Building change detection method, system, device and storage medium

Through the standardized processing and feature extraction network of synthetic aperture radar and optical remote sensing images, the building change detection model is trained, and the problem that the existing technology cannot determine the type of building change is solved, and accurate detection of building changes and illegal construction are achieved. Early warning.

CN118657785BActive Publication Date: 2024-11-15HUNAN SUKE INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411152516.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-15
Estimated Expiration
2044-08-21

AI Technical Summary

Technical Problem

The existing building change detection technology cannot determine the type of change of the target protected area based on image changes in different periods, resulting in the inability to effectively warn of illegal construction areas.

Method used

Standardized processing of synthetic aperture radar remote sensing images and optical remote sensing images is adopted, and the building change detection model is trained through the timing deformation matrix and feature extraction network combined with the difference algorithm, to detect the difference in the change value of the building at different moments, and to combine image information to perform the building Change detection.

Benefits of technology

The detection of different types of changes in the target protected areas has been achieved, and the illegal construction areas have been warned, which has improved the accuracy and efficiency of building change detection and reduced the false alarm rate.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118657785B_ABST
    Figure CN118657785B_ABST
Patent Text Reader

Abstract

The present application discloses a method, system, device and storage medium for detecting building changes. The method comprises standardizing a synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to a first historical moment and a second time series deformation matrix corresponding to a second historical moment; aligning each item in the first optical remote sensing image, the second optical remote sensing image, the first time series deformation matrix and the second time series deformation matrix with the other three items, performing building change detection based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned time series deformation matrix and the second aligned time series deformation matrix through a trained building change detection model, obtaining a building change detection result within a target time period, being able to detect different types of changes in a target protection zone, and warning illegal construction areas in the target protection zone.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field related to building change detection, and in particular to a building change detection method, system, device and storage medium. Background Art

[0002] The building protection zone is a key area to ensure the safety of construction. In this area, any irregular construction may cause damage to the surrounding structures, tracks, signal systems, etc.

[0003] At present, the main building change detection technologies include visual interpretation of images by professionals and traditional optical remote sensing image change detection methods, both of which mainly rely on the texture features and color features of the image, but cannot determine the type of change in the target protected area based on image changes in different periods. Summary of the invention

[0004] The present application aims to at least solve the technical problems existing in the prior art. To this end, the present application proposes a building change detection method, system, device and storage medium, which can detect different types of changes in the target protection area and warn of illegal construction areas in the target protection area.

[0005] A third aspect of the present application provides a building change detection method, comprising the following steps:

[0006] Acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing a building in a target area within a target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period;

[0007] Performing standardization processing on the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment;

[0008] Aligning each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the remaining three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix, and a second aligned temporal deformation matrix;

[0009] Based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix, building change detection is performed through a trained building change detection model to obtain a building change detection result within the target time period.

[0010] The control method according to the embodiment of the present application has at least the following beneficial effects:

[0011] The method obtains a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image by photographing the building in the target area within the target time period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target time period; the synthetic aperture radar remote sensing image is standardized to obtain a first temporal deformation matrix corresponding to the first historical moment and a second temporal deformation matrix corresponding to the second historical moment; each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix is ​​aligned with the other three items to obtain a first aligned optical remote sensing image. Remote sensing image, second aligned optical remote sensing image, first aligned time series deformation matrix and second aligned time series deformation matrix; based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned time series deformation matrix and the second aligned time series deformation matrix, building change detection is performed through a trained building change detection model to obtain the building change detection result within the target time period. The present application judges the difference in change values ​​of the building at different times according to the first aligned time series deformation matrix and the second aligned time series deformation matrix, and then detects the building changes through the difference in change values ​​in combination with the first aligned optical remote sensing image and the second aligned optical remote sensing image, so as to detect different types of changes in the target protection area and warn of illegal construction areas in the target protection area.

[0012] According to some embodiments of the present application, the step of performing standardization processing on the synthetic aperture radar remote sensing image to obtain a first temporal deformation matrix corresponding to the first historical moment and a second temporal deformation matrix corresponding to the second historical moment includes:

[0013] Determining the height and width of the synthetic aperture radar remote sensing image;

[0014] Preprocessing the synthetic aperture radar remote sensing image to obtain a preprocessed synthetic aperture radar remote sensing image;

[0015] Calculate the temporal deformation matrix of the target period by using the interference superposition technology on the pre-processed synthetic aperture radar remote sensing image;

[0016] Based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the first historical moment and the preset number before it to obtain a first initial temporal deformation matrix, and based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the second historical moment and the preset number before it to obtain a second initial temporal deformation matrix;

[0017] According to the height and width of the synthetic aperture radar remote sensing image and the preset number, the first initial time series deformation matrix is ​​subjected to matrix transformation to obtain a first transformed time series deformation matrix, and according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the second initial time series deformation matrix is ​​subjected to matrix transformation to obtain a second transformed time series deformation matrix;

[0018] The first temporal deformation matrix is ​​calculated according to the first transformation temporal deformation matrix by the following formula:

[0019]

[0020] in, is the value of the first temporal deformation matrix, is the value of the first transformation time series deformation matrix, for The absolute value of is the preset deformation threshold;

[0021] The second temporal deformation matrix is ​​calculated according to the second transformation temporal deformation matrix:

[0022]

[0023] in, is the value of the second temporal deformation matrix, is the value of the second transformation time series deformation matrix, for The absolute value of .

[0024] According to some embodiments of the present application, the training process of the building change detection model is:

[0025] Acquire a historical synthetic aperture radar remote sensing image, a third optical remote sensing image, and a fourth optical remote sensing image obtained by photographing buildings in a specified area within a preset historical period, wherein the third optical remote sensing image is an optical remote sensing image of a third historical moment when the building has no size change, the fourth optical remote sensing image is an optical remote sensing image of a fourth historical moment when the building has a size change, and the third historical moment and the fourth historical moment are two different moments within the preset historical period;

[0026] Performing standardization processing on the historical synthetic aperture radar remote sensing image to obtain a third temporal deformation matrix corresponding to the third historical moment and a fourth temporal deformation matrix corresponding to the fourth historical moment;

[0027] Aligning each item in the third optical remote sensing image, the fourth optical remote sensing image, the third temporal deformation matrix, and the fourth temporal deformation matrix with the remaining three items to obtain a third aligned optical remote sensing image, a fourth aligned optical remote sensing image, a third aligned temporal deformation matrix, and a fourth aligned temporal deformation matrix;

[0028] Adding building change annotations to the fourth aligned optical remote sensing image based on the third aligned optical remote sensing image, the third aligned temporal deformation matrix, and the fourth aligned temporal deformation matrix to obtain a fifth optical remote sensing image;

[0029] Performing feature extraction on the third aligned optical remote sensing image through a feature extraction network to obtain a first optical remote sensing feature map; performing feature extraction on the fifth optical remote sensing image through the feature extraction network to obtain a second optical remote sensing feature map;

[0030] Calculating a first characteristic map by using a difference algorithm according to the first optical remote sensing characteristic map and the second optical remote sensing characteristic map;

[0031] Calculating a second feature map by using a difference algorithm according to the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix;

[0032] Performing weighted fusion on the first feature map and the second feature map to obtain a fused feature map;

[0033] Inputting the fused feature map into the preset initial building change detection model to obtain a total loss value output by the initial building change detection model;

[0034] The initial building change detection model is reversely optimized according to the total loss value to obtain the trained building change detection model.

[0035] According to some embodiments of the present application, inputting the fused feature map into the preset initial building change detection model to obtain a total loss value output by the initial building change detection model includes:

[0036] Inputting the fused feature map into the initial building change detection model to obtain a predicted value output by the initial building change detection model;

[0037] Calculate a basic loss value according to the predicted value;

[0038] The deformation loss value is calculated according to the predicted value by the following formula:

[0039]

[0040] in, is the deformation loss value, is the total number of feature maps after fusion, For the The true value of the feature map after fusion, For the The predicted value of the feature map after fusion, is the norm;

[0041] The total loss value is calculated according to the basic loss value and the deformation loss value.

[0042] According to some embodiments of the present application, the total loss value is calculated by the following formula:

[0043]

[0044]

[0045] in, is the iteration parameter, is the current iteration number, is the preset maximum number of iterations, is the total loss value, is the basic loss value.

[0046] According to some embodiments of the present application, aligning each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the remaining three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix, and a second aligned temporal deformation matrix, includes:

[0047] Aligning each item of the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the other three items in terms of longitude and latitude based on the WGS-84 coordinate system to obtain a first initially aligned optical remote sensing image, a second initially aligned optical remote sensing image, a first initially aligned temporal deformation matrix, and a second initially aligned temporal deformation matrix;

[0048] The first initially aligned optical remote sensing image, the second initially aligned optical remote sensing image, the first initially aligned temporal deformation matrix, and the second initially aligned temporal deformation matrix are cropped in the unaligned area to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix, and the second aligned temporal deformation matrix.

[0049] According to some embodiments of the present application, acquiring a first optical remote sensing image and a second optical remote sensing image in a target area during a target period includes:

[0050] Acquire a first original optical remote sensing image and a second original optical remote sensing image, wherein the first original optical remote sensing image is an original optical remote sensing image at a first historical moment in the target area, and the second original optical remote sensing image is a second original optical remote sensing image at a second historical moment in the target area;

[0051] Performing geometric correction on the first original optical remote sensing image to obtain a first geometrically corrected optical remote sensing image; performing geometric correction on the second original optical remote sensing image to obtain a second geometrically corrected optical remote sensing image;

[0052] Performing radiation correction on the first geometrically corrected optical remote sensing image to obtain a first radiation-corrected optical remote sensing image; performing radiation correction on the second geometrically corrected optical remote sensing image to obtain a second radiation-corrected optical remote sensing image;

[0053] Performing atmospheric correction on the first radiation-corrected optical remote sensing image to obtain a first atmospheric-corrected optical remote sensing image; performing atmospheric correction on the second radiation-corrected optical remote sensing image to obtain a second atmospheric-corrected optical remote sensing image;

[0054] The first atmospherically corrected optical remote sensing image is geocoded to obtain the first optical remote sensing image; and the second atmospherically corrected optical remote sensing image is geocoded to obtain the second optical remote sensing image.

[0055] In a second aspect of the present application, a building change detection system is provided, the building change detection system comprising:

[0056] A data acquisition module, used to acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing a building in a target area within a target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period;

[0057] A standardization processing module, used for performing standardization processing on the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment;

[0058] A geographic location alignment module, used for aligning each item of the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix with the other three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix and a second aligned temporal deformation matrix;

[0059] A building change detection module is used to perform building change detection based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix through a trained building change detection model to obtain the building change detection result within the target time period.

[0060] The system obtains a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image by photographing buildings in a target area within a target time period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target time period; the synthetic aperture radar remote sensing image is standardized to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment; each item in the first optical remote sensing image, the second optical remote sensing image, the first time series deformation matrix, and the second time series deformation matrix is ​​aligned with the other three items to obtain a first aligned optical remote sensing image. Remote sensing image, second aligned optical remote sensing image, first aligned time series deformation matrix and second aligned time series deformation matrix; based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned time series deformation matrix and the second aligned time series deformation matrix, building change detection is performed through a trained building change detection model to obtain the building change detection result within the target time period. The present application judges the difference in change values ​​of the building at different times according to the first aligned time series deformation matrix and the second aligned time series deformation matrix, and then detects the building changes through the difference in change values ​​in combination with the first aligned optical remote sensing image and the second aligned optical remote sensing image, so as to detect different types of changes in the target protection area and warn of illegal construction areas in the target protection area.

[0061] The third aspect of the present application provides a building change detection electronic device, comprising at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor to enable the at least one control processor to execute the above-mentioned building change detection method.

[0062] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned building change detection method.

[0063] It should be noted that the beneficial effects between the second to fourth aspects of the present application and the prior art are the same as the beneficial effects between the above-mentioned building change detection system and the prior art, and will not be described in detail here.

[0064] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0066] Figure 1 is a flow chart of an embodiment of a building change detection method provided by the present application;

[0067] Figure 2 is a structural schematic diagram of an embodiment of a building change detection system provided by the present application;

[0068] Figure 3 It is a schematic diagram of the structure of an embodiment of the electronic device provided by the present application. DETAILED DESCRIPTION

[0069] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0070] In the description of this application, if there is a description of the third, fourth, etc., it is only for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0071] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0072] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0073] The building protection zone is a key area to ensure the safety of construction. In this area, any irregular construction may cause damage to the surrounding structures, tracks, signal systems, etc.

[0074] At present, the main building change detection technologies include visual interpretation of images by professionals and traditional optical remote sensing image change detection methods, both of which mainly rely on the texture features and color features of the image, but cannot determine the type of change in the target protected area based on image changes in different periods.

[0075] In order to solve the above technical defects, the embodiments of the present application provide a building change detection method, system, device and storage medium.

[0076] See also Figure 1 , is a flow chart of a building change detection method provided in an embodiment of the present application, the method is applied to an electronic device, and the electronic device may be a server, etc. Figure 1 As shown, the building change detection method includes:

[0077] Step S101, obtaining a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image by photographing buildings in a target area within a target time period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target time period.

[0078] Step S102: Standardize the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment.

[0079] Step S103, align each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix with the other three items to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix.

[0080] Step S104: Based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix, building change detection is performed through the trained building change detection model to obtain the building change detection result within the target time period.

[0081] Specifically, the above building change detection result may include that the building within the target time period is not under construction, the building within the target time period is under construction, and the building within the target time period is completed.

[0082] Specifically, the synthetic aperture radar remote sensing image obtained by photographing the buildings in the target area during the target period may include:

[0083] A first original synthetic aperture radar remote sensing image is obtained by photographing buildings in a target area within a target period of time.

[0084] Orbit correction is performed on the first original synthetic aperture radar remote sensing image to obtain a first orbit-corrected synthetic aperture radar remote sensing image.

[0085] A radiation correction is performed on the synthetic aperture radar remote sensing image after the first orbit correction to obtain a first radiation corrected synthetic aperture radar remote sensing image.

[0086] The synthetic aperture radar remote sensing image after the first radiation correction is geometrically corrected to obtain a synthetic aperture radar remote sensing image after the first geometric correction.

[0087] The synthetic aperture radar remote sensing image after the first geometric correction is geocoded to obtain a synthetic aperture radar remote sensing image.

[0088] The method obtains a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image by photographing the building in the target area within the target time period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target time period; the synthetic aperture radar remote sensing image is standardized to obtain a first temporal deformation matrix corresponding to the first historical moment and a second temporal deformation matrix corresponding to the second historical moment; each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix is ​​aligned with the other three items to obtain a first aligned optical remote sensing image. Remote sensing image, second aligned optical remote sensing image, first aligned time series deformation matrix and second aligned time series deformation matrix; based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned time series deformation matrix and the second aligned time series deformation matrix, building change detection is performed through a trained building change detection model to obtain the building change detection result within the target time period. The present application judges the difference in change values ​​of the building at different times according to the first aligned time series deformation matrix and the second aligned time series deformation matrix, and then detects the building changes through the difference in change values ​​in combination with the first aligned optical remote sensing image and the second aligned optical remote sensing image, so as to detect different types of changes in the target protection area and warn of illegal construction areas in the target protection area.

[0089] In some embodiments, in step S102, the synthetic aperture radar remote sensing image is standardized to obtain a first temporal deformation matrix corresponding to the first historical moment and a second temporal deformation matrix corresponding to the second historical moment, including:

[0090] Step S201: Determine the height and width of a synthetic aperture radar remote sensing image.

[0091] Step S202: preprocess the synthetic aperture radar remote sensing image to obtain a preprocessed synthetic aperture radar remote sensing image.

[0092] Step S203: Calculate the temporal deformation matrix of the target period using the interference superposition technology for the preprocessed synthetic aperture radar remote sensing image.

[0093] Step S204: based on the timing deformation matrix of the target time period, extract the timing deformation results of the first historical moment and the preset number before it to obtain a first initial timing deformation matrix; and based on the timing deformation matrix of the target time period, extract the timing deformation results of the second historical moment and the preset number before it to obtain a second initial timing deformation matrix.

[0094] Step S205, according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the first initial time series deformation matrix is ​​matrix transformed to obtain a first transformed time series deformation matrix, and according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the second initial time series deformation matrix is ​​matrix transformed to obtain a second transformed time series deformation matrix.

[0095] Step S206: Calculate the first temporal deformation matrix according to the first transformed temporal deformation matrix by the following formula:

[0096]

[0097] in, is the value of the first temporal deformation matrix, is the value of the first transformation time series deformation matrix, for The absolute value of is the preset deformation threshold;

[0098] Step S207: Calculate the second temporal deformation matrix according to the second transformed temporal deformation matrix:

[0099]

[0100] in, is the value of the second temporal deformation matrix, is the value of the second transformation time series deformation matrix, for The absolute value of .

[0101] Specifically, the above-mentioned Synthetic Aperture Radar (SAR) interferometric stacking technology is an advanced remote sensing technology, which is often used to monitor the deformation and movement of the surface. In this technology, there are two widely used methods, namely Persistent Scatterers (PS) and Small Baselines (SBAS). Both methods can provide time-series SAR image analysis with high sensitivity and high temporal and spatial resolution, providing strong support for geological disasters, urban subsidence, building deformation and other fields.

[0102] The above-mentioned permanent scatterer (PS) method is mainly used to monitor areas with many artificial features, such as cities, buildings, bridges, dams, etc. This method relies on scatterers that are stable in time and have unchanged scattering characteristics. These scatterers are called permanent scatterers. By analyzing multiple SAR images, the PS method can detect and extract the changes of these permanent scatterers, thereby obtaining information on surface deformation. The PS method has high sensitivity and stability and is suitable for monitoring and analysis over a long period of time.

[0103] The above-mentioned small baseline (SBAS) method is mainly used to monitor natural features or other types of non-geometric feature objects. Compared with the PS method, the SBAS method is more suitable for natural environments with complex surface features and large changes. The SBAS method obtains surface deformation information by selecting a set of relatively small baselines, that is, the spatial baselines between adjacent SAR images. This method can better cope with the heterogeneity and complexity of the surface and provide more comprehensive monitoring results.

[0104] Specifically, this embodiment generates a deformation rate map based on heavy-track SAR data through SBAS-Insar technology, and stacks the most recent deformation values ​​into a matrix according to the image width and height after processing, thereby providing a data reference standard for subsequent building change detection and improving the accuracy of building change detection.

[0105] In some embodiments, the training process of the building change detection model in step S104 is:

[0106] Step S301, obtaining a historical synthetic aperture radar remote sensing image, a third optical remote sensing image, and a fourth optical remote sensing image obtained by photographing buildings in a specified area within a preset historical period, wherein the third optical remote sensing image is an optical remote sensing image of a third historical moment when the building has no size change, and the fourth optical remote sensing image is an optical remote sensing image of a fourth historical moment when the building has a size change, and the third historical moment and the fourth historical moment are two different moments within the preset historical period.

[0107] Step S302: Standardize the historical synthetic aperture radar remote sensing images to obtain a third temporal deformation matrix corresponding to a third historical moment and a fourth temporal deformation matrix corresponding to a fourth historical moment.

[0108] Step S303, align each item in the third optical remote sensing image, the fourth optical remote sensing image, the third temporal deformation matrix and the fourth temporal deformation matrix with the other three items to obtain a third aligned optical remote sensing image, a fourth aligned optical remote sensing image, a third aligned temporal deformation matrix and a fourth aligned temporal deformation matrix.

[0109] Step S304: adding building change annotations to the fourth aligned optical remote sensing image based on the third aligned optical remote sensing image, the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix to obtain a fifth optical remote sensing image.

[0110] Step S305: extract features from the third aligned optical remote sensing image through a feature extraction network to obtain a first optical remote sensing feature map; and extract features from the fifth optical remote sensing image through a feature extraction network to obtain a second optical remote sensing feature map.

[0111] Step S306: Calculate the first characteristic map by using a difference algorithm according to the first optical remote sensing characteristic map and the second optical remote sensing characteristic map.

[0112] Step S307: Calculate a second feature map using a difference algorithm according to the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix.

[0113] Step S308: Perform weighted fusion on the first feature map and the second feature map to obtain a fused feature map.

[0114] Step S309: input the fused feature map into a preset initial building change detection model to obtain a total loss value output by the initial building change detection model.

[0115] Step S310: reversely optimize the initial building change detection model according to the total loss value to obtain a trained building change detection model.

[0116] Specifically, the difference algorithm may be image difference, or any one of image difference, image ratio, principal component analysis (PCA) or change vector analysis (CVA).

[0117] In this embodiment, the first feature map is calculated by using a differential algorithm based on the first optical remote sensing feature map and the second optical remote sensing feature map, and the second feature map is calculated by using a differential algorithm based on the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix, and the fused feature map is input into a preset initial building change detection model to obtain a total loss value output by the initial building change detection model. The temporal deformation information, image texture information, image color information, etc. are combined, and the network model is trained based on the total loss value. This can improve the detection rate of construction areas under construction with short cycles, small areas, and small changes, and reduce the false alarm rate of "pseudo-change" areas, thereby warning illegal construction areas within the protected areas.

[0118] In some embodiments, in step S309, the fused feature map is input into a preset initial building change detection model to obtain a total loss value output by the initial building change detection model, including:

[0119] Step S401: input the fused feature map into the initial building change detection model to obtain the predicted value output by the initial building change detection model.

[0120] Step S402: Calculate the basic loss value according to the predicted value.

[0121] Step S403: Calculate the deformation loss value according to the predicted value using the following formula:

[0122]

[0123] in, is the deformation loss value, is the total number of feature maps after fusion, For the The true value of the feature map after fusion, For the The predicted value of the feature map after fusion, is the norm.

[0124] Step S404: Calculate the total loss value according to the basic loss value and the deformation loss value.

[0125] Specifically, the above basic loss value can be calculated by an exponential loss function or a cross entropy loss function.

[0126] In this implementation, a deformation loss function is added to the model to calculate the total loss value, and then the initial building change detection model is reversely optimized using the total loss value, thereby improving the training efficiency and accuracy of the model.

[0127] In some embodiments, the total loss value is calculated in step S404 by the following formula:

[0128]

[0129]

[0130] in, is the iteration parameter, is the current iteration number, is the preset maximum number of iterations, is the total loss value, is the basic loss value.

[0131] In some embodiments, in step S103, each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix is ​​aligned with the other three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix, and a second aligned temporal deformation matrix, including:

[0132] Step S501, align each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix with the other three items in terms of longitude and latitude based on the WGS-84 coordinate system to obtain the first initially aligned optical remote sensing image, the second initially aligned optical remote sensing image, the first initially aligned temporal deformation matrix and the second initially aligned temporal deformation matrix.

[0133] Step S502, cropping the misaligned areas of the first initially aligned optical remote sensing image, the second initially aligned optical remote sensing image, the first initially aligned temporal deformation matrix, and the second initially aligned temporal deformation matrix to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix, and the second aligned temporal deformation matrix.

[0134] This embodiment improves the accuracy of the input data by aligning each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix with the other three items, and then crops the misaligned areas to serve as input data for the subsequent building change detection model, thereby improving the accuracy of the input data and thus improving the accuracy of building change detection.

[0135] In some embodiments, obtaining a first optical remote sensing image and a second optical remote sensing image in a target area during a target period in step S101 includes:

[0136] Step S601: Acquire a first original optical remote sensing image and a second original optical remote sensing image, wherein the first original optical remote sensing image is an original optical remote sensing image at a first historical moment in the target area, and the second original optical remote sensing image is a second original optical remote sensing image at a second historical moment in the target area.

[0137] Step S602: geometrically correct the first original optical remote sensing image to obtain a first geometrically corrected optical remote sensing image; and geometrically correct the second original optical remote sensing image to obtain a second geometrically corrected optical remote sensing image.

[0138] Step S603: performing radiation correction on the first geometrically corrected optical remote sensing image to obtain a first radiation corrected optical remote sensing image; performing radiation correction on the second geometrically corrected optical remote sensing image to obtain a second radiation corrected optical remote sensing image.

[0139] Step S604: performing atmospheric correction on the first radiation-corrected optical remote sensing image to obtain a first atmospheric-corrected optical remote sensing image; performing atmospheric correction on the second radiation-corrected optical remote sensing image to obtain a second atmospheric-corrected optical remote sensing image.

[0140] Step S605: geocoding the first atmospherically corrected optical remote sensing image to obtain a first optical remote sensing image; and geocoding the second atmospherically corrected optical remote sensing image to obtain a second optical remote sensing image.

[0141] This implementation improves the data accuracy of the optical remote sensing image by preprocessing the optical remote sensing image, thereby improving the accuracy of building change detection.

[0142] Specifically, in order to facilitate understanding by those skilled in the art, a set of optimal embodiments is provided below:

[0143] 1. Data acquisition:

[0144] Acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing the building in the target area within the target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period. Specifically:

[0145] A first original synthetic aperture radar remote sensing image is obtained by photographing buildings in a target area within a target period of time.

[0146] Orbit correction is performed on the first original synthetic aperture radar remote sensing image to obtain a first orbit-corrected synthetic aperture radar remote sensing image.

[0147] A radiation correction is performed on the synthetic aperture radar remote sensing image after the first orbit correction to obtain a first radiation corrected synthetic aperture radar remote sensing image.

[0148] The synthetic aperture radar remote sensing image after the first radiation correction is geometrically corrected to obtain a synthetic aperture radar remote sensing image after the first geometric correction.

[0149] The synthetic aperture radar remote sensing image after the first geometric correction is geocoded to obtain a synthetic aperture radar remote sensing image.

[0150] A first original optical remote sensing image and a second original optical remote sensing image are acquired, wherein the first original optical remote sensing image is an original optical remote sensing image at a first historical moment in the target area, and the second original optical remote sensing image is a second original optical remote sensing image at a second historical moment in the target area.

[0151] The first original optical remote sensing image is geometrically corrected to obtain a first geometrically corrected optical remote sensing image; the second original optical remote sensing image is geometrically corrected to obtain a second geometrically corrected optical remote sensing image.

[0152] The first geometrically corrected optical remote sensing image is subjected to radiation correction to obtain a first radiation-corrected optical remote sensing image; the second geometrically corrected optical remote sensing image is subjected to radiation correction to obtain a second radiation-corrected optical remote sensing image.

[0153] Performing atmospheric correction on the first radiation corrected optical remote sensing image to obtain a first atmospheric corrected optical remote sensing image; performing atmospheric correction on the second radiation corrected optical remote sensing image to obtain a second atmospheric corrected optical remote sensing image.

[0154] The first atmospherically corrected optical remote sensing image is geocoded to obtain a first optical remote sensing image; the second atmospherically corrected optical remote sensing image is geocoded to obtain a second optical remote sensing image.

[0155] 2. Data standardization:

[0156] Determine the height and width of a synthetic aperture radar remote sensing image.

[0157] The synthetic aperture radar remote sensing image is preprocessed to obtain a preprocessed synthetic aperture radar remote sensing image.

[0158] The preprocessed synthetic aperture radar remote sensing image is used to calculate the temporal deformation matrix of the target period through interference superposition technology.

[0159] Based on the timing deformation matrix of the target time period, the timing deformation results of the first historical moment and the preset number before it are extracted to obtain a first initial timing deformation matrix, and based on the timing deformation matrix of the target time period, the timing deformation results of the second historical moment and the preset number before it are extracted to obtain a second initial timing deformation matrix.

[0160] According to the height and width of the synthetic aperture radar remote sensing image and the preset number, the first initial time series deformation matrix is ​​transformed into a first transformed time series deformation matrix, and according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the second initial time series deformation matrix is ​​transformed into a second transformed time series deformation matrix.

[0161] The first temporal deformation matrix is ​​calculated according to the first transformation temporal deformation matrix by the following formula:

[0162]

[0163] in, is the value of the first temporal deformation matrix, is the value of the first transformation time series deformation matrix, for The absolute value of The preset deformation threshold.

[0164] The second temporal deformation matrix is ​​calculated according to the second transformation temporal deformation matrix:

[0165]

[0166] in, is the value of the second temporal deformation matrix, is the value of the second transformation time series deformation matrix, for The absolute value of .

[0167] 3. Data alignment:

[0168] Each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix is ​​aligned with the other three items in terms of longitude and latitude based on the WGS-84 coordinate system to obtain the first initial aligned optical remote sensing image, the second initial aligned optical remote sensing image, the first initial aligned temporal deformation matrix and the second initial aligned temporal deformation matrix.

[0169] The first initial aligned optical remote sensing image, the second initial aligned optical remote sensing image, the first initial aligned temporal deformation matrix and the second initial aligned temporal deformation matrix are cropped in the unaligned area to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix.

[0170] 4. Building change detection:

[0171] Based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix, the trained building change detection model is used to perform building change detection to obtain the building change detection result within the target time period. The training process of the building change detection model is as follows:

[0172] A historical synthetic aperture radar remote sensing image, a third optical remote sensing image, and a fourth optical remote sensing image are obtained by photographing buildings in a designated area within a preset historical period, wherein the third optical remote sensing image is an optical remote sensing image of a third historical moment when the building has no size change, and the fourth optical remote sensing image is an optical remote sensing image of a fourth historical moment when the building has a size change, and the third historical moment and the fourth historical moment are two different moments within the preset historical period.

[0173] The historical synthetic aperture radar remote sensing images are standardized to obtain a third time series deformation matrix corresponding to a third historical moment and a fourth time series deformation matrix corresponding to a fourth historical moment.

[0174] Each item in the third optical remote sensing image, the fourth optical remote sensing image, the third temporal deformation matrix and the fourth temporal deformation matrix is ​​aligned with the other three items to obtain a third aligned optical remote sensing image, a fourth aligned optical remote sensing image, a third aligned temporal deformation matrix and a fourth aligned temporal deformation matrix.

[0175] Building change annotations are added to the fourth aligned optical remote sensing image based on the third aligned optical remote sensing image, the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix to obtain a fifth optical remote sensing image.

[0176] The third aligned optical remote sensing image is subjected to feature extraction through a feature extraction network to obtain a first optical remote sensing feature map; the fifth optical remote sensing image is subjected to feature extraction through a feature extraction network to obtain a second optical remote sensing feature map.

[0177] The first characteristic map is calculated according to the first optical remote sensing characteristic map and the second optical remote sensing characteristic map by using a difference algorithm.

[0178] Calculating a second feature map by using a difference algorithm according to the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix;

[0179] The first feature map and the second feature map are weighted fused to obtain a fused feature map.

[0180] The fused feature map is input into the initial building change detection model to obtain the predicted value output by the initial building change detection model.

[0181] Calculate the base loss value based on the predicted value.

[0182] The deformation loss value is calculated based on the predicted value using the following formula:

[0183]

[0184] in, is the deformation loss value, is the total number of feature maps after fusion, For the The true value of the feature map after fusion, For the The predicted value of the feature map after fusion, is the norm.

[0185] The total loss value is calculated based on the basic loss value and the deformation loss value using the following formula:

[0186]

[0187]

[0188] in, is the iteration parameter, is the current iteration number, is the preset maximum number of iterations, is the total loss value, is the basic loss value.

[0189] The initial building change detection model is reversely optimized according to the total loss value to obtain a trained building change detection model.

[0190] In addition, refer to Figure 2In one embodiment of the present application, a building change detection system is provided, including a data acquisition module 1100, a standardization processing module 1200, a geographic location alignment module 1300 and a building change detection module 1400, wherein:

[0191] The data acquisition module 1100 is used to acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing a building in a target area within a target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period;

[0192] The standardization processing module 1200 is used to perform standardization processing on the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment;

[0193] The geographic location alignment module 1300 is used to align each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the other three items to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix, and the second aligned temporal deformation matrix;

[0194] The building change detection module 1400 is used to perform building change detection based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix through a trained building change detection model to obtain the building change detection results within the target time period.

[0195] The system obtains a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image by photographing buildings in a target area within a target time period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target time period; the synthetic aperture radar remote sensing image is standardized to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment; each item in the first optical remote sensing image, the second optical remote sensing image, the first time series deformation matrix, and the second time series deformation matrix is ​​aligned with the other three items to obtain a first aligned optical remote sensing image. Remote sensing image, second aligned optical remote sensing image, first aligned time series deformation matrix and second aligned time series deformation matrix; based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned time series deformation matrix and the second aligned time series deformation matrix, building change detection is performed through a trained building change detection model to obtain the building change detection result within the target time period. The present application judges the difference in change values ​​of the building at different times according to the first aligned time series deformation matrix and the second aligned time series deformation matrix, and then detects the building changes through the difference in change values ​​in combination with the first aligned optical remote sensing image and the second aligned optical remote sensing image, so as to detect different types of changes in the target protection area and warn of illegal construction areas in the target protection area.

[0196] It should be noted that the present system embodiment and the above-mentioned method embodiment are based on the same inventive concept, so the relevant contents of the above-mentioned method embodiment are also applicable to the present system embodiment and will not be repeated here.

[0197] Figure 3 A schematic diagram of the hardware structure of building change detection provided in an embodiment of the present application is shown.

[0198] The building change detection device may include a processor 301 and a memory 302 storing computer program instructions.

[0199] Specifically, the processor 301 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0200] The memory 302 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 302 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 302 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 302 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 302 is a non-volatile solid-state memory.

[0201] In some embodiments, the memory 302 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage media device, an optical storage media device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0202] The processor 301 implements any one of the building change detection methods in the above embodiments by reading and executing computer program instructions stored in the memory 302 .

[0203] In one example, the building change detection device may further include a communication interface 303 and a bus 310. Figure 3 As shown, the processor 301, the memory 302, and the communication interface 303 are connected via a bus 310 and communicate with each other.

[0204] The communication interface 303 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0205] The bus 310 includes hardware, software or both, coupling the components of the building change detection device to each other. For example and not limitation, the bus may include an accelerated graphics port (AGP) or other graphics bus, an enhanced industrial standard architecture (EISA) bus, a front-side bus (FSB), a hypertransport (HT) interconnect, an industrial standard architecture (ISA) bus, an infinite bandwidth interconnect, a low pin count (LPC) bus, a memory bus, a microchannel architecture (MCA) bus, a peripheral component interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a serial advanced technology attachment (SATA) bus, a video electronics standard association local (VLB) bus or other suitable bus or a combination of two or more of these. Where appropriate, the bus 310 may include one or more buses. Although the present application embodiment describes and illustrates a specific bus, the present application considers any suitable bus or interconnect.

[0206] The building change detection device can execute the building change detection method in the embodiment of the present application based on the three-dimensional design model, thereby realizing the combination of Figure 1 and Figure 2 A building change detection method and system are described.

[0207] In addition, in combination with the building change detection method in the above embodiment, the embodiment of the present application can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the building change detection methods in the above embodiment is implemented.

[0208] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0209] The functional blocks shown in the structural block diagram described above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier. "Machine-readable medium" may include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memories, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0210] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0211] Aspects of the present disclosure are described above with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0212] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A building change detection method, characterized in that: The building change detection method comprises: Acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing a building in a target area within a target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period; The synthetic aperture radar remote sensing image is subjected to standardization processing to obtain a first temporal deformation matrix corresponding to a first historical moment and a second temporal deformation matrix corresponding to the second historical moment; the synthetic aperture radar remote sensing image is subjected to standardization processing to obtain a first temporal deformation matrix corresponding to the first historical moment and a second temporal deformation matrix corresponding to the second historical moment, including: Determining the height and width of the synthetic aperture radar remote sensing image; Preprocessing the synthetic aperture radar remote sensing image to obtain a preprocessed synthetic aperture radar remote sensing image; Calculate the temporal deformation matrix of the target period by using the interference superposition technology on the pre-processed synthetic aperture radar remote sensing image; Based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the first historical moment and the preset number before it to obtain a first initial temporal deformation matrix, and based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the second historical moment and the preset number before it to obtain a second initial temporal deformation matrix; According to the height and width of the synthetic aperture radar remote sensing image and the preset number, the first initial time-series deformation matrix is ​​subjected to matrix transformation to obtain a first transformed time-series deformation matrix, and according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the second initial time-series deformation matrix is ​​subjected to matrix transformation to obtain a second transformed time-series deformation matrix; The first temporal deformation matrix is ​​calculated according to the first transformation temporal deformation matrix by the following formula: ; in, is the value of the first temporal deformation matrix, is the value of the first transformation time series deformation matrix, for The absolute value of is the preset deformation threshold; The second temporal deformation matrix is ​​calculated according to the second transformation temporal deformation matrix: ; in, is the value of the second temporal deformation matrix, is the value of the second transformation time series deformation matrix, for The absolute value of Aligning each item in the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the remaining three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix, and a second aligned temporal deformation matrix; Based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix, building change detection is performed through a trained building change detection model to obtain a building change detection result within the target time period.

2. The building change detection method according to claim 1, characterized in that: The training process of the building change detection model is: Acquire a historical synthetic aperture radar remote sensing image, a third optical remote sensing image, and a fourth optical remote sensing image obtained by photographing buildings in a specified area within a preset historical period, wherein the third optical remote sensing image is an optical remote sensing image of a third historical moment when the building has no size change, the fourth optical remote sensing image is an optical remote sensing image of a fourth historical moment when the building has a size change, and the third historical moment and the fourth historical moment are two different moments within the preset historical period; Performing standardization processing on the historical synthetic aperture radar remote sensing image to obtain a third temporal deformation matrix corresponding to the third historical moment and a fourth temporal deformation matrix corresponding to the fourth historical moment; Aligning each item in the third optical remote sensing image, the fourth optical remote sensing image, the third temporal deformation matrix, and the fourth temporal deformation matrix with the remaining three items to obtain a third aligned optical remote sensing image, a fourth aligned optical remote sensing image, a third aligned temporal deformation matrix, and a fourth aligned temporal deformation matrix; Adding building change annotations to the fourth aligned optical remote sensing image based on the third aligned optical remote sensing image, the third aligned temporal deformation matrix, and the fourth aligned temporal deformation matrix to obtain a fifth optical remote sensing image; Performing feature extraction on the third aligned optical remote sensing image through a feature extraction network to obtain a first optical remote sensing feature map; performing feature extraction on the fifth optical remote sensing image through the feature extraction network to obtain a second optical remote sensing feature map; Calculating a first characteristic map by using a difference algorithm according to the first optical remote sensing characteristic map and the second optical remote sensing characteristic map; Calculating a second feature map by using a difference algorithm according to the third aligned temporal deformation matrix and the fourth aligned temporal deformation matrix; Performing weighted fusion on the first feature map and the second feature map to obtain a fused feature map; Inputting the fused feature map into the preset initial building change detection model to obtain a total loss value output by the initial building change detection model; The initial building change detection model is reversely optimized according to the total loss value to obtain the trained building change detection model.

3. The building change detection method according to claim 2, characterized in that: The step of inputting the fused feature map into the preset initial building change detection model to obtain a total loss value output by the initial building change detection model includes: Inputting the fused feature map into the initial building change detection model to obtain a predicted value output by the initial building change detection model; Calculating a basic loss value according to the predicted value, wherein the basic loss value is calculated by an exponential loss function or a cross entropy loss function; The deformation loss value is calculated according to the predicted value by the following formula: ; in, is the deformation loss value, is the total number of feature maps after fusion, For the The true value of the feature map after fusion, For the The predicted value of the feature map after fusion, is the norm; The total loss value is calculated according to the basic loss value and the deformation loss value.

4. The building change detection method according to claim 3, characterized in that: The total loss value is calculated by the following formula: ; ; in, is the iteration parameter, is the current iteration number, is the preset maximum number of iterations, is the total loss value, is the basic loss value.

5. The building change detection method according to claim 1, characterized in that: The step of aligning each item of the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the remaining three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix, and a second aligned temporal deformation matrix comprises: Aligning each item of the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix, and the second temporal deformation matrix with the other three items in terms of longitude and latitude based on the WGS-84 coordinate system to obtain a first initially aligned optical remote sensing image, a second initially aligned optical remote sensing image, a first initially aligned temporal deformation matrix, and a second initially aligned temporal deformation matrix; The first initially aligned optical remote sensing image, the second initially aligned optical remote sensing image, the first initially aligned temporal deformation matrix, and the second initially aligned temporal deformation matrix are cropped in the unaligned area to obtain the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix, and the second aligned temporal deformation matrix.

6. The building change detection method according to claim 1, characterized in that: Acquiring a first optical remote sensing image and a second optical remote sensing image in a target area during a target period of time, comprising: Acquire a first original optical remote sensing image and a second original optical remote sensing image, wherein the first original optical remote sensing image is an original optical remote sensing image of the target area at a first historical moment, and the second original optical remote sensing image is an original optical remote sensing image of the target area at a second historical moment; Performing geometric correction on the first original optical remote sensing image to obtain a first geometrically corrected optical remote sensing image; performing geometric correction on the second original optical remote sensing image to obtain a second geometrically corrected optical remote sensing image; Performing radiation correction on the first geometrically corrected optical remote sensing image to obtain a first radiation-corrected optical remote sensing image; performing radiation correction on the second geometrically corrected optical remote sensing image to obtain a second radiation-corrected optical remote sensing image; Performing atmospheric correction on the first radiation-corrected optical remote sensing image to obtain a first atmospheric-corrected optical remote sensing image; performing atmospheric correction on the second radiation-corrected optical remote sensing image to obtain a second atmospheric-corrected optical remote sensing image; The first atmospherically corrected optical remote sensing image is geocoded to obtain the first optical remote sensing image; and the second atmospherically corrected optical remote sensing image is geocoded to obtain the second optical remote sensing image.

7. A building change detection system, characterized in that: The building change detection system comprises: A data acquisition module, used to acquire a synthetic aperture radar remote sensing image, a first optical remote sensing image, and a second optical remote sensing image obtained by photographing a building in a target area within a target period, wherein the first optical remote sensing image is an optical remote sensing image of the building at a first historical moment, and the second optical remote sensing image is an optical remote sensing image of the building at a second historical moment, and the first historical moment and the second historical moment are two different moments within the target period; A standardization processing module is used to perform standardization processing on the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment; the standardization processing on the synthetic aperture radar remote sensing image to obtain a first time series deformation matrix corresponding to the first historical moment and a second time series deformation matrix corresponding to the second historical moment includes: Determining the height and width of the synthetic aperture radar remote sensing image; Preprocessing the synthetic aperture radar remote sensing image to obtain a preprocessed synthetic aperture radar remote sensing image; Calculate the temporal deformation matrix of the target period by using the interference superposition technology on the pre-processed synthetic aperture radar remote sensing image; Based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the first historical moment and the preset number before it to obtain a first initial temporal deformation matrix, and based on the temporal deformation matrix of the target time period, extracting the temporal deformation results of the second historical moment and the preset number before it to obtain a second initial temporal deformation matrix; According to the height and width of the synthetic aperture radar remote sensing image and the preset number, the first initial time-series deformation matrix is ​​subjected to matrix transformation to obtain a first transformed time-series deformation matrix, and according to the height and width of the synthetic aperture radar remote sensing image and the preset number, the second initial time-series deformation matrix is ​​subjected to matrix transformation to obtain a second transformed time-series deformation matrix; The first temporal deformation matrix is ​​calculated according to the first transformation temporal deformation matrix by the following formula: ; in, is the value of the first temporal deformation matrix, is the value of the first transformation time series deformation matrix, for The absolute value of is the preset deformation threshold; The second temporal deformation matrix is ​​calculated according to the second transformation temporal deformation matrix: ; in, is the value of the second temporal deformation matrix, is the value of the second transformation time series deformation matrix, for The absolute value of A geographic location alignment module, used for aligning each item of the first optical remote sensing image, the second optical remote sensing image, the first temporal deformation matrix and the second temporal deformation matrix with the other three items to obtain a first aligned optical remote sensing image, a second aligned optical remote sensing image, a first aligned temporal deformation matrix and a second aligned temporal deformation matrix; A building change detection module is used to perform building change detection based on the first aligned optical remote sensing image, the second aligned optical remote sensing image, the first aligned temporal deformation matrix and the second aligned temporal deformation matrix through a trained building change detection model to obtain the building change detection result within the target time period.

8. A building change detection device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute a building change detection method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute a building change detection method as described in any one of claims 1 to 6.