Two-stage graph patch pushing determination method and tool for remote sensing change detection

Abstract

The embodiment of the present disclosure discloses a two-stage graph spot pushing determination method and tool for remote sensing change detection, and belongs to the technical field of change graph spot pushing. The specific implementation of the method comprises: acquiring remote sensing high-resolution image data, detecting the remote sensing high-resolution image data, and determining an initial first change detection graph spot; preprocessing the initial first change detection graph spot to determine an initial first change graph spot pushing frame; based on a preset detection algorithm, acquiring images within the initial first change graph spot pushing frame, and extracting an initial second change graph spot; based on a preset rule, processing the initial second change graph spot to determine a second change graph spot, and finally merging the initial first change detection graph spot and the second change graph spot for pushing. The embodiment can reduce the scope and workload of manual operation, reduce the gap with actual demand, and thus improve the operation efficiency of change detection.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of change patch pushing, and in particular to a two-level patch pushing determination method and tool for remote sensing change detection. BACKGROUND

[0002] With the wide application of artificial intelligence and deep learning technology in the field of remote sensing, high-resolution remote sensing image interpretation is gradually developing towards automation and intelligence. High-resolution remote sensing image interpretation is divided into two directions of typical feature extraction and change detection. After the efforts of researchers in recent years, great progress has been made in interpretation, and high interpretation accuracy can be achieved. However, the interpretation result accuracy still cannot meet the business requirements and is difficult to be used in actual work. Especially in the direction of change detection, multi-temporal images are used, and the change result is greatly different from the actual requirement due to the influence of factors such as homogeneity and change rules, which cannot improve the work efficiency. SUMMARY

[0003] Therefore, the embodiments of the present disclosure provide a two-level patch pushing determination method and tool for remote sensing change detection to solve the problem that the change result obtained in the direction of change detection is greatly different from the requirement in the prior art.

[0004] In a first aspect, the embodiments of the present disclosure provide a two-level patch pushing determination method for remote sensing change detection, comprising: acquiring remote sensing high-resolution image data, detecting the remote sensing high-resolution image data to determine an initial first-level change detection patch; preprocessing the initial first-level change detection patch to determine an initial first-level change patch pushing frame; acquiring images within the initial first-level change patch pushing frame based on a preset detection algorithm to perform initial second-level change patch extraction; processing the initial second-level change patch based on a preset rule to determine a second-level change patch, and finally merging the initial first-level change detection patch and the second-level change patch for pushing.

[0005] In some optional implementations of some embodiments, the remote sensing high-resolution image data, and the detection of the remote sensing high-resolution image data to determine the initial first-level change detection patch, comprise: acquiring first-period remote sensing high-resolution image data and second-period remote sensing high-resolution image data, performing radiation correction on the first-period remote sensing high-resolution image data and the second-period remote sensing high-resolution image data to obtain first-period target image data and second-period target image data; determining the correlation coefficient of the first-period target image data and the second-period target image data according to a preset formula; and determining the initial first-level change detection patch in the first-period target image data and the second-period target image data according to the correlation coefficient.

[0006] In some optional implementations of some embodiments, the preprocessing of the initial primary change detection image patch, determining the initial primary change patch push box, includes: determining an image pixel threshold, simplifying the initial primary change detection image patch according to the image pixel threshold, and determining a target change detection image patch; drawing at least two original image push boxes based on the initial primary change detection image patch, and determining the initial primary change patch push box according to the at least two original image push boxes.

[0007] In some optional implementations of some embodiments, the drawing of the at least two original image push boxes based on the initial primary change detection image patch, and the determination of the initial primary change patch push box according to the at least two original image push boxes, includes: determining a merging distance threshold, determining a merged original image push box based on the vertices of the at least two original image push boxes; and merging the merged original image push box to determine the initial primary change patch push box.

[0008] In some optional implementations of some embodiments, the initial secondary change patch extraction based on the preset detection algorithm on the image in the initial primary change patch push box includes: cutting a potential change area based on the initial primary change patch push box; inputting the potential change area into a deep learning change detection model to extract difference features and determine an initial secondary change patch.

[0009] In some optional implementations of some embodiments, the processing of the initial secondary change patch based on the preset rule to determine a secondary change patch, and the final merging of the initial primary change detection image patch and the secondary change patch for pushing includes: superimposing the initial secondary change patch and the initial primary change detection image patch to obtain superimposed image data; extracting image edge information in the superimposed image data based on an edge detection algorithm; performing interpolation processing on the image edge information to determine a secondary change patch; and merging the initial primary change detection image patch and the secondary change patch for pushing.

[0010] In some optional implementations of some embodiments, the method further includes: determining a recall rate of the initial secondary change patch, and determining a precision rate of the secondary change patch; determining an evaluation result according to the precision rate and the recall rate, and transmitting the evaluation result to a target management platform.

[0011] In a second aspect, the embodiment of the present disclosure provides a two-stage graph patch pushing determination tool for remote sensing change detection. The device comprises: an acquisition detection unit configured to acquire remote sensing high-resolution image data, detect the remote sensing high-resolution image data, and determine an initial first-stage change detection graph patch; a preprocessing unit configured to preprocess the initial first-stage change detection graph patch, and determine an initial first-stage change graph patch pushing frame; an extraction unit configured to acquire images within the initial first-stage change graph patch pushing frame based on a preset detection algorithm, and perform initial second-stage change graph patch extraction; and a determination unit configured to process the initial second-stage change graph patch based on a preset rule, determine a second-stage change graph patch, and finally merge the initial first-stage change detection graph patch and the second-stage change graph patch, and perform pushing.

[0012] In a third aspect, the embodiment of the present disclosure provides a computer readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above method.

[0013] The above-mentioned embodiment of the present disclosure has the following beneficial effects: first, remote sensing high-resolution image data is acquired, the remote sensing high-resolution image data is detected, and an initial first-stage change detection graph patch is determined; then, the initial first-stage change detection graph patch is preprocessed, and an initial first-stage change graph patch pushing frame is determined; then, images within the initial first-stage change graph patch pushing frame are acquired based on a preset detection algorithm, and initial second-stage change graph patch extraction is performed; finally, the initial second-stage change graph patch is processed based on a preset rule, a second-stage change graph patch is determined, and the initial first-stage change detection graph patch and the second-stage change graph patch are finally merged, and pushing is performed, thereby reducing the scope and amount of manual work, reducing the gap with actual demand, and improving the work efficiency of change detection. BRIEF DESCRIPTION OF DRAWINGS

[0014] The above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements and elements are not necessarily drawn to scale.

[0015] Figure 1 is a schematic diagram of one application scenario of a two-stage graph patch pushing determination method for remote sensing change detection according to some embodiments of the present disclosure;

[0016] Figure 2 is a flowchart of the two-stage graph patch pushing determination method for remote sensing change detection according to some embodiments of the present disclosure;

[0017] Figure 3is a structural schematic diagram of some embodiments of a two-level map patch pushing determination device for remote sensing change detection according to the present disclosure;

[0018] Figure 4 is a structural schematic diagram of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0019] In the following description, specific details are set forth, such as a particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present disclosure. However, persons skilled in the art will understand that the present disclosure can be implemented in other embodiments that deviate from these specific details. In other instances, well-known systems, devices, circuits, and methods have not been described in detail in order to avoid unnecessary detail to make the description of the present disclosure more clear.

[0020] A two-level map patch pushing determination method and tool for remote sensing change detection according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0021] Figure 1 is a schematic diagram of an application scenario of a two-level map patch pushing determination method for remote sensing change detection according to some embodiments of the present disclosure.

[0022] In Figure 1 In the application scenario of, first, the computing device 101 can obtain remote sensing high-resolution image data 102, detect the remote sensing high-resolution image data 102, and determine an initial first-level change detection map patch 103. Then, the computing device 101 can preprocess the initial first-level change detection map patch 103 and determine an initial first-level change map patch pushing frame 104. After that, the computing device 101 can obtain the image within the initial first-level change map patch pushing frame 104 based on a preset detection algorithm 105, and perform initial second-level change map patch 106 extraction. Finally, the computing device 101 can process the initial second-level change map patch 106 based on a preset rule 107, determine a second-level change map patch 108, and finally merge the initial first-level change detection map patch 103 and the second-level change map patch 108 for pushing.

[0023] When the computing device body 101 is software, it can be installed in the hardware devices listed above. It can be implemented as a plurality of software or software modules, for example, to provide distributed services, or as a single software or software module. No specific limitation is made here.

[0024] It should be understood that Figure 1 The number of computing devices 101 in is only illustrative. Depending on the implementation needs, there can be any number of computing devices 101.

[0025] Figure 2This is a flowchart illustrating some embodiments of the two-level patch push determination method for remote sensing change detection according to this disclosure. Figure 2 A two-level patch push determination method for remote sensing change detection can be derived from... Figure 1 The computing device 101 performs the operation. For example... Figure 2 As shown, this two-level patch push determination method for remote sensing change detection includes the following steps:

[0026] Step S201: Acquire high-resolution remote sensing image data, perform detection on the aforementioned high-resolution remote sensing image data, and determine the initial first-level change detection patches.

[0027] In some embodiments, the execution entity of the two-level patch push determination method for remote sensing change detection (such as...) Figure 1 The computing device 101 shown can acquire remote sensing high-resolution image data through the following steps, and detect the aforementioned remote sensing high-resolution image data to determine the initial first-level change detection patches:

[0028] The first step involves the aforementioned implementing entity acquiring high-resolution remote sensing imagery data from both the first and second periods. Radiometric correction is then performed on these two sets of data to obtain target imagery data from the first and second periods. Here, the first and second periods of high-resolution remote sensing imagery data represent data from the same region at two different times. To ensure the accuracy of the change detection patches, the sensors used to acquire the two sets of high-resolution remote sensing imagery data are the same or have similar resolutions.

[0029] In the second step, the aforementioned executing entity can determine the correlation coefficient between the target image data of the first period and the target image data of the second period according to a preset formula.

[0030] Third, the aforementioned executing entity can determine the initial first-level change detection patches in the target image data of the first period and the target image data of the second period based on the aforementioned correlation coefficient.

[0031] Specifically, the aforementioned implementing entity can perform relative radiometric correction on the two sets of high-resolution remote sensing image data X and Y using a histogram specification method, for the spectral vectors at the same location in X and Y. and First, standardization is required. Then, the minimum correlation coefficient between the pixels of the two images X and Y is calculated according to the above formula. , where θ is the projection matrix. From the above linear relationship, we can obtain The algorithm determines whether a pixel has changed based on α, thereby obtaining an initial first-level change detection patch at the pixel level.

[0032] In step S202, the initial primary change detection image spot is preprocessed to determine the initial primary change image spot push box.

[0033] In some embodiments, the execution subject can preprocess the initial primary change detection image spot to determine the initial primary change image spot push box. Here, the execution subject can determine an image pixel threshold, simplify the initial primary change detection image spot according to the image pixel threshold, and determine a target change detection image spot. Then, the execution subject can draw at least two original image push boxes based on the initial primary change detection image spot, determine the initial primary change image spot push box according to the at least two original image push boxes. In actual application, small area image spots in the pixel-level change detection image spot are removed, the minimum circumscribed rectangle of the remaining pixel-level change image spot is calculated, and rectangles with similar distances are merged to generate the initial primary change image spot push box with positioning function.

[0034] Specifically, the pixel-level change is converted into a planar vector image spot, and the change area is calculated by using the vector image spot. The vectors with an area smaller than a certain threshold are removed. For the remaining vector image spots , the four boundary coordinates of the vector image spot are calculated one by one , and then the minimum circumscribed rectangle is drawn with the four coordinates as the vertices , , , . The merging distance r of the rectangle is set, one vertex of each rectangle is taken as the coordinate, and the rectangle is searched with r as the radius. All searched rectangles are merged, and the small rectangles after merging are deleted, thereby generating the initial primary change image spot push box.

[0035] In some optional implementations of some embodiments, determining the initial primary change image spot push box according to the at least two original image push boxes includes determining a merging distance threshold, determining a merged original image push box based on the vertices of the at least two original image push boxes, and merging the merged original image push box to determine the initial primary change image spot push box.

[0036] In step S203, the image in the initial primary change image spot push box is obtained based on a preset detection algorithm to extract an initial secondary change image spot.

[0037] In some embodiments, the execution subject can obtain the image in the initial first-level change patch pushing box based on a preset detection algorithm, and then the execution subject can perform initial second-level change patch extraction to obtain the initial second-level change patch. Specifically, the execution subject can intercept a potential change area based on the initial first-level change patch pushing box; then the execution subject can input the potential change area into a deep learning change detection model to perform difference feature extraction and determine the initial second-level change patch. Here, the potential change area in the remote sensing high-resolution image data of two periods is intercepted by the initial first-level change patch pushing box, and then the intercepted image is input into a deep learning change detection algorithm with a twin structure to perform deep difference feature extraction, extract a more accurate change detection result, and output a vector result with accurate boundaries.

[0038] In step S204, the initial second-level change patch is processed based on a preset rule to determine a second-level change patch, and the initial first-level change detection patch and the second-level change patch are finally merged and pushed.

[0039] In some embodiments, the execution subject can superimpose the initial second-level change patch and the initial first-level change detection patch to obtain superimposed image data. Then, the execution subject can extract image edge information in the superimposed image data based on an edge detection algorithm (for example, a Canny edge detection algorithm). After that, the execution subject can perform interpolation processing on the image edge information to determine a second-level change patch. Finally, the execution subject can merge the initial first-level change detection patch and the second-level change patch and push them.

[0040] In some optional implementations of some embodiments, the method further includes determining the recall rate of the initial second-level change patch and determining the precision rate of the second-level change patch; determining an evaluation result according to the precision rate and the recall rate, and transmitting the evaluation result to a target management platform. Specifically, the execution subject can perform precision evaluation on the initial second-level change patch and the second-level change patch, count the number of change patch detections according to the true value of the change, calculate the recall rate of the initial second-level change patch and the precision rate of the second-level change patch, and then calculate a score as an evaluation result according to the following formula:

[0041] ,

[0042] wherein Recall represents the recall rate of the initial second-level change patch; Precision represents the precision rate of the second-level change patch; TS represents the number of detected true value patches; SS represents the total number of true value patches; JS represents the total number of detected changes; and F1 represents the score as the evaluation result.

[0043] One of the above-mentioned various embodiments of the present disclosure has the following beneficial effects: first, remote sensing high-resolution image data is acquired, the remote sensing high-resolution image data is detected to determine initial first-level change detection patches; then, the initial first-level change detection patches are preprocessed to determine initial first-level change patch push boxes; thereafter, based on a preset detection algorithm, images within the initial first-level change patch push boxes are acquired to extract initial second-level change patches; finally, based on a preset rule, the initial second-level change patches are processed to determine second-level change patches, and the initial first-level change detection patches and the second-level change patches are finally merged and pushed, thereby reducing the scope and workload of manual work, reducing the gap with actual needs, and thereby improving the work efficiency of change detection.

[0044] All the above-mentioned optional technical solutions can be combined to form optional embodiments of the present application, which will not be described here.

[0045] The following is an embodiment of the device of the present disclosure, which can be used to execute the method embodiments of the present disclosure. For details not disclosed in the device embodiments of the present disclosure, please refer to the method embodiments of the present disclosure.

[0046] Figure 3 is a structural schematic diagram of some embodiments of a two-level patch push determination tool for remote sensing change detection according to the present disclosure. As shown in Figure 3 the two-level patch push determination tool for remote sensing change detection includes an acquisition and detection unit 301, a processing unit 302, an extraction unit 303, and a determination unit 304. The acquisition and detection unit 301 is configured to acquire remote sensing high-resolution image data, detect the remote sensing high-resolution image data, and determine initial first-level change detection patches. The preprocessing unit 302 is configured to preprocess the initial first-level change detection patches to determine initial first-level change patch push boxes. The extraction unit 303 is configured to acquire images within the initial first-level change patch push boxes based on a preset detection algorithm to extract initial second-level change patches. The determination unit 304 is configured to process the initial second-level change patches based on a preset rule to determine second-level change patches, and finally merge the initial first-level change detection patches and the second-level change patches for pushing. Here, the two-level patch push determination tool for remote sensing change detection refers to a control or virtual device that implements a two-level patch push determination method for remote sensing change detection.

[0047] In some optional implementations of some embodiments, the acquisition detection unit 301 of the two-stage graph push determination tool for remote sensing change detection is further configured to: acquire first period remote sensing high-resolution image data and second period remote sensing high-resolution image data, perform radiation correction on the first period remote sensing high-resolution image data and the second period remote sensing high-resolution image data to obtain first period target image data and second period target image data; determine a correlation coefficient of the first period target image data and the second period target image data according to a preset formula; and determine an initial first-stage change detection graph in the first period target image data and the second period target image data according to the correlation coefficient.

[0048] In some optional implementations of some embodiments, the preprocessing unit 302 of the two-stage graph push determination tool for remote sensing change detection is further configured to: determine an image pixel threshold, simplify the initial first-stage change detection graph according to the image pixel threshold, and determine a target change detection graph; draw at least two original image push boxes based on the initial first-stage change detection graph, and determine an initial first-stage change graph push box according to the at least two original image push boxes.

[0049] In some optional implementations of some embodiments, the drawing of the at least two original image push boxes based on the initial first-stage change detection graph and the determination of the first-stage change graph push box according to the at least two original image push boxes include: determining a merging distance threshold, determining a merged original image push box based on vertices of the at least two original image push boxes; and merging the merged original image push box to determine the initial first-stage change graph push box.

[0050] In some optional implementations of some embodiments, the extraction unit 303 of the two-stage graph push determination tool for remote sensing change detection is further configured to: cut a potential change area based on the initial first-stage change graph push box; and input the potential change area into a deep learning change detection model to perform difference feature extraction and determine an initial second-stage change graph.

[0051] In some optional implementations of some embodiments, the determination unit 304 of the two-stage graph push determination tool for remote sensing change detection is further configured to: superimpose the initial second-stage change graph and the initial first-stage change detection graph to obtain superimposed image data; extract image edge information in the superimposed image data based on an edge detection algorithm; and perform interpolation processing on the image edge information to determine a second-stage change graph.

[0052] In some optional implementations of some embodiments, the two-level graph patch pushing determination tool for remote sensing change detection is further configured to: determine a recall rate of the initial secondary change graph patch, and determine a precision rate of the initial primary change graph patch; determine an evaluation result according to the precision rate and the recall rate, and transmit the evaluation result to the target management platform.

[0053] It can be understood that the units described in the device correspond to the respective steps in the described method. Therefore, the operations, features and advantages described above for the method also apply to the device and the units contained therein, which will not be described here again. Figure 2 It can be understood that the units described in the device correspond to the respective steps in the described method. Therefore, the operations, features and advantages described above for the method also apply to the device and the units contained therein, which will not be described here again.

[0054] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0055] Figure 4 is a schematic diagram of a computer device 4 provided by the embodiments of the present disclosure. As shown in the figure, the computer device 4 of this embodiment includes a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. The processor 401 implements the steps in each of the above method embodiments when executing the computer program 403. Alternatively, the processor 401 implements the functions of each module / unit in the above device embodiments when executing the computer program 403. Figure 4

[0056] By way of example, the computer program 403 can be segmented into one or more modules / units stored in the memory 402 and executed by the processor 401 to complete the present disclosure. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 403 in the computer device 4.

[0057] The computer device 4 can be a desktop computer, a notebook, a palm computer, and a cloud server, etc. The computer device 4 can include but is not limited to the processor 401 and the memory 402. Those skilled in the art can understand that the computer device 4 is only an example of the computer device 4 and does not constitute a limitation on the computer device 4, and can include more or fewer components than the figure, or combine certain components, or different components, for example, the computer device can also include an input / output device, a network access device, a bus, etc. Figure 4 The computer device 4 is only an example of the computer device 4 and does not constitute a limitation on the computer device 4, and can include more or fewer components than the figure, or combine certain components, or different components, for example, the computer device can also include an input / output device, a network access device, a bus, etc.

[0058] ​The processor 401 can be a central processing unit (CPU), or other general purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, etc. The general purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0059] The memory 402 can be an internal storage unit of the computer device 4, for example, a hard disk or a memory of the computer device 4. The memory 402 can also be an external storage device of the computer device 4, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 402 can also include both the internal storage unit and the external storage device of the computer device 4. The memory 402 is used to store computer programs and other programs and data required by the computer device. The memory 402 can also be used to temporarily store data that has been output or will be output.

[0060] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is taken as an example for illustration, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit, and the integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0061] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0062] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present disclosure.

[0063] In the embodiments provided by the present disclosure, it should be understood that the disclosed apparatus / computer device and method can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely schematic, for example, the division of the modules or units is only a logical function division, and there can be another division manner in actual implementation, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling or direct coupling or communication connection between each other can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0064] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0065] In addition, each functional unit in each embodiment of the present disclosure can be integrated into a processing unit, or each unit can exist physically independently, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0066] The integrated modules / units, if implemented in the form of software functional units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various method embodiments can be implemented. The computer program can include computer program code, which can be in the form of source code, object code, executable files or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), electric carrier wave signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electric carrier wave signal and telecommunication signal.

[0067] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than limit them; although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in the protection scope of the present disclosure.

Claims

1. A two-level patch push and determination method for remote sensing change detection, characterized in that, include: Acquire high-resolution remote sensing image data, perform detection on the high-resolution remote sensing image data, and determine the initial first-level change detection patches; Preprocessing the initial first-level change detection patch includes: determining an image pixel threshold, simplifying the initial first-level change detection patch according to the image pixel threshold, and determining the target change detection patch; At least two original image push boxes are drawn based on the initial first-level change detection patches, and the initial first-level change patch push box is determined based on the at least two original image push boxes; The image within the initial first-level change patch push box is obtained based on a preset detection algorithm, and the initial second-level change patch is extracted. The initial secondary change patches are processed based on preset rules to determine secondary change patches, and finally the initial primary change detection patches and the secondary change patches are merged and pushed.

2. The two-level patch pushing and determination method for remote sensing change detection according to claim 1, characterized in that, The process of acquiring high-resolution remote sensing image data and detecting the high-resolution remote sensing image data to determine initial first-level change detection patches includes: Acquire high-resolution remote sensing image data of the first period and high-resolution remote sensing image data of the second period, and perform radiometric correction on the high-resolution remote sensing image data of the first period and the high-resolution remote sensing image data of the second period to obtain target image data of the first period and target image data of the second period. The correlation coefficient between the target image data of the first period and the target image data of the second period is determined according to a preset formula; The initial first-level change detection patches in the target image data of the first period and the target image data of the second period are determined based on the correlation coefficient.

3. The two-level patch pushing and determination method for remote sensing change detection according to claim 2, characterized in that, The process of drawing at least two original image push boxes based on the initial first-level change detection patches, and determining the initial first-level change patch push box based on the at least two original image push boxes, includes: Determine the merging distance threshold, and determine the merged original image push box based on the vertices of the at least two original image push boxes; The original image push boxes are merged to determine the initial first-level change patch push box.

4. The two-level patch pushing and determination method for remote sensing change detection according to claim 1, characterized in that, The step of obtaining the image within the initial primary change patch push frame based on a preset detection algorithm and extracting the initial secondary change patches includes: Based on the initial first-level change patch push frame, potential change areas are extracted; The potential change regions are input into a deep learning change detection model to extract differential features and determine the initial secondary change patches.

5. The two-level patch pushing and determination method for remote sensing change detection according to claim 1, characterized in that, The process of processing the initial secondary change patches based on preset rules to determine secondary change patches, and finally merging the initial primary change detection patches and the secondary change patches for push, includes: The initial secondary change patch is superimposed with the initial primary change detection patch to obtain superimposed image data; Image edge information is extracted from the overlaid image data based on an edge detection algorithm; Interpolation processing is performed on the image edge information to determine secondary variation patches; The initial first-level change detection patch and the second-level change patch are merged and then pushed.

6. The two-level patch pushing and determination method for remote sensing change detection according to claim 1, characterized in that, The method further includes: Determine the recall of the initial secondary change patch and the precision of the secondary change patch; The evaluation result is determined based on the precision and recall, and the evaluation result is transmitted to the target management platform.

7. A two-level patch push and determination tool for remote sensing change detection, characterized in that, include: The detection unit is configured to acquire remote sensing high-resolution image data, detect the remote sensing high-resolution image data, and determine the initial first-level change detection patches; The preprocessing unit is configured to preprocess the initial first-level change detection patch, including: determining an image pixel threshold, simplifying the initial first-level change detection patch according to the image pixel threshold, and determining a target change detection patch; At least two original image push boxes are drawn based on the initial first-level change detection patches, and the initial first-level change patch push box is determined based on the at least two original image push boxes; The extraction unit is configured to acquire the image within the initial first-level change patch push box based on a preset detection algorithm, and to extract the initial second-level change patches; The determining unit is configured to process the initial secondary change patch based on preset rules, determine the secondary change patch, and finally merge the initial primary change detection patch and the secondary change patch for push.

Images