensing image as possible. The subsequent steps will further screen and determine the stable and reliable matching point pairs.

[0113] So far, through the key point detection process, all possible key points in the source image and the target image, as well as all possible matching point pairs between the source image and the target image, are obtained.

[0114] S1032. Obtain the first stable matching point pair possibility value for each pair of matching point pairs.

[0115] The main purpose of this step is to initially determine some matching point pairs with obvious feature manifestations, stable and reliable in the source image and the target image, that is, to determine stable and reliable key points. The stable and reliable key points in the source image and the target image have obvious feature differences in the local area. The self-similar structure means that there are multiple local areas with similar patterns in the source image and the target image. For example, there are large areas of farmland in the source image and the target image, and a certain feature in the large area of farmland appears as a similar pattern in the source image and the target image multiple times. The self-similar structure often distributes in areas with more complex key points. If the distance difference between the nearest neighbor key point corresponding to a key point in the source image and the second-nearest neighbor point in the target image is larger, it indicates that this key point in the source image and the nearest key point in the target image are very likely to be a correct match, with obvious feature differences and the least influence from the self-similar structure, and both are stable key points; if the distance difference is smaller, it means that the distances between this key point in the source image and the two key points in the target image are relatively close, and the matching point may be a false match.

[0116] In the following embodiments, the initial nearest neighbor key point and the current key point form a pair of matching point pairs, where the current key point is the key point in the source image, and the initial nearest neighbor key point is the key point in the target image. In one embodiment, step S1032 includes performing the following steps S10321 - S10327 for each pair of matching point pairs:

[0117] S10321. Obtain the key point distribution density within the first preset neighborhood corresponding to the current key point.

[0118] For the current key point in the source image, in the source image, a square range of 11 * 11 centered on this current key point is used as the range of the first preset neighborhood, and the key point distribution density within the first preset neighborhood range is obtained. In the present disclosure, the number of key points within the first preset neighborhood range can be used to represent the key point distribution density within the first preset neighborhood range.

[0119] S10322. Obtain the key point distribution density within the second preset neighborhood corresponding to the initial nearest neighbor key point.

[0120] Among them, the ranges corresponding to the first preset neighborhood and the second preset neighborhood are the same.

[0121] Take the other key point of the matching point pair corresponding to the current key point as the initial nearest neighbor key point of the current key point. For the initial nearest neighbor key point in the target image, in the target image, a square range of 11 * 11 centered on the initial nearest neighbor key point is used as the second preset neighborhood range of the initial nearest neighbor key point, and the key point distribution density within the second preset neighborhood range is obtained. In the present disclosure, the number of key points within the second preset neighborhood range can be used to represent the key point distribution density within the second preset neighborhood range.

[0122] S10323. Obtain the feature performance credibility of the current key point and the initial nearest neighbor key point according to the key point distribution density within the first preset neighborhood and the key point distribution density within the second preset neighborhood.

[0123] The feature performance credibility of the current key point and the initial nearest neighbor key point can be obtained through the following formula:

[0124] ;

[0125] In the formula, is the key point distribution density of the th key point in the source image within its first preset neighborhood; is the key point distribution density of the initial nearest neighbor key point corresponding to the th key point in the source image within its second preset neighborhood range in the target image; is the exponential function with the natural constant as the base, used to achieve the inverse proportional relationship; is the feature performance credibility of the th key point in the source image and the initial nearest neighbor key point corresponding to it in the target image. Among them, the th key point in the source image represents the current key point of the source image in the above embodiments.

[0126] In the above formula, the smaller the key point distribution density of the th key point in the source image within its first preset neighborhood range, and at the same time, the closer the key point distribution density of the initial nearest neighbor key point corresponding to the th key point in the source image within its second preset neighborhood range in the target image is to the key point distribution density of the th key point in the source image within its first preset neighborhood range, it indicates that the sparse feature performance of the th key point in the source image and the initial nearest neighbor key point corresponding to it in the target image is better, the influence of the self-similar structure is the smallest, and the feature performance credibility is higher.

[0127] Through the preliminary analysis of the above steps, the reliability of the feature representation of the current key point in the source image and the corresponding initial nearest neighbor key point in the target image is obtained. Considering that the remote sensing images are taken from approximately the same viewpoint during the flight of the unmanned aerial vehicle, there will be a certain offset between the images. Therefore, it is also necessary to combine the changes in the nearest neighbor distance and the next-nearest neighbor distance to obtain stable and reliable reliable matching point pairs.

[0128] S10324. Obtain the nearest neighbor key point and the next-nearest neighbor key point of the current key point in the target image.

[0129] In this step, since the current key point in the source image and the initial nearest neighbor key point in the target image are a pair of matching point pairs, therefore, the nearest neighbor key point and the next-nearest neighbor key point corresponding to the initial nearest neighbor key point can be obtained in the target image. The nearest neighbor key point and the next-nearest neighbor key point corresponding to the initial nearest neighbor key point obtained here represent the nearest neighbor key point and the next-nearest neighbor key point of the current key point in the target image in the above embodiments.

[0130] S10325. Respectively obtain the relative distances between the current key point and the initial nearest neighbor key point, the nearest neighbor key point, and the next-nearest neighbor key point.

[0131] S10326. Respectively obtain the angular values corresponding to the initial nearest neighbor key point and the nearest neighbor key point in the target image.

[0132] S10327. Obtain the first stable matching point pair possibility value between the current key point and the initial nearest neighbor key point according to the reliability of the feature representation, the relative distance, and the angular value.

[0133] The first stable matching point pair possibility value is obtained through the following formula:

[0134] ;

[0135] In the formula, is the reliability of the feature representation of the th key point in the source image and the corresponding initial nearest neighbor key point in the target image; is the relative distance between the th key point in the source image and the corresponding initial nearest neighbor key point in the target image. The relative distance is specifically the Euclidean distance between the coordinates of the th key point and the coordinates of the corresponding initial nearest neighbor key point; is the relative distance between the th key point in the source image and the corresponding nearest neighbor key point in the target image; is the relative distance parameter, is the angular parameter. In the present disclosure, can take 0.6, It can be 0.4. The smaller the relative distance, the more it indicates a correct match. Therefore, the relative distance parameter can be 0.6. There will be a certain offset between the images themselves. Although it is better if the offset is smaller, the correct match is mainly judged by the relative distance, and its importance is slightly lower. Therefore, the angle parameter can be 0.4; is a hyperparameter to prevent the denominator from being 0. In the present disclosure, can be 0.1; is the angle value corresponding to the initial nearest neighbor key point corresponding to the th key point in the source image in the target image. The angle value corresponding to the initial nearest neighbor key point in the target image is the included angle value between the line segment formed by the initial nearest neighbor key point and the lower left corner of the target image and the horizontal line; is the angle value corresponding to the nearest neighbor key point corresponding to the th key point in the source image in the target image; The angle value corresponding to the nearest neighbor key point in the target image is the included angle value between the line segment formed by the nearest neighbor key point and the lower left corner of the target image and the horizontal line; is the relative distance between the th key point in the source image and the corresponding second nearest neighbor key point in the target image; is the first stable matching point pair possibility value between the th key point in the source image and the corresponding initial nearest neighbor key point in the target image. Through the first stable matching point pair possibility value, it can be judged whether the th key point in the source image and the corresponding initial nearest neighbor key point in the target image are a pair of stable and reliable matching point pairs.

[0136] When the feature expression credibility is greater and the influence of the self-similar structure is minimized, the relative distance difference between the th key point in the source image and the corresponding initial nearest neighbor key point and nearest neighbor key point in the target image is smaller. At the same time, the angle difference between the th key point in the source image and the corresponding initial nearest neighbor key point and nearest neighbor key point in the target image is smaller, indicating that the th key point in the source image and the corresponding initial nearest neighbor key point in the target image are more likely to be correctly corresponding; If the relative distance between the th key point in the source image and the corresponding initial nearest neighbor key point in the target image and the relative distance difference between the th key point in the source image and the corresponding second nearest neighbor key point in the target image are larger, it indicates that the th key point in the source image and the corresponding initial nearest neighbor key point in the target image are more likely to be a correct match, there are obvious feature differences, the influence of the self-similar structure is minimized, and both are stable key points.

[0137] S1033. Obtain a first target matching point pair whose possibility value of the first stable matching point pair is greater than a first preset threshold.

[0138] The first preset threshold can be set to 0.8. The matching point pairs whose possibility values of the first stable matching point pairs obtained in the above steps are greater than the set first preset threshold are used as stable and reliable matching point pairs, and vice versa as unstable and unreliable matching point pairs. Among them, the stable and reliable matching point pairs represent the first target matching points.

[0139] In one implementable manner, before comparison, the possibility values of the first stable matching point pairs obtained in the above steps can also be linearly normalized to obtain the normalized possibility. The matching point pairs with the normalized possibility greater than the set first preset threshold are used as stable and reliable matching point pairs, and vice versa as unstable and unreliable matching point pairs.

[0140] So far, by analyzing the changes in the nearest neighbor distance and the second nearest neighbor distance, some relatively reliable partial matching point pairs in the source image and the target image are initially obtained.

[0141] Next, analyze the remaining key points, obtain the local network structure of the polygon formed by each remaining key point in the source image and the target image and its surrounding similar remaining key points, analyze the similarity of the chain codes of the local network structures of the polygons formed in the source image and the target image, as well as the brightness change and density distribution change of the remaining key points in the local network structure of the polygon, and determine more reliable matching point pairs in the source image and the target image.

[0142] For the automatic registration of remote sensing images, it is hoped to determine as many, relatively accurate and evenly distributed matching point pairs as possible. Some stable and reliable small numbers of matching point pairs in the source image and the target image are initially obtained through the above analysis. Self-similar structures in the source image and the target image often exist in areas with complex local key points. In order to achieve better automatic registration of the source image and the target image, it is also necessary to verify the remaining possible stable key points to expand more stable matching point pairs.

[0143] Self-similar structures in the source image and the target image often have similar patterns in multiple local areas, making it difficult to judge the correct corresponding relationship when selecting matching point pairs and resulting in false matches. Therefore, if the matching point pairs are determined only by analyzing the performance of the local areas where the remaining key points are located, there will be a large deviation.

[0144] Stable remaining key points often form an approximately fixed local network structure. Therefore, the similarity of the local network structure can be analyzed to reduce the influence of self-similar structures in the image. At the same time, if the color performance of the remaining key point regions corresponding to the local network structure on the remote sensing image is similar, it indicates that the corresponding actual ground performance is similar within a certain neighborhood range, which can also reduce the influence of self-similar structures. By analyzing the density distribution, if the distribution is relatively uniform, it indicates that the two remaining key points are a pair of correct matching point pairs, so as to reduce the interference of self-similarity on the matching point pairs in the image and expand more stable matching point pairs to achieve automatic registration.

[0145] The key points in the unstable and reliable matching point pairs obtained above are used as the remaining key points in the source image and the target image. That is, the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pair in the source image, and the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pair in the target image. Next, further analysis is performed on these remaining key points.

[0146] S1034. Obtain the first local network structure corresponding to each remaining key point of the source image.

[0147] In one embodiment, step S1034 includes performing the following steps S10341 - S10342 on each remaining key point in the source image:

[0148] S10341. Among all the remaining key points in the source image, obtain a preset number of first remaining key points that are closest to the current remaining key point in the source image.

[0149] S10342. Connect the preset number of first remaining key points in sequence to obtain the first local network structure corresponding to the current remaining key point in the source image, where the current remaining key point in the source image is the source key point of the first local network structure, and the preset number of first remaining key points are the edge key points of the first local network structure.

[0150] Exemplarily, the preset number is 6. For any remaining key point in any source image, obtain 6 (which can also be set to other values) first remaining key points that are closest to this remaining key point, and connect these 6 first remaining key points that are closest in sequence to obtain the first local network structure of this remaining key point.

[0151] S1035. Obtain the second local network structure corresponding to each remaining key point of the target image;

[0152] In one embodiment, step S1035 includes performing the following steps S10351 - S10352 on each remaining key point in the target image:

[0153] S10351. Among all the remaining key points in the target image, obtain a preset number of second remaining key points that are closest to the current remaining key point of the target image.

[0154] S10352. Connect the preset number of second remaining key points in sequence to obtain a second local network structure corresponding to the current remaining key point of the target image, where the current remaining key point of the target image is the source key point of the second local network structure, and the preset number of second remaining key points are the edge key points of the second local network structure.

[0155] Exemplarily, the preset number is 6. For any remaining key point in any target image, obtain 6 (which can also be set to other values) second remaining key points that are closest to this remaining key point, and connect these 6 closest second remaining key points in sequence to obtain the second local network structure of this remaining key point.

[0156] S1036. According to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure, obtain the reliable matching points of each remaining key point in the source image among the remaining key points in the target image.

[0157] In one embodiment, step S1036 includes performing the following steps S10361 - S10366 on each first local network structure:

[0158] S10361. Calculate the initial similarity between the current first local network structure and each second local network structure.

[0159] Obtain the chain code of the current first local network structure, where the starting point of the chain code of the current first local network structure starts from an edge key point closest to the source key point of this current first local network structure and is obtained by cycling counterclockwise.

[0160] Obtain the chain code of the second local network structure, where the starting point of the chain code of the second local network structure starts from an edge key point closest to the source key point of this second local network structure and is obtained by cycling counterclockwise.

[0161] Obtain the initial similarity between the first local network structure and the second local network structure through the following formula:

[0162] ;

[0163] In the formula, is the number of remaining key points in the target image; is the The chain code sequence vector of the first local network structure of the remaining key points. The specific method for obtaining the chain code sequence vector is as follows: Arrange the chain codes of the first local network structure of the remaining key points from the starting point to obtain the corresponding chain code sequence, and use this chain code sequence as a row vector, that is, the chain code sequence vector; is the chain code sequence vector of the second local network structure of the remaining key points in the target image; is the cosine function, represents the cosine similarity of two chain code sequence vectors; is the sigmoid function for normalization processing; is the initial similarity between the first local network structure of the remaining key points in the source image and the second local network structure of the remaining key points in the target image.

[0164] Among them, the first local network structure of the remaining key points in the source image represents the current first local network structure; the remaining key points in the source image represent the source key points of the current first local network structure; the remaining key points in the target image represent the source key points of the corresponding second local network structure.

[0165] S10362. Obtain the target second local network structure corresponding to the maximum initial similarity.

[0166] For each first local network structure, find the target second local network structure corresponding to the maximum initial similarity. Among them, the first local network structure and the target second local network structure corresponding to the maximum initial similarity are initially similar.

[0167] S10363. Obtain the final structural similarity between the current first local network structure and the target second local network structure.

[0168] If the chain code sequence vector of the first local network structure of the remaining key points in the source image is very similar to the chain code sequence vector of the second local network structure of a remaining key point in the target image, it indicates that the source key points corresponding to the two local network structures are more likely to be a pair of correct matches, and the initial similarity of the two local network structures is better.

[0169] Considering that there may be multiple local network structures that are relatively similar, resulting in errors in the similarity judgment of local network structures, the distribution relationship of key points on the local network structure is analyzed here, and combined with the initial similarity, the final structural similarity of the local network structures of key points in the source image and the target image is determined.

[0170] In one embodiment, step S10363 includes the following sub-steps S103631 - S103635:

[0171] S103631. Obtain the first distance between the source key point of the current first local network structure and each edge key point of the current first local network structure.

[0172] S103632. Obtain the second distance between the source key point of the target second local network structure and each edge key point of the target second local network structure.

[0173] S103633. Obtain the first included angle between the line connecting the source key point of the current first local network structure and each edge key point of the current first local network structure and the horizontal line.

[0174] S103634. Obtain the second included angle between the line connecting the source key point of the target second local network structure and each edge key point of the target second local network structure and the horizontal line.

[0175] S103635. Obtain the final structural similarity between the current first local network structure and the target second local network structure according to the first distance, the second distance, the first included angle, and the second included angle.

[0176] The final structural similarity between the current first local network structure and the target second local network structure can be obtained through the following formula:

[0177] ;

[0178] In the formula, is the number of edge key points on any local network structure, which can be the number of edge key points on the current first local network structure or the number of edge key points on the target second local network structure. is the first distance between the th remaining key point on the first local network structure of the source image and the th remaining key point on the first local network structure of the source image; is the second distance between the th remaining key point on the target second local network structure, which has the highest initial similarity with the first local network structure of the th remaining key point in the source image, and the source key point of the target second local network structure; is the first included angle value between the line connecting the th remaining key point and the th remaining key point on the first local network structure of the source image and the horizontal line; is the first included angle value between the line connecting the th remaining key point and the is the second included angle value between the horizontal line and the line connecting the th remaining key point on the target second local network structure with the highest initial similarity to the first local network structure of the th remaining key point in the source image; is the initial similarity between the first local network structure of the th remaining key point in the source image and the second local network structure of the th remaining key point in the target image. Here, it is assumed that the second local network structure of the th remaining key point is initially similar to the first local network structure of the th remaining key point, that is, the second local network structure that is initially similar to the first local network structure of the th remaining key point in the source image; is the exponential function with the natural constant as the base, used to achieve the inverse proportional relationship and normalization; is the final structural similarity between the first local network structure of the th remaining key point in the source image and the second local network structure of the th remaining key point in the target image.

[0179] Among them, the first local network structure of the th remaining key point in the source image represents the current first local network structure, the th remaining key point in the source image represents the source key point in the current first local network structure, the th remaining key point in the first local network structure of the th remaining key point in the source image represents the edge key point in the current first local network structure; the th remaining key point in the second local network structure represents the target second local network structure corresponding to the current first local network structure; the th remaining key point represents the source key point in the target second local network structure; the th remaining key point on the target second local network structure represents the edge key point in the target second local network structure.

[0180] When the corresponding local network structures in the source image and the target image are initially similar, if the differences in the distance distribution and angle distribution of the local network structures are both small, it indicates that these two local network structures are truly similar, and the two remaining key points corresponding to the local network structures are more likely to be a pair of stable and reliable matching point pairs.

[0181] Next, analyze the brightness change of the remaining key points in the two initially similar local network structures.

[0182] S10364. Obtain the ground vegetation color similarity between the current first local network structure and the target second local network structure according to the brightness value of the current first local network structure and the brightness value of the target second local network structure.

[0183] In one embodiment, step S10364 includes the following sub-steps S103641 - S103643:

[0184] S103641. Obtain the first gray value of each edge key point of the current first local network structure;

[0185] S103642. Obtain the second gray value of each edge key point of the target second local network structure;

[0186] S103643. Obtain the ground vegetation color similarity according to all the first gray values and all the second gray values.

[0187] Specifically, obtain the ground vegetation color similarity according to the following formula:

[0188] ;

[0189] In the formula, is the number of edge key points on any local network structure, which can be the number of edge key points on the current first local network structure or the number of edge key points on the target second local network structure; is the gray value of the th remaining key point on the first local network structure of the th remaining key point in the source image; is the gray value of the th remaining key point on the target second local network structure that is initially similar to the first local network structure of the th remaining key point in the target image corresponding to the th remaining key point in the source image; is the ground vegetation color similarity between the first local network structure of the th remaining key point in the source image and the target second local network structure of the th remaining key point in the target image. Here, the ground vegetation color similarity characterizes the consistency of the ground vegetation color performance in the actual scene. It is assumed that the target second local network structure of the th remaining key point is initially similar to the first local network structure of the

[0190] If the corresponding remaining key points on the two initially similar local network structures have the same color representation on the actual ground, that is, the gray values of the corresponding remaining key points are similar, it further indicates that these two local network structures are truly similar, and the more likely the two remaining key points corresponding to the local network structures are a pair of stable and reliable matching point pairs.

[0191] S10365. Obtain the possible value of the second stable matching point pair between the source key points corresponding to the current first local network structure and the source key points in the target second local network structure according to the final structure similarity and the ground vegetation color similarity.

[0192] Combining the final structure similarity and the ground vegetation color similarity, obtain the possible value of the second stable matching point pair through the following formula:

[0193] ;

[0194] In the formula, is the final structure similarity between the first local network structure of the th remaining key point in the source image and the target second local network structure of the th remaining key point in the target image; is the ground vegetation color similarity between the first local network structure of the th remaining key point in the source image and the second local network structure of the th remaining key point in the target image; is the linear normalization function; is the th remaining key point in the source image and the th remaining key point in the target image, and the possible value of the second stable matching point pair.

[0195] S10366. When the possible value of the second stable matching point pair is greater than the second preset threshold, determine the source key point in the target second local network structure as the reliable matching point of the source key point corresponding to the current first local network structure.

[0196] Take those greater than the set second preset threshold as stable and reliable matching point pairs, and vice versa as unstable and reliable matching point pairs.

[0197] So far, through the similarity of the network structure chain codes of the remaining key points and the analysis of the brightness changes of the remaining key points in the local network structure of the polygon, more stable and reliable matching point pairs are obtained.

[0198] S1037. Obtain the reliable matching point pairs according to the first target matching point pairs and the second target matching point pairs, and the second target matching point pairs are composed of the remaining key points in the source image and the corresponding reliable matching points.

[0199] S104. Match all reliable matching point pairs in two remotely sensed images to be registered to obtain the target remotely sensed image after registration.

[0200] Match all reliable matching point pairs in the two remotely sensed images to be registered to obtain the target remotely sensed image after registration. That is, as Figure 5 shown, use the SIFT algorithm to match all stable and reliable matching point pairs in the source image and the target image to obtain the target remotely sensed image after registration.

[0201] In this application, by analyzing the changes in the nearest neighbor distance and the second nearest neighbor distance of the corresponding key points in the source image and the target image, and considering the distances between two pairs of corresponding key points in the two images, initially obtain some relatively reliable partial matching point pairs in the remotely sensed image, and analyze the remaining key points to obtain the polygon network structure formed by each key point in the remotely sensed image and its surrounding similar key points. Analyze the similarity of the chain codes of the polygon network structures formed in the two remotely sensed images to be registered, as well as the brightness changes of the key points in the polygon network structure, and expand more reliable matching point pairs, so that the automatic registration can be well achieved even when the remotely sensed image has a slight offset and a self-similar structure image. Through this method, the problem of incorrect matching caused by the self-similar structure can be effectively reduced, and the accuracy of remotely sensed image registration can be improved.

[0202] Figure 7 The structural block diagram of the automatic registration device for remotely sensed image data provided by an embodiment of the present invention is shown as Figure 7 shown, and the device includes:

[0203] The first acquisition module 11 is used to acquire remotely sensed images at different times of the area to be acquired;

[0204] The second acquisition module 12 is used to analyze any two remotely sensed images to be registered and acquire the key points in the two remotely sensed images to be registered;

[0205] The third acquisition module 13 is used to acquire the reliable matching point pairs in the two remotely sensed images to be registered according to the local network structure corresponding to each key point in the two remotely sensed images to be registered and the brightness change of the key point corresponding to the local network structure;

[0206] The matching module 14 is used to match all the reliable matching point pairs in the two remotely sensed images to be registered to obtain the target remotely sensed image after registration.

[0207] In one embodiment, the third acquisition module is specifically used for:

[0208] Obtain matching point pairs among the key points of the source image and the key points of the target image; the source image and the target image are the two remotely sensed images to be registered.

[0209] Obtain the first stable matching point pair possibility value for each pair of the matching point pairs.

[0210] Obtain the first target matching point pairs for which the first stable matching point pair possibility value is greater than the first preset threshold.

[0211] Obtain the first local network structure corresponding to each remaining key point of the source image; the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pairs in the source image.

[0212] Obtain the second local network structure corresponding to each remaining key point of the target image; the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pairs in the target image.

[0213] According to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure, obtain the reliable matching points of each remaining key point in the source image among the remaining key points in the target image.

[0214] Obtain the reliable matching point pairs according to the first target matching point pairs and the second target matching point pairs, where the second target matching point pairs are composed of the remaining key points in the source image and the corresponding reliable matching points.

[0215] In one embodiment, in terms of obtaining the matching point pairs among the key points of the source image and the key points of the target image, the third obtaining module is specifically configured to:

[0216] Obtain the first descriptor of each key point of the source image.

[0217] Obtain the second descriptor of each key point in the target image.

[0218] Perform the following steps on the first descriptor of each key point of the source image:

[0219] Obtain the distance between the current first descriptor and each of the second descriptors.

[0220] Obtain the target second descriptor corresponding to the minimum distance.

[0221] The key point of the source image corresponding to the current first descriptor and the key point in the target image corresponding to the target second descriptor are a pair of the matching point pairs.

[0222] In one embodiment, in terms of obtaining the first stable matching point pair possibility value for each pair of the matching point pairs, the third obtaining module is specifically configured to:

[0223] Perform the following steps for each pair of the matching point pairs:

[0224] Obtain the key point distribution density within the first preset neighborhood corresponding to the current key point;

[0225] Obtain the key point distribution density within the second preset neighborhood corresponding to the initial nearest neighbor key point; the initial nearest neighbor key point and the current key point are a pair of the matching point pairs; the current key point is a key point in the source image, and the initial nearest neighbor key point is a key point in the target image;

[0226] Obtain the feature expression credibility of the current key point and the initial nearest neighbor key point according to the key point distribution density within the first preset neighborhood and the key point distribution density within the second preset neighborhood;

[0227] Obtain the nearest neighbor key point and the second nearest neighbor key point of the current key point in the target image;

[0228] Respectively obtain the relative distances between the current key point and the initial nearest neighbor key point, between the current key point and the nearest neighbor key point, and between the current key point and the second nearest neighbor key point;

[0229] Respectively obtain the angular values corresponding to the initial nearest neighbor key point and the nearest neighbor key point in the target image;

[0230] Obtain the first stable matching point pair possibility value of the current key point and the initial nearest neighbor key point according to the feature expression credibility, the relative distance, and the angular value.

[0231] In one embodiment, in terms of obtaining the first local network structure corresponding to each remaining key point of the source image, the third obtaining module is specifically configured to:

[0232] Perform the following steps for each remaining key point of the source image:

[0233] Among all the remaining key points of the source image, obtain a preset number of first remaining key points that are closest to the current remaining key point in the source image;

[0234] Connect the preset number of first remaining key points in sequence to obtain the first local network structure corresponding to the current remaining key points in the source image, where the current remaining key points in the source image are the source key points of the first local network structure, and the preset number of first remaining key points are the edge key points of the first local network structure.

[0235] In one embodiment, in terms of obtaining the second local network structure corresponding to each remaining key point of the target image, the third acquisition module is specifically configured to:

[0236] Execute the following steps for each of the remaining key points of the target image:

[0237] Among all the remaining key points of the target image, obtain the preset number of second remaining key points that are closest to the current remaining key point in the target image;

[0238] Connect the preset number of second remaining key points in sequence to obtain the second local network structure corresponding to the current remaining key point in the target image, where the current remaining key point in the target image is the source key point of the second local network structure, and the preset number of second remaining key points are the edge key points of the second local network structure.

[0239] In one embodiment, in terms of obtaining the reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure, the third acquisition module is specifically configured to:

[0240] Execute the following steps for each of the first local network structures:

[0241] Calculate the initial similarity between the current first local network structure and each of the second local network structures;

[0242] Obtain the target second local network structure corresponding to the maximum initial similarity;

[0243] Obtain the final structural similarity between the current first local network structure and the target second local network structure;

[0244] According to the brightness value of the current first local network structure and the brightness value of the target second local network structure, obtain the ground vegetation color similarity between the current first local network structure and the target second local network structure;

[0245] Obtain the possibility value of the second stable matching point pair between the source key points corresponding to the current first local network structure and the source key points in the target second local network structure according to the final structure similarity and the ground vegetation color similarity;

[0246] When the possibility value of the second stable matching point pair is greater than the second preset threshold, determine that the source key points in the target second local network structure are reliable matching points of the source key points corresponding to the current first local network structure.

[0247] In one embodiment, in terms of obtaining the final structure similarity between the current first local network structure and the target second local network structure, the third obtaining module is specifically configured to:

[0248] Obtain the first distance between the source key points of the current first local network structure and each of the edge key points of the current first local network structure;

[0249] Obtain the second distance between the source key points of the target second local network structure and each of the edge key points of the target second local network structure;

[0250] Obtain the first angle between the line connecting the source key points of the current first local network structure and each of the edge key points of the current first local network structure and the horizontal line;

[0251] Obtain the second angle between the line connecting the source key points of the target second local network structure and each of the edge key points of the target second local network structure and the horizontal line;

[0252] Obtain the final structure similarity between the current first local network structure and the target second local network structure according to the first distance, the second distance, the first angle, and the second angle.

[0253] In one embodiment, in terms of obtaining the ground vegetation color similarity between the current first local network structure and the target second local network structure according to the brightness value of the current first local network structure and the brightness value of the target second local network structure, the third obtaining module is specifically configured to:

[0254] Obtain the first gray value of each of the edge key points of the current first local network structure;

[0255] Obtain the second gray value of each of the edge key points of the target second local network structure;

[0256] Obtain the ground vegetation color similarity according to all the first gray values and all the second gray values.

[0257] The present disclosure also provides an automatic registration system for remote sensing image data. The system includes a drone and an automatic registration device for remote sensing image data as described in the above embodiments.

[0258] It should be noted that the above order of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0259] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for automatic registration of remote sensing image data, characterized in that: The method comprises: Obtain remote sensing images of the area to be collected at different times; Analyze any two remote sensing images to be registered to obtain key points in the two remote sensing images to be registered; Obtain matching point pairs among key points of the source image and key points of the target image, where the source image and the target image are two remote sensing images to be registered; Obtain a first target matching point pair, the first target matching point pair being a matching point pair in which the likelihood value of the first stable matching point pair in each pair of matching point pairs is greater than a first preset threshold; wherein the likelihood value of the first stable matching point pair between the current key point and the initial nearest neighbor key point is obtained based on the feature expression credibility, relative distance, and the angle value corresponding to the initial nearest neighbor key point and the nearest neighbor key point of the current key point in the target image in the target image, the current key point is a key point in the source image, the initial nearest neighbor key point is a key point in the target image, the initial nearest neighbor key point and the current key point are a pair of matching point pairs, the relative distance includes the relative distance between the current key point and the initial nearest neighbor key point, the nearest neighbor key point, and the next nearest neighbor key point of the current key point in the target image, the feature expression credibility is obtained based on the key point distribution density in the first preset neighborhood corresponding to the current key point and the key point distribution density in the second preset neighborhood corresponding to the initial nearest neighbor key point; Obtain a first local network structure corresponding to each remaining key point of the source image and a second local network structure corresponding to each remaining key point of the target image, wherein the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pair in the source image, and the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pair in the target image; Obtaining reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure; Acquire reliable matching point pairs in the two remote sensing images to be registered according to the first target matching point pair and the second target matching point pair, wherein the second target matching point pair is composed of the remaining key points in the source image and the corresponding reliable matching points; All the reliable matching point pairs in the two remote sensing images to be registered are matched to obtain the registered target remote sensing image.

2. The automatic registration method for remote sensing image data according to claim 1, characterized in that: The step of obtaining matching point pairs between key points of the source image and key points of the target image includes: Obtaining a first descriptor of each key point of the source image; Obtaining a second descriptor of each key point in the target image; For each of the first descriptors of the key points of the source image, perform the following steps: Obtaining the distance between the current first descriptor and each of the second descriptors; Get the second descriptor of the target corresponding to the minimum distance; The key point of the source image corresponding to the current first descriptor and the key point in the target image corresponding to the target second descriptor form a pair of matching points.

3. The automatic registration method for remote sensing image data according to claim 1, characterized in that: The obtaining of a first local network structure corresponding to each remaining key point of the source image includes: Perform the following steps for each of the remaining key points of the source image: Acquire a preset number of first remaining key points that are closest to the current remaining key point in the source image from among all the remaining key points in the source image; Connecting the preset number of first remaining key points in sequence to obtain the first local network structure corresponding to the current remaining key points in the source image, wherein the current remaining key points in the source image are source key points of the first local network structure, and the preset number of first remaining key points are edge key points of the first local network structure; The obtaining of a second local network structure corresponding to each remaining key point of the target image includes: For each of the remaining key points of the target image, perform the following steps: Acquire, from all the remaining key points of the target image, the preset number of second remaining key points that are closest to the current remaining key point in the target image; Connect the preset number of second remaining key points in sequence to obtain the second local network structure corresponding to the current remaining key points in the target image, wherein the current remaining key points in the target image are the source key points of the second local network structure, and the preset number of second remaining key points are the edge key points of the second local network structure.

4. The automatic registration method for remote sensing image data according to claim 1, characterized in that: The step of obtaining reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure includes: Perform the following steps for each of the first local network structures: Calculating an initial similarity between the current first local network structure and each of the second local network structures; Obtain the target second local network structure corresponding to the maximum initial similarity; Obtaining a final structural similarity between the current first partial network structure and the target second partial network structure; According to the brightness value of the current first local network structure and the brightness value of the target second local network structure, obtaining the ground vegetation color similarity between the current first local network structure and the target second local network structure; According to the final structural similarity and the ground vegetation color similarity, obtaining a second stable matching point pair possibility value between the source key point corresponding to the current first local network structure and the source key point in the target second local network structure; When the likelihood value of the second stable matching point pair is greater than a second preset threshold, the source key point in the target second local network structure is determined to be a reliable matching point of the source key point corresponding to the current first local network structure.

5. The automatic registration method for remote sensing image data according to claim 4, characterized in that: The obtaining of the final structural similarity between the current first partial network structure and the target second partial network structure includes: Acquire a first distance between a source key point of the current first partial network structure and each edge key point of the current first partial network structure; Acquire a second distance between a source key point of the target second partial network structure and each edge key point of the target second partial network structure; Acquire a first angle between a line connecting a source key point of the current first partial network structure and each edge key point of the current first partial network structure and a horizontal line; Acquire a second angle between a source key point of the target second partial network structure and a line connecting each edge key point of the target second partial network structure and a horizontal line; According to the first distance, the second distance, the first angle, and the second angle, a final structural similarity between the current first local network structure and the target second local network structure is obtained.

6. The automatic registration method for remote sensing image data according to claim 4, characterized in that: The obtaining, according to the brightness value of the current first local network structure and the brightness value of the target second local network structure, the ground vegetation color similarity between the current first local network structure and the target second local network structure includes: Obtaining a first grayscale value of each edge key point of the current first local network structure; Obtaining a second grayscale value of each edge key point of the target second local network structure; The ground vegetation color similarity is acquired according to all the first grayscale values ​​and all the second grayscale values.

7. An automatic registration device for remote sensing image data, characterized in that: The device comprises: The first acquisition module is used to acquire remote sensing images of the area to be collected at different times; A second acquisition module is used to analyze any two remote sensing images to be registered, and obtain key points in the two remote sensing images to be registered; The third acquisition module is used to obtain matching point pairs among key points of the source image and key points of the target image, where the source image and the target image are two remote sensing images to be registered; And, obtaining a first target matching point pair, the first target matching point pair being a matching point pair in which the probability value of the first stable matching point pair in each pair of matching point pairs is greater than a first preset threshold value; wherein the probability value of the first stable matching point pair between the current key point and the initial nearest neighbor key point is obtained based on the feature expression credibility, relative distance, and the corresponding angle value of the initial nearest neighbor key point and the nearest neighbor key point of the current key point in the target image, the current key point is a key point in the source image, the initial nearest neighbor key point is a key point in the target image, the initial nearest neighbor key point and the current key point are a pair of matching point pairs, the relative distance includes the relative distance between the current key point and the initial nearest neighbor key point, the nearest neighbor key point, and the next nearest neighbor key point of the current key point in the target image, the feature expression credibility is obtained based on the key point distribution density in the first preset neighborhood corresponding to the current key point and the key point distribution density in the second preset neighborhood corresponding to the initial nearest neighbor key point; And, obtaining a first local network structure corresponding to each remaining key point of the source image and a second local network structure corresponding to each remaining key point of the target image, wherein the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pair in the source image, and the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pair in the target image; and, obtaining reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure; And, obtaining reliable matching point pairs in the two remote sensing images to be registered according to the first target matching point pairs and the second target matching point pairs, wherein the second target matching point pairs are composed of the remaining key points in the source image and the corresponding reliable matching points; The matching module is used to match all the reliable matching point pairs in the two remote sensing images to be registered to obtain the target remote sensing image after registration.

8. An automatic registration system for remote sensing image data, characterized in that: The system comprises a drone and an automatic registration device for remote sensing image data as claimed in claim 7; The drone is used to capture remote sensing images of the area to be collected at different times, and send the remote sensing images of the collection area at different times to the first acquisition module; The first acquisition module is used to acquire remote sensing images of the area to be collected at different times; The second acquisition module is used to analyze any two remote sensing images to be registered, and obtain key points in the two remote sensing images to be registered; The third acquisition module is used to obtain matching point pairs among key points of a source image and key points of a target image, wherein the source image and the target image are two remote sensing images to be registered; And, obtaining a first target matching point pair, the first target matching point pair being a matching point pair in which the probability value of the first stable matching point pair in each pair of matching point pairs is greater than a first preset threshold value; wherein the probability value of the first stable matching point pair between the current key point and the initial nearest neighbor key point is obtained based on the feature expression credibility, relative distance, and the corresponding angle value of the initial nearest neighbor key point and the nearest neighbor key point of the current key point in the target image, the current key point is a key point in the source image, the initial nearest neighbor key point is a key point in the target image, the initial nearest neighbor key point and the current key point are a pair of matching point pairs, the relative distance includes the relative distance between the current key point and the initial nearest neighbor key point, the nearest neighbor key point, and the next nearest neighbor key point of the current key point in the target image, the feature expression credibility is obtained based on the key point distribution density in the first preset neighborhood corresponding to the current key point and the key point distribution density in the second preset neighborhood corresponding to the initial nearest neighbor key point; And, obtaining a first local network structure corresponding to each remaining key point of the source image and a second local network structure corresponding to each remaining key point of the target image, wherein the remaining key points in the source image are the key points remaining after removing the key points corresponding to the first target matching point pair in the source image, and the remaining key points in the target image are the key points remaining after removing the key points corresponding to the first target matching point pair in the target image; and, obtaining reliable matching points of each remaining key point in the source image among the remaining key points in the target image according to the first local network structure, the second local network structure, the brightness of the first local network structure, and the brightness of the second local network structure; And, obtaining reliable matching point pairs in the two remote sensing images to be registered according to the first target matching point pairs and the second target matching point pairs, wherein the second target matching point pairs are composed of the remaining key points in the source image and the corresponding reliable matching points; The matching module is used to match all the reliable matching point pairs in the two remote sensing images to be registered to obtain the registered target remote sensing image.

Images