Geometric correction method and device for images without geographic information

Abstract

The present invention provides a method and device for geometric correction of an image without geographic information, and relates to the technical field of image processing. The method comprises: resampling each reference image N times to construct a reference image pyramid; storing feature points corresponding to reference images of different levels in the reference image pyramid into different storage areas, and storing feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas to construct a feature point library; determining a target level that matches the resolution of the image without geographic information to be corrected; extracting target feature points from the feature point library according to the target level; taking out feature points corresponding to the same sub-geographical area from the target feature points each time to match the feature points of the image without geographic information to be corrected; and geometrically correcting the image without geographic information to be corrected based on the successfully matched target feature points. The present invention improves the efficiency of geometric correction of images without geographic coordinates.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method and device for geometric correction of an image without geographic information. Background Art

[0002] Non-geographic information images refer to images without projection information and geographic coordinates, and without rational polynomial coefficient (RPC) parameters. In practical applications, it is necessary to perform geometric correction on non-geographic information images to generate images with high geometric accuracy. Geometric correction is generally based on high geometric accuracy images. First, a large number of high-precision control points are obtained through automatic matching, and then these control points are used to build a geometric correction model to complete the geometric correction of the image.

[0003] For images without geographic information, it is impossible to determine their corresponding geographic areas, resulting in the inability to automatically determine their corresponding reference images. The corresponding reference images can only be determined manually or through image retrieval, and the geometric correction process of images without geographic information cannot be completed quickly. Summary of the invention

[0004] The present invention provides a method and device for geometric correction of images without geographic information, which is used to solve the problem in the prior art that, for images without geographic information, their corresponding reference images can only be determined manually or by image retrieval, and the geometric correction process of the images without geographic information cannot be completed quickly.

[0005] The present invention provides a method for geometric correction of an image without geographic information, comprising:

[0006] Resampling each reference image N times to construct a reference image pyramid; reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid;

[0007] The feature points corresponding to the reference images at different levels in the reference image pyramid are stored in different storage areas, and the feature points corresponding to different sub-geographical areas in the same storage area are stored in different sub-storage areas to construct a feature point library;

[0008] Determining a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected;

[0009] Extracting target feature points from the feature point library; the target feature points are feature points stored in a storage area corresponding to the target level;

[0010] From the target feature points, feature points corresponding to the same sub-geographical area are taken out each time to match with feature points of the image without geographic information to be corrected;

[0011] The image without geographic information to be corrected is geometrically corrected based on the successfully matched target feature points.

[0012] In some embodiments, before storing the feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas, the method includes:

[0013] A large geographical area is divided into blocks to obtain a plurality of first-level sub-geographical areas;

[0014] According to the preset level, the following steps are executed cyclically:

[0015] Each sub-geographical area of ​​the current level is divided into blocks to obtain a plurality of sub-geographical areas of the next level;

[0016] Stop the loop and obtain the constructed geographic region segmentation standard library; the geographic region segmentation standard library includes sub-geographic regions of various levels;

[0017] According to the construction of the geographic area segmentation standard library, the storage area is segmented to obtain sub-storage areas of various levels; one sub-storage area corresponds to one sub-geographic area.

[0018] In some embodiments, extracting target feature points from the feature point library includes:

[0019] In the case of determining the initial geographic range corresponding to the image without geographic information to be corrected, determining a target sub-geographic area; the target sub-geographic area is a sub-geographic area of ​​the next level corresponding to the sub-geographic area matching the initial geographic range;

[0020] From the feature points stored in the storage area corresponding to the target level, feature points stored in the sub-storage area corresponding to the target sub-geographical area are extracted.

[0021] In some embodiments, the step of extracting feature points corresponding to the same sub-geographical area from the target feature points each time and matching them with feature points of the image without geographic information to be corrected includes:

[0022] When the feature points corresponding to a certain sub-geographical area are successfully matched with the feature points of the image without geographic information to be corrected, the feature points corresponding to the sub-geographical area adjacent to the certain sub-geographical area among the target feature points are matched with the feature points of the image without geographic information to be corrected.

[0023] In some embodiments, before storing the feature points corresponding to the reference images at different levels in the reference image pyramid in different storage areas, the method includes:

[0024] Performing image segmentation on the reference image of each level in the reference image pyramid to obtain a sub-reference image of each level in the reference image pyramid;

[0025] Feature points are extracted from the sub-reference images at each level in the reference image pyramid.

[0026] In some embodiments, after extracting feature points corresponding to the same sub-geographical area from the target feature points each time and matching them with feature points of the image without geographic information to be corrected, the method further includes:

[0027] In case of a matching failure, replacing the target level;

[0028] The feature points stored in the storage area corresponding to the replaced target level are matched with the feature points of the image without geographic information to be corrected.

[0029] The present invention also provides a geometric correction device for an image without geographic information, comprising:

[0030] A first construction module is used to resample each reference image N times to construct a reference image pyramid; reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid;

[0031] A second construction module is used to store feature points corresponding to reference images at different levels in the reference image pyramid into different storage areas, and to store feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas to construct a feature point library;

[0032] A determination module, configured to determine a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected;

[0033] A first extraction module is used to extract target feature points from the feature point library; the target feature points are feature points stored in a storage area corresponding to the target level;

[0034] A first matching module is used to extract feature points corresponding to the same sub-geographical area from the target feature points each time and match them with feature points of the image without geographic information to be corrected;

[0035] The correction module is used to perform geometric correction on the image without geographic information to be corrected based on the successfully matched target feature points.

[0036] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for geometric correction of an image without geographic information as described above is implemented.

[0037] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for geometric correction of an image without geographic information as described in any one of the above is implemented.

[0038] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned methods for geometric correction of images without geographic information.

[0039] The geometric correction method and device of the image without geographic information provided by the present invention firstly constructs a reference image pyramid, and the resolutions corresponding to the reference images at different levels in the reference image pyramid are different, so as to realize the resolution of the annotated reference image, which is conducive to the unified extraction of feature points; then, by constructing a nested storage feature point library, it is conducive to the rapid extraction of target feature points; then, each time, the feature points corresponding to the same sub-geographical area are taken out from the target feature points to match the feature points of the image without geographic information to be corrected, and the matching time is reduced by reducing the number of feature points matched each time; finally, the image without geographic information to be corrected is geometrically corrected based on the successfully matched target feature points. The present invention does not need to obtain the reference image corresponding to the image without geographic information to be corrected, and automatically and quickly obtains the feature points of the reference image from the feature point library, and matches the obtained feature points with the feature points of the image without geographic information to be corrected according to the corresponding sub-geographical area, thereby greatly reducing the matching time and improving the geometric correction efficiency of the image without geographic coordinates. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the present invention or 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.

[0041] Figure 1 It is a schematic flow chart of a method for geometric correction of an image without geographic information provided by the present invention;

[0042] Figure 2 It is a structural schematic diagram of a geometric correction device for an image without geographic information provided by the present invention;

[0043] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] Figure 1 is a flow chart of the geometric correction method for images without geographic information provided by the present invention, such as Figure 1 As shown, the geometric correction method of the image without geographic information provided by the present invention includes:

[0046] Step 110 , resample each reference image N times to construct a reference image pyramid; reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid.

[0047] Specifically, in actual use, the reference images may have different resolutions, which makes it inconvenient to extract and manage feature points in a unified manner.

[0048] In order to solve the above problem, a reference image pyramid is constructed, in which reference images at the same level in the reference image pyramid have the same resolution, and reference images at different levels in the reference image pyramid have different resolutions. For example, the larger the level, the higher the resolution, or the larger the level, the lower the resolution.

[0049] Exemplarily, the setting rules of the reference image resolutions at different levels in the reference image pyramid are as follows: let the reference image resolution of the first level of the reference image pyramid be A, and the reference image resolution of the second level be 2 A, the base image resolution of level N is N A. The values ​​of N and A are determined according to the possible resolution of the non-geographic coordinate image to be corrected and the difference between the reference image.

[0050] According to the resolution corresponding to each level in the reference image pyramid, each reference image is resampled so that the resolution of the resampled reference image is the same as the resolution corresponding to the level in the reference image pyramid. If the number of levels of the reference image pyramid is N, each reference image is resampled N times so that each reference image exists in each level of the reference image pyramid.

[0051] By constructing a benchmark image pyramid, the resolution of the benchmark images is standardized so that all benchmark images have standardized resolutions.

[0052] Step 120 , storing feature points corresponding to reference images at different levels in the reference image pyramid into different storage areas, and storing feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas, to construct a feature point library.

[0053] Specifically, feature points are extracted from the reference image at each level in the reference image pyramid. After feature point extraction is completed, each feature point includes the following information: feature point geographic coordinates, feature point pixel coordinates, feature vectors, and the level of the reference image pyramid.

[0054] In order to facilitate the query of feature points, feature points corresponding to reference images at different levels in the reference image pyramid are stored in different storage areas, and one level in the reference image pyramid corresponds to one storage area.

[0055] In order to further improve the efficiency of feature point query, the feature points in the same storage area are stored in different sub-storage areas according to the different sub-geographical areas corresponding to the feature points, and one sub-geographical area corresponds to one sub-storage area, so as to realize the construction of the feature point library. The sub-geographical area is obtained by dividing a large geographical area.

[0056] The stored feature point information includes: feature point geographic coordinates, feature point pixel coordinates, feature vector and other information.

[0057] Step 130 : determining a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected.

[0058] Specifically, the resolution of the image without geographic information to be corrected is determined according to known information of the image without geographic information to be corrected or through manual analysis.

[0059] The reference image resolutions of different levels in the reference image pyramid are matched with the resolution of the image without geographic information to be corrected, and the level corresponding to the reference image resolution closest to the resolution of the image without geographic information to be corrected is determined, and the level is determined as the target level.

[0060] Step 140, extracting target feature points from the feature point library; the target feature points are feature points stored in the storage area corresponding to the target level.

[0061] Specifically, feature points stored in a storage area corresponding to a target level are extracted from the feature point library, and the extracted feature points are regarded as target feature points.

[0062] Step 150 , from the target feature points, feature points corresponding to the same sub-geographical area are taken out each time to match with feature points of the image without geographic information to be corrected.

[0063] Specifically, since the geographical area corresponding to the target feature points may be very large and the number of target feature points is also large, if the target feature points are directly matched with the feature points of the image without geographic information to be corrected, there will be problems such as long matching time and inability to accurately locate the geographical area corresponding to the image without geographic information to be corrected.

[0064] In order to avoid the above problems, only the feature points corresponding to the same sub-geographical area are taken out from the target feature points each time to match with the feature points of the image without geographic information to be corrected. That is, only the feature points stored in the same sub-storage area are taken out from the target feature points each time to match with the feature points of the image without geographic information to be corrected. This can not only reduce the matching time but also accurately locate the geographic area corresponding to the image without geographic information to be corrected.

[0065] In some embodiments, after determining the target feature points, the feature points stored in the same sub-storage area of ​​the target feature points can be extracted from the feature point library each time, so as to avoid extracting a large number of feature points at one time and wasting resources.

[0066] Step 160 , geometrically correct the image without geographic information to be corrected based on the successfully matched target feature points.

[0067] Specifically, the successfully matched target feature points are used to construct models such as polynomials and triangulated networks to complete the geometric correction of the image without geographic information to be corrected, and the geometrically corrected image is output.

[0068] The geometric correction method of the image without geographic information provided by the present invention firstly constructs a reference image pyramid, and the resolutions corresponding to the reference images at different levels in the reference image pyramid are different, so as to realize the resolution of the annotated reference image, which is conducive to the unified extraction of feature points; then, by constructing a nested storage feature point library, it is conducive to the rapid extraction of target feature points; then, each time, the feature points corresponding to the same sub-geographical area are taken out from the target feature points to match the feature points of the image without geographic information to be corrected, and the matching time is reduced by reducing the number of feature points matched each time; finally, the image without geographic information to be corrected is geometrically corrected based on the successfully matched target feature points. The present invention does not need to obtain the reference image corresponding to the image without geographic information to be corrected, and automatically and quickly obtains the feature points of the reference image from the feature point library, and matches the obtained feature points with the feature points of the image without geographic information to be corrected according to the corresponding sub-geographical area, thereby greatly reducing the matching time and improving the geometric correction efficiency of the image without geographic coordinates.

[0069] In some embodiments, before storing feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas, the method includes:

[0070] A large geographical area is divided into blocks to obtain a plurality of first-level sub-geographical areas;

[0071] According to the preset level, the following steps are executed cyclically:

[0072] Each sub-geographical area of ​​the current level is divided into blocks to obtain a plurality of sub-geographical areas of the next level;

[0073] Stop the loop and obtain the constructed geographic region block standard library; the geographic region block standard library includes sub-geographic regions of various levels;

[0074] According to the construction of the geographic area segmentation standard library, the storage area is segmented to obtain sub-storage areas of various levels; one sub-storage area corresponds to one sub-geographic area.

[0075] Specifically, the feature points extracted from the reference image are massive. If the massive feature points are directly stored in the storage area, as the amount of data increases, the storage speed and retrieval speed will become slower and slower.

[0076] In order to improve the storage and retrieval speed of feature points, a large geographical area is divided into multiple levels of sub-geographical areas, and a standard library of geographical area blocks is constructed.

[0077] The specific process of constructing a geographic region segmentation standard library may be: first, a large-scale geographic region is segmented to obtain a plurality of first-level sub-geographic regions; then, each first-level sub-geographic region is segmented to obtain a plurality of second-level sub-geographic regions; then, each second-level sub-geographic region is segmented to obtain a plurality of third-level sub-geographic regions; and so on, until a preset level of sub-geographic regions is obtained, segmentation is stopped, and a constructed geographic region segmentation standard library is obtained.

[0078] The large-scale geographical area may be a global geographical area, and the geographical range includes a longitude range and a latitude range, wherein the longitude range is from -180 degrees to 180 degrees, and the latitude range is from -90 degrees to 90 degrees.

[0079] For example, the global geographic region is first divided into four first-level sub-geographic regions; then, each first-level sub-geographic region is further divided into four second-level sub-geographic regions, so that a total of a second-level sub-geographical area; and so on.

[0080] In actual use, the specific selection of the preset level needs to be determined according to the size of the geographical area corresponding to the reference image of the feature points to be extracted. The larger the geographical area corresponding to the reference image, the higher the preset level.

[0081] According to the construction of the geographic area block standard library, the storage area is divided into blocks so that one sub-geographic area corresponds to one sub-storage area, thereby obtaining sub-storage areas of various levels.

[0082] The geometric correction method for images without geographic information provided by the present invention divides a large-scale geographic area into sub-geographic areas of multiple levels, constructs a geographic area block standard library, and performs block processing on the storage area according to the construction of the geographic area block standard library to obtain sub-storage areas of various levels, so that feature points can be stored in sub-storage areas of different levels according to the corresponding sub-geographic areas of different levels, thereby improving the storage speed and retrieval speed of feature points.

[0083] In some embodiments, extracting target feature points from a feature point library includes:

[0084] When the initial geographic range corresponding to the image without geographic information to be corrected is determined, a target sub-geographic area is determined; the target sub-geographic area is a sub-geographic area of ​​the next level corresponding to the sub-geographic area matching the initial geographic range;

[0085] From the feature points stored in the storage area corresponding to the target level, feature points stored in the sub-storage area corresponding to the target sub-geographical area are extracted.

[0086] Specifically, when the initial geographical range corresponding to the image without geographical information to be corrected is determined, that is, when the approximate geographical range corresponding to the image without geographical information to be corrected is known, a sub-geographical area matching the initial geographical range is determined.

[0087] Since the initial geographic scope is roughly determined and not precise enough, the sub-geographic area at the next level corresponding to the sub-geographic area matching the initial geographic scope is determined as the target sub-geographic area.

[0088] From the feature points stored in the storage area corresponding to the target level, feature points stored in the sub-storage area corresponding to the target sub-geographical area are extracted, thereby narrowing the range of the target feature points.

[0089] The geometric correction method for images without geographic information provided by the present invention, when determining the initial geographic range corresponding to the image without geographic information to be corrected, determines the sub-geographic area of ​​the next level corresponding to the sub-geographic area matching the initial geographic range as the target sub-geographic area, and extracts the feature points stored in the sub-storage area corresponding to the target sub-geographic area from the feature points stored in the storage area corresponding to the target level, thereby narrowing the range of the target feature points and further improving the matching efficiency.

[0090] In some embodiments, from the target feature points, feature points corresponding to the same sub-geographical area are taken out each time to match with feature points of the image without geographic information to be corrected, including:

[0091] When the feature points corresponding to a certain sub-geographical area are successfully matched with the feature points of the image without geographic information to be corrected, the feature points corresponding to the sub-geographical area adjacent to the certain sub-geographical area among the target feature points are matched with the feature points of the image without geographic information to be corrected.

[0092] Specifically, when the feature points corresponding to a certain sub-geographical area successfully match the feature points of the image without geographic information to be corrected, it indicates that the feature points corresponding to the sub-geographical area adjacent to the certain sub-geographical area have a higher probability of successfully matching the feature points of the image without geographic information to be corrected; while the feature points corresponding to the sub-geographical area not adjacent to the certain sub-geographical area have a lower probability of successfully matching the feature points of the image without geographic information to be corrected.

[0093] In order to further reduce the matching time, feature points corresponding to sub-geographical areas that are not adjacent to the sub-geographical area are temporarily not considered, and feature points corresponding to sub-geographical areas adjacent to the sub-geographical area among the target feature points are matched with feature points of the image without geographic information to be corrected.

[0094] In the case that the feature points corresponding to the sub-geographical areas adjacent to the certain sub-geographical area fail to match, the feature points corresponding to the sub-geographical areas not adjacent to the certain sub-geographical area may be matched with the feature points of the image without geographic information to be corrected.

[0095] The geometric correction method for images without geographic information provided by the present invention, when the feature points corresponding to a certain sub-geographic area are successfully matched with the feature points of the image without geographic information to be corrected, the feature points corresponding to the sub-geographic areas adjacent to the certain sub-geographic area among the target feature points are matched with the feature points of the image without geographic information to be corrected, thereby further reducing the matching time.

[0096] In some embodiments, before storing feature points corresponding to reference images at different levels in the reference image pyramid in different storage areas, the process includes:

[0097] Performing image segmentation on the reference image of each level in the reference image pyramid to obtain a sub-reference image of each level in the reference image pyramid;

[0098] Feature points are extracted from the sub-reference images at each level in the reference image pyramid.

[0099] Specifically, in order to improve the efficiency of feature point extraction, the reference image of each level in the reference image pyramid is divided into blocks, and the number of blocks corresponding to the same level is the same, and the number of blocks corresponding to different levels can be the same or different.

[0100] Each reference image is divided into multiple sub-reference images. For example, each reference image is divided into M sub-reference images. There are M sub-reference images, and the value of M is determined according to the hardware resources required by the feature point extraction algorithm and the computer configuration used.

[0101] Feature points are extracted from the sub-reference images at each level in the reference image pyramid. The extraction method may be parallel extraction, that is, feature points are extracted from multiple sub-reference images at the same level at the same time.

[0102] The commonly used feature point extraction methods with rotation and scale invariance, such as Scale-Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF) and Oriented FAST and Rotated BRIEF (ORB) in image matching, are used for feature point extraction.

[0103] The geometric correction method for images without geographic information provided by the present invention divides the reference image of each level in the reference image pyramid into sub-reference images, uses the sub-reference images as objects for feature point extraction, and improves the feature point extraction efficiency by reducing the image size of the object for feature point extraction.

[0104] In some embodiments, after extracting feature points corresponding to the same sub-geographical area from the target feature points each time and matching them with feature points of the image without geographic information to be corrected, the method further includes:

[0105] In case of matching failure, change the target level;

[0106] The feature points stored in the storage area corresponding to the replaced target level are matched with the feature points of the image without geographic information to be corrected.

[0107] Specifically, after the feature point matching is completed, if the number of matching points is less than the first preset threshold (for example, 10), it is considered that the sub-geographic area corresponding to the feature point fails to match; if the number of matching points is greater than the first preset threshold, the random sample consensus (RANSAC) method is used to eliminate erroneous matching points. If the elimination fails or the number of eliminated matching points is less than the second preset threshold (for example, 4), it is considered that the sub-geographic area corresponding to the feature point fails to match.

[0108] If all sub-geographical areas corresponding to the target feature points fail to match, it indicates that the target level determined to match the resolution of the image without geographic information to be corrected is wrong and needs to be replaced. The replaced target level can be a level adjacent to the original target level.

[0109] The feature points stored in the storage area corresponding to the replaced target level are matched with the feature points of the image without geographic information to be corrected.

[0110] The geometric correction method for images without geographic information provided by the present invention replaces the target level when matching fails; the feature points stored in the storage area corresponding to the replaced target level are matched with the feature points of the image without geographic information to be corrected, so as to automatically correct errors and improve the matching success rate.

[0111] The geometric correction device for images without geographic information provided by the present invention is described below. The geometric correction device for images without geographic information described below and the geometric correction method for images without geographic information described above can be referenced to each other.

[0112] Figure 2 Schematic diagram of the structure of the geometric correction device for the image without geographic information provided by the present invention. Figure 2 As shown, the present invention provides a geometric correction device for an image without geographic information, comprising:

[0113] A first construction module 210 is used to resample each reference image N times to construct a reference image pyramid; reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid;

[0114] The second construction module 220 is used to store the feature points corresponding to the reference images at different levels in the reference image pyramid into different storage areas, and store the feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas to construct a feature point library;

[0115] A determination module 230, configured to determine a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected;

[0116] A first extraction module 240 is used to extract target feature points from the feature point library; the target feature points are feature points stored in a storage area corresponding to the target level;

[0117] A first matching module 250 is used to extract feature points corresponding to the same sub-geographical area from the target feature points each time and match them with feature points of the image without geographic information to be corrected;

[0118] The correction module 260 is used to perform geometric correction on the image without geographic information to be corrected based on the successfully matched target feature points.

[0119] In some embodiments, the apparatus further comprises:

[0120] A third construction module is used to perform block processing on a large-scale geographical area to obtain a plurality of first-level sub-geographical areas;

[0121] According to the preset level, the following steps are executed cyclically:

[0122] Each sub-geographical area of ​​the current level is divided into blocks to obtain a plurality of sub-geographical areas of the next level;

[0123] Stop the loop and obtain the constructed geographic region segmentation standard library; the geographic region segmentation standard library includes sub-geographic regions of various levels;

[0124] The first segmentation module is used to segment the storage area into blocks according to the construction of the geographic area segmentation standard library to obtain sub-storage areas of various levels; one sub-storage area corresponds to one sub-geographic area.

[0125] In some embodiments, the first extraction module 240 is specifically used to:

[0126] In the case of determining the initial geographic range corresponding to the image without geographic information to be corrected, determining a target sub-geographic area; the target sub-geographic area is a sub-geographic area of ​​the next level corresponding to the sub-geographic area matching the initial geographic range;

[0127] From the feature points stored in the storage area corresponding to the target level, feature points stored in the sub-storage area corresponding to the target sub-geographical area are extracted.

[0128] In some embodiments, the first matching module 250 is specifically used for:

[0129] When the feature points corresponding to a certain sub-geographical area are successfully matched with the feature points of the image without geographic information to be corrected, the feature points corresponding to the sub-geographical area adjacent to the certain sub-geographical area among the target feature points are matched with the feature points of the image without geographic information to be corrected.

[0130] In some embodiments, the apparatus further comprises:

[0131] A second blocking module is used to block the reference image of each level in the reference image pyramid to obtain a sub-reference image of each level in the reference image pyramid;

[0132] The second extraction module is used to extract feature points from the sub-reference images at each level in the reference image pyramid.

[0133] In some embodiments, the apparatus further comprises:

[0134] A replacement module, used to replace the target level in case of a matching failure;

[0135] The second matching module is used to match the feature points stored in the storage area corresponding to the replaced target level with the feature points of the image without geographic information to be corrected.

[0136] It should be noted here that the above-mentioned geometric correction device for images without geographic information provided by the present invention can implement all the method steps implemented by the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0137] Figure 3 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 3 As shown, the electronic device may include: a processor 310 , a communications interface 320 , a memory 330 and a communication bus 340 , wherein the processor 310 , the communications interface 320 , and the memory 330 communicate with each other via the communication bus 340 . The processor 310 can call the logic instructions in the memory 330 to execute the geometric correction method of the image without geographic information, which includes: resampling each reference image N times to construct a reference image pyramid; the reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid; the feature points corresponding to the reference images at different levels in the reference image pyramid are stored in different storage areas, and the feature points corresponding to different sub-geographical areas in the same storage area are stored in different sub-storage areas to construct a feature point library; determine a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected; extract target feature points from the feature point library; the target feature points are feature points stored in the storage area corresponding to the target level; from the target feature points, each time the feature points corresponding to the same sub-geographical area are taken out to match the feature points of the image without geographic information to be corrected; and the image without geographic information to be corrected is geometrically corrected based on the successfully matched target feature points.

[0138] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0139] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the geometric correction method of the non-geographic information image provided by the above methods, the method including: resampling each reference image N times to construct a reference image pyramid; the reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid; storing feature points corresponding to reference images at different levels in the reference image pyramid in different storage areas, and storing feature points corresponding to different sub-geographic areas in the same storage area in different sub-storage areas to construct a feature point library; determining a target level in the reference image pyramid that matches the resolution of the non-geographic information image to be corrected; extracting target feature points from the feature point library; the target feature points are feature points stored in the storage area corresponding to the target level; from the target feature points, taking out feature points corresponding to the same sub-geographic area each time to match with feature points of the non-geographic information image to be corrected; and performing geometric correction on the non-geographic information image to be corrected based on the successfully matched target feature points.

[0140] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the processor executes the method for geometric correction of the image without geographic information provided by the above methods, the method comprising: resampling each reference image N times to construct a reference image pyramid; the reference images at different levels in the reference image pyramid have different resolutions, and N is the number of levels of the reference image pyramid; storing feature points corresponding to reference images at different levels in the reference image pyramid in different storage areas, and storing feature points corresponding to different sub-geographical areas in the same storage area in different sub-storage areas to construct a feature point library; determining a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected; extracting target feature points from the feature point library; the target feature points are feature points stored in the storage area corresponding to the target level; taking out feature points corresponding to the same sub-geographical area from the target feature points each time to match them with feature points of the image without geographic information to be corrected; and geometrically correcting the image without geographic information to be corrected based on the successfully matched target feature points.

[0141] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0142] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0143] 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 geometric correction of an image without geographic information, characterized in that: include: Resampling each reference image N times respectively to construct a reference image pyramid, wherein each reference image exists at each level of the reference image pyramid; The reference images at different levels in the reference image pyramid have different corresponding resolutions, and N is the number of levels of the reference image pyramid; The feature points corresponding to the reference images at different levels in the reference image pyramid are stored in different storage areas, and the feature points corresponding to different sub-geographical areas in the same storage area are stored in different sub-storage areas to construct a feature point library; Determining a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected; Extracting target feature points from the feature point library; The target feature point is a feature point stored in a storage area corresponding to the target level; From the target feature points, feature points corresponding to the same sub-geographical area are taken out each time to match with feature points of the image without geographic information to be corrected; The image without geographic information to be corrected is geometrically corrected based on the successfully matched target feature points.

2. The method for geometric correction of an image without geographic information according to claim 1, characterized in that: Before storing the feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas, the method includes: A large geographical area is divided into blocks to obtain a plurality of first-level sub-geographical areas; According to the preset level, the following steps are executed cyclically: Each sub-geographical area of ​​the current level is divided into blocks to obtain a plurality of sub-geographical areas of the next level; Stop the loop and obtain the constructed geographic region segmentation standard library; the geographic region segmentation standard library includes sub-geographic regions of various levels; According to the construction of the geographic area segmentation standard library, the storage area is segmented to obtain sub-storage areas of various levels; one sub-storage area corresponds to one sub-geographic area.

3. The method for geometric correction of an image without geographic information according to claim 2, characterized in that: The extracting target feature points from the feature point library comprises: In the case of determining the initial geographic range corresponding to the image without geographic information to be corrected, determining a target sub-geographic area; the target sub-geographic area is a sub-geographic area of ​​the next level corresponding to the sub-geographic area matching the initial geographic range; From the feature points stored in the storage area corresponding to the target level, feature points stored in the sub-storage area corresponding to the target sub-geographical area are extracted.

4. The method for geometric correction of an image without geographic information according to claim 1, characterized in that: The step of taking out feature points corresponding to the same sub-geographical area from the target feature points each time and matching them with feature points of the image without geographic information to be corrected comprises: When the feature points corresponding to a certain sub-geographical area are successfully matched with the feature points of the image without geographic information to be corrected, the feature points corresponding to the sub-geographical area adjacent to the certain sub-geographical area among the target feature points are matched with the feature points of the image without geographic information to be corrected.

5. The method for geometric correction of an image without geographic information according to claim 1, characterized in that: Before storing the feature points corresponding to the reference images at different levels in the reference image pyramid in different storage areas, the method includes: Performing image segmentation on the reference image of each level in the reference image pyramid to obtain a sub-reference image of each level in the reference image pyramid; Feature points are extracted from the sub-reference images at each level in the reference image pyramid.

6. The method for geometric correction of an image without geographic information according to claim 1, characterized in that: After extracting the feature points corresponding to the same sub-geographical area from the target feature points each time and matching them with the feature points of the image without geographic information to be corrected, the method further includes: In case of a matching failure, replacing the target level; The feature points stored in the storage area corresponding to the replaced target level are matched with the feature points of the image without geographic information to be corrected.

7. A geometric correction device for an image without geographic information, characterized in that: include: A first construction module is used to resample each reference image N times to construct a reference image pyramid, wherein each reference image exists at each level of the reference image pyramid; The reference images at different levels in the reference image pyramid have different corresponding resolutions, and N is the number of levels of the reference image pyramid; A second construction module is used to store feature points corresponding to reference images at different levels in the reference image pyramid into different storage areas, and to store feature points corresponding to different sub-geographical areas in the same storage area into different sub-storage areas to construct a feature point library; A determination module, configured to determine a target level in the reference image pyramid that matches the resolution of the image without geographic information to be corrected; A first extraction module, used to extract target feature points from the feature point library; The target feature point is a feature point stored in a storage area corresponding to the target level; A first matching module is used to extract feature points corresponding to the same sub-geographical area from the target feature points each time and match them with feature points of the image without geographic information to be corrected; The correction module is used to perform geometric correction on the image without geographic information to be corrected based on the successfully matched target feature points.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for geometric correction of an image without geographic information as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for geometric correction of an image without geographic information as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for geometric correction of an image without geographic information as described in any one of claims 1 to 6 is implemented.

Images