Remote sensing image processing method, storage medium and electronic device

Abstract

The application discloses a remote sensing image processing method, a storage medium and an electronic device. The method comprises the following steps: interpreting image features of a remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing map patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the map patch regularization processing is used for performing map patch regularization on map patch grids in the first processing result and map patch vectors in the second processing result to adjust a map patch visualization result. The application solves the technical problem of poor regularization and poor visualization effect of the extraction result of the method for directly calling a contour detection algorithm to extract a map patch in the related art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a remote sensing image processing method, a storage medium and an electronic device. BACKGROUND

[0002] With the advent of the era of remote sensing big data, processing of remote sensing images has become one of the important problems in the related field. In the related technology, a deep learning algorithm is mainly used to intelligently interpret the remote sensing image to obtain a ground object classification extraction result. The ground object classification extraction result is raster data, which needs to be converted into vector data to be displayed in the form of a graph patch. The commonly used raster data conversion method at present is to directly call a contour detection algorithm to extract a vector graph patch from the raster data. However, the defect of this method is that the vector graph patch obtained has poor regularity, and problems such as topological error, boundary gap, small patch fragmentation and obvious jagged patch often occur.

[0003] At present, no effective solution has been proposed for the above problems. SUMMARY

[0004] The embodiments of the present application provide a remote sensing image processing method, a storage medium and an electronic device to at least solve the technical problem of poor regularity and poor visualization effect of the extraction result of the method of directly calling a contour detection algorithm to extract a graph patch in the related technology.

[0005] According to an aspect of the embodiments of the present application, a remote sensing image processing method is provided, including: interpreting image features of a remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing graph patch regularization processing based on the first processing result and the second processing result to obtain a target regularity result, wherein the graph patch regularization processing is used to perform graph patch regularization on graph patch raster in the first processing result and graph patch vector in the second processing result to adjust a graph patch visualization result.

[0006] According to another aspect of the embodiments of the present application, a remote sensing image processing method is also provided, including: performing ground object classification interpretation on image features of a target regional type remote sensing image to obtain target regional type raster data; performing data format conversion on the target regional type raster data to obtain target regional type vector data; and performing graph patch regularization processing based on the target regional type raster data and the target regional type vector data to obtain a target regional type regularity result, wherein the graph patch regularization processing is used to perform graph patch regularization on graph patch raster in the target regional type raster data and graph patch vector in the target regional type vector data to adjust a graph patch visualization result.

[0007] According to another aspect of the embodiments of the present application, a remote sensing image processing method is also provided, which comprises: performing ground object classification interpretation on image features of a target water area type remote sensing image to obtain target water area type raster data; performing data format conversion on the target water area type raster data to obtain target water area type vector data; and performing graph patch regularization processing based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on graph patch raster in the target water area type raster data and graph patch vector in the target water area type vector data to adjust a graph patch visualization result.

[0008] According to another aspect of the embodiments of the present application, a computer readable storage medium is also provided, which comprises a stored program, wherein the program, when executed, controls a device in which the computer readable storage medium is located to perform any one of the remote sensing image processing methods described above.

[0009] According to another aspect of the embodiments of the present application, an electronic device is also provided, which comprises: a processor; and a memory connected to the processor and configured to provide the processor with instructions for processing the following processing steps: interpreting image features of a remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing graph patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on graph patch raster in the first processing result and graph patch vector in the second processing result to adjust a graph patch visualization result.

[0010] In the embodiments of the present application, by interpreting image features of a remote sensing image to obtain a first processing result, performing data format conversion on the first processing result to obtain a second processing result, and further performing graph patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on graph patch raster in the first processing result and graph patch vector in the second processing result to adjust a graph patch visualization result.

[0011] It is easy to note that, by interpreting image features of a remote sensing image, performing data format conversion and graph patch regularization processing, the embodiments of the present application obtain a graph patch regularization result corresponding to the remote sensing image, which achieves the purpose of performing regularization processing on a ground object classification interpretation result of a remote sensing image to optimize a graph patch visualization result, thereby realizing the technical effect of improving the regularization degree and visualization effect of a graph patch extracted in a ground object classification process of a remote sensing image, and further solving the technical problem of poor regularization degree and poor visualization effect of a graph patch extraction result of a method of directly calling a contour detection algorithm in the related art. BRIEF DESCRIPTION OF DRAWINGS

[0012] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0013] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing the remote sensing image processing method is shown;

[0014] Figure 2 is a remote sensing ground feature classification image according to the prior art;

[0015] Figure 3 is a schematic diagram of an interpretation result of a remote sensing ground feature classification image according to the prior art;

[0016] Figure 4 is a schematic diagram of a vectorization graph patch of a remote sensing ground feature classification image according to the prior art;

[0017] Figure 5 is a schematic diagram of a graph patch topology error according to the prior art;

[0018] Figure 6 is a schematic diagram of another graph patch topology error according to the prior art;

[0019] Figure 7 is a schematic diagram of a graph patch gap according to the prior art;

[0020] Figure 8 is a schematic diagram of a graph patch sawtooth boundary according to the prior art;

[0021] Figure 9 is a flowchart of a remote sensing image processing method according to an embodiment of the application;

[0022] Figure 10 is a schematic diagram of an optional graph patch topology error according to an embodiment of the application;

[0023] Figure 11 is a schematic diagram of a result of an optional graph patch topology regularization processing according to an embodiment of the application;

[0024] Figure 12 is a schematic diagram of another optional graph patch topology error according to an embodiment of the application;

[0025] Figure 13 is a schematic diagram of another result of an optional graph patch topology regularization processing according to an embodiment of the application;

[0026] Figure 14 is a schematic diagram of an optional graph patch contour according to an embodiment of the application;

[0027] Figure 15 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0028] Figure 16 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0029] Figure 17 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0030] Figure 18 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0031] Figure 19 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0032] Figure 20 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0033] Figure 21 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0034] Figure 22 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0035] Figure 23 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0036] Figure 24 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0037] Figure 25 is a schematic diagram of another alternative pattern spot outline according to an embodiment of the application;

[0038] Figure 26 is a schematic diagram of an alternative pattern spot boundary anomaly according to an embodiment of the application;

[0039] Figure 27 is a schematic diagram of an alternative pattern spot boundary regularization processing result according to an embodiment of the application;

[0040] Figure 28 is a schematic diagram of an alternative to-be-deleted boundary vector point shape according to an embodiment of the application;

[0041] Figure 29 is a schematic diagram of another alternative to-be-deleted boundary vector point shape according to an embodiment of the application;

[0042] Figure 30is a schematic diagram of another optional result of the patch boundary regularization processing according to an embodiment of the present application;

[0043] Figure 31 is a flow chart of another remote sensing image processing method according to an embodiment of the present application;

[0044] Figure 32 is a flow chart of another remote sensing image processing method according to an embodiment of the present application;

[0045] Figure 33 is a structural schematic diagram of a remote sensing image processing device according to an embodiment of the present application;

[0046] Figure 34 is a structural schematic diagram of an optional remote sensing image processing device according to an embodiment of the present application;

[0047] Figure 35 is a structural schematic diagram of another remote sensing image processing device according to an embodiment of the present application;

[0048] Figure 36 is a structural schematic diagram of an optional remote sensing image processing device according to an embodiment of the present application;

[0049] Figure 37 is a structural schematic diagram of another remote sensing image processing device according to an embodiment of the present application;

[0050] Figure 38 is a structural block diagram of another computer terminal according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the personnel in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by the personnel in the art without creative labor should belong to the protection scope of the present application.

[0052] It should be noted that the terms "first", "second", and the like in the description and claims of the application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0053] First, some nouns or terms that appear in the description of the embodiments of the application are explained as follows:

[0054] Remote sensing feature classification: refers to the classification of each pixel unit of a remote sensing image according to feature types. Feature types mainly include: cultivated land, water area, building, road, structure, forest land, grassland and bare land, etc.

[0055] Remote sensing intelligent interpretation: refers to the process of using deep learning algorithms to analyze and process remote sensing images to extract valuable information in remote sensing images.

[0056] Map patch regularization: the output of remote sensing intelligent interpretation is raster data, which needs to be processed and converted into vector data to be displayed in the form of vector map patches. The process of post-processing the vector map patches to make the visualization results of the vector map patches more concise and natural is called map patch regularization.

[0057] Embodiment 1

[0058] According to the embodiments of the application, a remote sensing image processing method embodiment is also provided. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0059] The method embodiment provided by the embodiment of the application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a remote sensing image processing method is shown. As Figure 1As shown, the computer terminal 10 (or mobile device 10) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0060] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of the present invention, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0061] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the remote sensing image processing method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned remote sensing image processing method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0062] The transmission device 106 is configured to receive or send data via a network. The network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network interface controller (NIC) that can connect to other network devices through a base station to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module that is configured to communicate with the Internet wirelessly.

[0063] The display can be a liquid crystal display (LCD) that is touch screen, for example, which can enable a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0064] It is noted that in some alternative embodiments, the above-mentioned Figure 1 The computer device (or mobile device) can include hardware elements (including circuitry), software elements (including computer code stored on a computer readable medium), or a combination of both hardware and software elements. It should be noted that in some embodiments, the computer device (or mobile device) can be a stand-alone device or can be part of a larger device, such as a mobile phone or a tablet computer. Figure 1 is merely one example of a particular implementation and is intended to illustrate the types of components that can be present in the above-mentioned computer device (or mobile device).

[0065] Figure 2 is a remote sensing feature classification image according to the prior art, Figure 3 is a schematic diagram of an interpretation result of a remote sensing feature classification image according to the prior art, Figure 4 is a schematic diagram of a vectorized plot of a remote sensing feature classification image according to the prior art. As Figure 2 , Figure 3 and Figure 4 In the process of intelligent interpretation of remote sensing images using deep learning algorithms for feature classification, raster data is generated, and further data format conversion of the raster data can obtain a vectorized plot result.

[0066] Specifically, a remote sensing image with a height of H and a width of W (such as the remote sensing feature classification image shown in Figure 2 ) is classified into features, and a gray image with a height of H and a width of W corresponding to the remote sensing image (such as the interpretation result shown in Figure 3 ) can be obtained. The gray image is raster data (the raster value is between 1 and K, and K is the number of feature types of feature classification). The gray image (or interpretation result, raster result) shown in Figure 3 is divided into K binary images according to the feature classification result ( Figure 3 and Figure 4(Not shown in the image), then the contours of the K binary images are extracted to obtain contour vectors. The K sets of contour vectors are then merged to obtain the following: Figure 4 The vectorized patch results are shown.

[0067] However, the above-mentioned remote sensing ground feature classification image processing methods provided by the existing technology have the following drawbacks.

[0068] First, it is prone to topology errors. Figure 5 This is a schematic diagram of a topological error in a patch based on existing technology, such as... Figure 5 As shown, the land cover classification results in raster data format (such as...) Figure 3 The grayscale image shown may contain pixels with a width of 1. After contour extraction, these pixels with a width of 1 will form a line segment with a width of 1. However, vectorized patches do not allow line segments with a width of 1, therefore, the following will occur: Figure 5 The topology error shown. Figure 6 This is a schematic diagram of another type of topological error in a patch, based on existing technology, such as... Figure 6 As shown, for the same category (such as...) Figure 6 The diagonal points of the two parts of the shading line in the middle may also be connected after the contour is extracted and vectorized. If they are two sets of intersecting diagonal points of the same category, they may also be connected by intersecting lines.

[0069] Second, gaps are prone to appear at the boundaries of the image patches. Figure 7 This is a schematic diagram of a patch gap based on existing technology, such as... Figure 7 As shown, the remote sensing land cover classification results in raster format ( Figure 7 In the grid to the left of the middle arrow, adjacent different types ( Figure 7 Prediction results (where grid values ​​of 1 and 2 represent different types) are stored in adjacent pixels. After contour extraction and merging of the prediction results for each type, gaps will appear between the different types of patches, such as... Figure 7 As shown, the diagonal shading patch corresponds to the prediction result with a grid value of 1, and the black patch corresponds to the prediction result with a grid value of 2. The pixels corresponding to the prediction results with a grid value of 1 and the prediction results with a grid value of 2 are adjacent, but the diagonal shading patch and the black patch do not fit together.

[0070] Third, the patterns have many jagged edges and are irregular. Figure 8 This is a schematic diagram of a jagged boundary of a pattern based on existing technology, such as... Figure 8As shown, in the process of directly extracting the contour of the grid format ground feature classification prediction result by using the above remote sensing ground feature classification image processing method provided by the prior art, more sawtooth-shaped patches are prone to appear at the inclined angle points, resulting in that the obtained vectorized patches are not regular enough, and the number of vector points stored in the corresponding vector file is too much, and the file is too large.

[0071] In view of the above defects of the prior art, no effective solution has been proposed so far.

[0072] Under the above operating environment, the present application provides a remote sensing image processing method as shown in Figure 9 Figure 9 is a flowchart of a remote sensing image processing method according to an embodiment of the present application, as shown in Figure 9

[0073] Step S91 interprets the image features of the remote sensing image to obtain a first processing result;

[0074] Step S92 performs data format conversion on the first processing result to obtain a second processing result;

[0075] Step S93 performs patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the patch regularization processing is used to perform patch regularization on the patch grid in the first processing result and the patch vector in the second processing result to adjust the patch visualization result.

[0076] In the embodiment of the present application, the image features of the remote sensing image can be color features, shape features, etc. The interpretation of the image features of the remote sensing image can be ground feature classification intelligent interpretation, ground feature recognition intelligent interpretation, ground feature change detection intelligent interpretation, and ground feature multi-classification change intelligent interpretation, etc. achieved by using a deep learning algorithm. In the following, ground feature classification intelligent interpretation is taken as an example for illustration, for example, the above-mentioned first processing result can be a ground feature classification result in grid format.

[0077] The data format conversion on the above-mentioned first processing result in grid format can obtain the above-mentioned second processing result. The second processing result can be a ground feature classification result in vector data format. The second processing result can be displayed in the form of a vector patch.

[0078] In order to make the visualization of the ground feature classification result more concise and natural, patch regularization processing can be performed based on the above-mentioned first processing result and the above-mentioned second processing result, and then a target regularization result is obtained. The patch regularization processing can be used to perform patch regularization on the patch grid in the first processing result and the patch vector in the second processing result to adjust the patch visualization result.

[0079] ​​Specifically, the first processing result and the second processing result are used for the graph spot regularization processing, to obtain a target regularization result, and other method steps are included, which can be referred to the further introduction of the embodiments of the present application below, and will not be described here.

[0080] In the embodiments of the present application, the image features of the remote sensing image are interpreted to obtain a first processing result, and the first processing result is converted in data format to obtain a second processing result, and the first processing result and the second processing result are further used for the graph spot regularization processing to obtain a target regularization result, wherein the graph spot regularization processing is used to regularize the graph spot grid in the first processing result and the graph spot vector in the second processing result to adjust the graph spot visualization result.

[0081] It is easy to note that, through the embodiments of the present application, the image features of the remote sensing image are interpreted, converted in data format and subjected to the graph spot regularization processing, to obtain the graph spot regularization result corresponding to the remote sensing image, so as to achieve the purpose of regularizing the ground object classification interpretation result of the remote sensing image to optimize the graph spot visualization result, thereby realizing the technical effect of improving the regularization degree and the visualization effect of the graph spot extracted in the ground object classification process of the remote sensing image, and further solving the technical problem that the extraction result of the method of directly calling the contour detection algorithm to extract the graph spot is poor in regularization degree and poor in visualization effect.

[0082] It should be noted that the embodiments of the present application can be applied to, but are not limited to, the remote sensing image ground object classification actual application scene, the remote sensing image change detection actual application scene and the remote sensing image ground object recognition actual application scene. For example, it can also be applied to the following technical fields: meteorological field (for example, cloud extraction, weather forecast, weather warning, etc.); natural resources and ecological environment field (for example, weather forecast, change detection, ecological red line change detection, multi-classification change detection, ground object classification, greenhouse extraction, road network extraction, building extraction, building change detection (satellite, unmanned aerial vehicle), etc.); water conservancy field (for example, water area change detection, greenhouse extraction, water body extraction (optical, radar), forest extraction, net cage culture extraction, sand mining field extraction, riverfront house extraction, river dam extraction, photovoltaic power plant extraction, etc.); agriculture and forestry field (for example, crop extraction (wheat, rice, potato, etc.), unmanned aerial vehicle crop recognition (corn, flue-cured tobacco, adlay, etc.), land identification, growth monitoring (index calculation), agricultural yield estimation, pest monitoring, planting suggestion pushing, etc.); secondary disaster field (for example, disaster monitoring, disaster warning, etc.); life service (travel, takeout, logistics) field (for example, travel path planning, travel suggestion pushing, personnel mobilization, price adjustment, etc.); urban planning field (for example, road network extraction (satellite, unmanned aerial vehicle), building extraction, building change detection (satellite, unmanned aerial vehicle), fire fighting, etc.).

[0083] The above method of the embodiment of the present application is further described below.

[0084] In an optional embodiment, in step S93, the polygon regularization processing is performed based on the first processing result and the second processing result to obtain a target regularized result, including the following method steps:

[0085] In step S931, polygon topology regularization processing is performed based on the first processing result to obtain a first regularized result;

[0086] In step S932, polygon boundary regularization processing is performed based on the second processing result to obtain a second regularized result;

[0087] In step S933, the first regularized result and the second regularized result are used to determine the target regularized result.

[0088] In the above optional embodiment, the polygon topology regularization processing based on the first processing result can be performed to solve the problem of polygon topology error in the first processing result, and then the first regularized result is obtained. The problem of polygon topology error can include, but is not limited to, the following: a line segment with a pixel width of 1, a line connecting focus points of the same category, and a cross line connecting multiple groups of focus points of the same category.

[0089] In the above optional embodiment, the polygon boundary regularization processing based on the second processing result can be performed to solve the problem of polygon boundary abnormality in the second processing result, and then the second regularized result is obtained. The problem of polygon boundary abnormality can include, but is not limited to, the following: adjacent polygon boundaries do not fit and have gaps, polygon boundaries have many sawteeth and are irregular, etc.

[0090] The first regularized result and the second regularized result can be used to determine the target regularized result. The target regularized result can be a vector polygon visualization regularized result corresponding to the remote sensing image.

[0091] It is easy to understand that through the polygon topology regularization processing and the polygon boundary regularization processing, the target regularized result without topology error, with boundary fitting, and with high vector polygon regularity can be obtained, and then a remote sensing image map classification result with higher quality can be generated.

[0092] In an optional embodiment, in step S931, the polygon topology regularization processing is performed based on the first processing result to obtain a first regularized result, including the following method steps:

[0093] In step S9311, a target raster polygon is obtained from the first processing result, wherein the target raster polygon includes a first raster polygon and a second raster polygon, the first raster polygon is a raster polygon with a pixel width of a preset value, and the second raster polygon is a diagonal point polygon belonging to the same category;

[0094] Step S9312: Perform topological regularization on the target grid patch and its adjacent grid patches to obtain the first regularization result.

[0095] In the optional embodiments described above, the preset value can be a pixel width value for line segments that are restricted from appearing in the vector map, specified by a technician. For example, the preset value can be 1, indicating that line segments with a pixel width of 1 are not allowed to appear in the vector map, otherwise it will lead to a topology error in the map.

[0096] In the above optional embodiments, "same category" refers to the same land cover type determined by land cover classification interpretation. The aforementioned diagonal patch belonging to the same category can be diagonal patch belonging to the same land cover type (with the same corresponding raster values).

[0097] The target grid patch obtained from the first processing result above may include a first grid patch and a second grid patch. The first grid patch may be a grid patch with a pixel width value of a preset value, and the second grid patch may be a diagonal point patch belonging to the same category.

[0098] The aforementioned target raster patch can be used as the patch to be processed in patch topology normalization. In order to eliminate topological errors in the land cover classification results and make the visualization of the land cover classification results more concise and natural, patch topology normalization is performed on the target raster patch and its adjacent raster patches to obtain the first normalization result.

[0099] For example, when performing land cover classification processing on a remote sensing image P1 of a certain location, the method provided in this embodiment of the invention can be used. Land cover classification interpretation is performed on the remote sensing image P1 to obtain a raster-format land cover classification result (equivalent to the first processing result described above). Patches with a pixel width of 1 in the land cover classification result (equivalent to the first raster patch described above) are detected, and these patches are merged with their adjacent patches. This can widen the pixel width of the patch, thereby eliminating topological errors in the patch.

[0100] Figure 10 This is a schematic diagram of an optional topological error of a patch according to an embodiment of the present invention, such as... Figure 10 As shown, in the raster format land cover classification result 1001 of remote sensing image P1, the four grids along the diagonal from the upper right to the lower left correspond to the same land cover type. Correspondingly, in the vector data format land cover classification result 1002 of remote sensing image P1, the patch in the square shaded area is a patch with a pixel width of 1 (denoted as B1). It is easy to see that there are two boundaries in this patch B1 ( Figure 10 The phenomenon of overlapping (white areas in the middle) will result in topological errors in the image.

[0101] Figure 11is a schematic diagram of an optional result of a graph spot topology regularization processing according to an embodiment of the present application. For the graph spot topology regularization processing of the ground object classification results 1001 and 1002 shown in Figure 10 , the graph spot B1 with a pixel width of 1 can be merged with its adjacent graph spots, that is, the adjacent grid values of the graph spot B1 (grid value of 1) are adjusted, and the ground object classification results 1101 in the grid format and the corresponding ground object classification results 1102 in the vector data format of the remote sensing image P1 can be obtained as shown in Figure 11 . It can be found that, after the adjustment, the phenomenon of the boundary (white area in the middle) fitting in the ground object classification results 1102 is eliminated, and the graph spot topology error is further eliminated. Figure 11

[0102] In an optional embodiment, in step S9312, the graph spot topology regularization processing is performed on the target grid graph spot and the adjacent grid graph spot to obtain a first regularization result, including the following method steps:

[0103] In step S9313, a first category of the first grid graph spot and a second category of the adjacent grid graph spot are determined.

[0104] In step S9314, based on the first category and the second category, the graph spot topology merging processing is performed on the target grid graph spot and the adjacent grid graph spot to obtain the first regularization result.

[0105] In the above optional embodiment, the adjacent grid graph spot and the above first grid graph spot belong to different ground object types. According to the ground object classification result, the first category of the first grid graph spot and the second category of the adjacent grid graph spot can be determined.

[0106] Based on the first category and the second category, the graph spot topology merging processing is performed on the target grid graph spot and the adjacent grid graph spot, which can avoid the situation of the graph spot appearing cross connection, and the above first regularization result is obtained. Still taking the ground object classification processing of the remote sensing image P1 as an example, the ground object classification interpretation of the remote sensing image P1 is performed to obtain the ground object classification result in the grid format (equivalent to the above first processing result). The graph spot with a pixel width of 1 (equivalent to the above first grid graph spot) and the corner point graph spot (equivalent to the above second grid graph spot) existing in the ground object classification result in the grid format are detected, and the surrounding point graph spots of the corner point graph spot are replaced with other graph spots of the same category to avoid the corner point connection, and the graph spot topology error is further eliminated.

[0107] Figure 12 is another schematic diagram of a graph spot topology error according to another embodiment of the present application, as shown in Figure 12 ​As shown in the ground feature classification result 1201 of the raster format of the remote sensing image P1, the category of the pixel block with a pixel width of 1 is a raster value of 2 (equivalent to the first category described above), and there is a diagonal point pixel block belonging to the same category (the raster value is also 2) in the ground feature classification result 1201. Accordingly, in the ground feature classification result 1202 of the vector data format of the remote sensing image P1, the pixel blocks in the left and right diagonal line shaded areas form a diagonal point pixel block, and the pixel blocks in the upper and lower black areas are equivalent to the adjacent raster pixel blocks (the pixel blocks in the black areas correspond to a raster value of 1, and the category is equivalent to the second category described above). It can be found that there is a cross connection in the ground feature classification result 1202, and at this time, a pixel block topology error occurs.

[0108] Figure 13 FIG. 13 is a schematic diagram of another optional pixel block topology regularization processing result according to an embodiment of the present application. As shown in the ground feature classification result 1201 and 1202, the pixel block topology regularization processing can replace the pixel blocks around the diagonal point pixel block with other pixel blocks of the same category. As shown in the ground feature classification result 1201, the raster values of the four central raster blocks are all adjusted to 2. Accordingly, as shown in the ground feature classification result 1302, the diagonal line shaded pixel blocks near the area with the cross connection shown in the ground feature classification result 1301 are replaced with black pixel blocks. It can be found that, after the adjustment, the cross connection is eliminated in the ground feature classification result 1302, and thus the pixel block topology error is eliminated. Figure 12 Figure 13 Figure 12

[0109] In an optional embodiment, in step S932, pixel block boundary regularization processing is performed based on the second processing result to obtain a second regularization result, including the following method steps:

[0110] In step S9321, pixel block boundary vector point translation processing is performed on the boundary vector points of the pixel blocks of different categories in the second processing result to obtain a third processing result.

[0111] In step S9322, pixel block boundary vector point thinning processing is performed based on the third processing result to obtain the second regularization result.

[0112] In the above optional embodiment, the second processing result can be a ground feature classification result in a vector data format. The pixel blocks of different categories can be pixel blocks belonging to different ground feature types determined through ground feature classification interpretation. In the second processing result, there can be gaps between the pixel blocks of different categories, and therefore, the pixel block boundary vector point translation processing is performed on the boundary vector points of the pixel blocks of different categories in the second processing result to eliminate the gaps between the pixel blocks of different categories, and obtain the third processing result.

[0113] ​​​In the optional embodiment, in the third processing result, there can be many sawtooth edges of the polygon boundary, and thus the polygon boundary vector points of the third processing result are subjected to vector point thinning processing (e.g., deleting some sawtooth angle points on the vector polygon boundary) to make the vector polygon more smooth and regular, and the second regular result is obtained.

[0114] In an optional embodiment, in step S9321, the boundary vector points of the polygons of different categories in the second processing result are subjected to vector point translation processing to obtain the third processing result, including the following method steps:

[0115] In step S9323, the first vector point and the second vector point are obtained from the second processing result, wherein the first vector point is a boundary vector point of a polygon of a first category, and the second vector point is a boundary vector point of a polygon of a second category.

[0116] In step S9324, the boundary vector points of the polygons are subjected to vector point translation processing based on the relative positions between the first vector point and the second vector point to obtain the third processing result, so that the boundary of the polygon of the first category and the boundary of the polygon of the second category are mutually fitted.

[0117] In the optional embodiment, the first category polygon and the second category polygon can be adjacent vector polygons belonging to different ground object types. The first vector point is a boundary vector point of a polygon of a first category, and the second vector point is a boundary vector point of a polygon of a second category.

[0118] When the boundary vector points of the polygons of different categories in the second processing result are subjected to vector point translation processing, the translation processing modes of different boundary vector points can be different, and thus the translation processing mode of the first vector point and / or the second vector point can be determined based on the relative positions between the first vector point and the second vector point, and then the vector point translation processing is performed to obtain the third processing result. In this way, the boundary of the polygon of the first category and the boundary of the polygon of the second category can be mutually fitted, the gap between the polygon of the first category and the polygon of the second category can be eliminated, and the visualization effect of the vector polygon can be improved.

[0119] Still taking the ground object classification processing of a remote sensing image P1 as an example, the remote sensing image P1 is subjected to ground object classification interpretation and data format conversion to obtain a ground object classification result in a vector data format (equivalent to the second processing result). The vector points of the polygon boundary in the ground object classification result in the vector data format are subjected to translation, and adjacent polygons can be fitted.

[0120] Specifically, based on the feature classification result in the vector data format, the relative position of a graph boundary vector point (such as a jagged vector point in the feature classification result) and an adjacent point is traversed, and then the graph boundary vector point meeting a preset condition is translated (such as being translated downward and / or to the right), so that the graph boundary vector point meeting the preset condition is attached to another graph boundary.

[0121] The following gives an example of a part of possible translation modes of the graph boundary vector point, in the following examples involving the accompanying drawings, the black line (except the frame) represents the graph contour edge, the black circle point represents the graph boundary vector point to be translated, and the oblique hatched area represents the graph interior area.

[0122] Figure 14 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 15 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 16 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 17 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 18 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 19 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 20 is a schematic diagram of another optional graph contour according to an embodiment of the present application. When the graph shape and the relative position of the graph boundary vector point and the adjacent point are as shown in any one of the accompanying drawings of Figure 14 to Figure 20 , the graph boundary vector point to be translated is controlled to be translated downward by one step.

[0123] Figure 21 is a schematic diagram of another optional graph contour according to an embodiment of the present application. When the graph shape and the relative position of the graph boundary vector point and the adjacent point are as shown in Figure 21 , the graph boundary vector point to be translated is controlled to be translated downward by one step and to the right by one step.

[0124] Figure 22 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 23 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 24 is a schematic diagram of another optional graph contour according to an embodiment of the present application, Figure 25 is a schematic diagram of another optional graph contour according to an embodiment of the present application. When the graph shape and the relative position of the graph boundary vector point and the adjacent point are as shown in any one of the accompanying drawings of Figure 22 to Figure 25 , the graph boundary vector point to be translated is controlled to be translated to the right by one step.

[0125] It should be noted that after the to-be-translated polygon boundary vector point is translated by one step, the relative position of the to-be-translated polygon boundary vector point and the adjacent point can be continuously judged, and the next step of translation can be performed until the to-be-translated polygon boundary vector point is in contact with another vector polygon boundary.

[0126] Figure 26 is a schematic diagram of an optional polygon boundary abnormality according to an embodiment of the present application, Figure 27 is a schematic diagram of an optional polygon boundary regularization processing result according to an embodiment of the present application. As shown in Figure 26 , in the feature classification result, two adjacent polygon boundaries of different categories are not in contact, and there is a gap (white area in Figure 26 ). As shown in Figure 26 , the feature classification result is subjected to polygon boundary vector point translation processing, and the polygon boundary regularization processing result (equivalent to the third processing result) as shown in Figure 27 is obtained. It can be found that, after the polygon boundary vector point translation processing, the gap between the two adjacent polygon boundaries of different categories in the feature classification result is eliminated.

[0127] In an optional embodiment, in step S9322, based on the third processing result, polygon boundary vector point thinning processing is performed to obtain a second regularization result, including the following method steps:

[0128] Step S9325, a third vector point is obtained from the third processing result, wherein the third vector point is a boundary vector point formed after the boundary of the first category polygon and the boundary of the second category polygon are in contact with each other;

[0129] Step S9326, a fourth vector point and a fifth vector point are obtained, wherein the fourth vector point is a vector point adjacent to the third vector point, and the fifth vector point is a vector point adjacent to the third vector point;

[0130] Step S9327, in response to the vector point position of the fourth vector point and the fifth vector point satisfying a first preset condition, and / or in response to the vector point shape of the fourth vector point and the fifth vector point satisfying a second preset condition, the third vector point is deleted from the second processing result to obtain the second regularization result.

[0131] In the above optional embodiment, the third processing result can be a feature classification result in the form of vector data subjected to polygon boundary vector point translation processing. The third vector point obtained from the third processing result can be a boundary vector point formed after the boundary of the first category polygon and the boundary of the second category polygon are in contact with each other.

[0132] The fourth vector point is a previous vector point adjacent to the third vector point on the boundary of the vector plot, and the fifth vector point is a subsequent vector point adjacent to the third vector point on the boundary of the vector plot.

[0133] Based on the fourth vector point and the fifth vector point, the first preset condition and the second preset condition are used to determine whether to delete the third vector point, so as to reduce the boundary sawtooth of the vector plot and obtain the second regularization result.

[0134] Still taking the ground object classification processing of the remote sensing image P1 of a certain place as an example, the vector plot boundary regularization processing result (equivalent to the third processing result) after the vector point translation processing of the plot boundary is detected, and the redundant parallel points and the redundant corner points are deleted, so that the vector plot is more regular and smooth.

[0135] Based on the vector plot boundary regularization processing result after the vector point translation processing of the plot boundary, the boundary vector points are traversed. When traversing, it is checked whether the horizontal coordinates or the vertical coordinates of the current vector point and the two adjacent vector points are the same. If the horizontal coordinates or the vertical coordinates of the current vector point and at least one adjacent vector point (the previous adjacent vector point and / or the subsequent adjacent vector point) are the same, the current vector point is determined as a redundant parallel point, and the current vector point is deleted.

[0136] In addition, when traversing the boundary vector points, the shapes of the two adjacent vector points of the current vector point can also be checked, and the boundary vector points of a specified shape are deleted. Figure 28 is a schematic diagram of an optional boundary vector point to be deleted according to an embodiment of the present application, Figure 29 is a schematic diagram of another optional boundary vector point to be deleted according to an embodiment of the present application, as Figure 28 and Figure 29 As shown in Figure 28 and Figure 29 If the shapes of the current vector point and the two adjacent vector points are as shown in

[0137] It should be noted that, in order to avoid deleting the vector boundary right angle when deleting the redundant corner point, after the current vector point is determined as a redundant corner point, the lengths of the adjacent edges of the current vector point are determined again. If the length of the shortest adjacent edge of the current vector point is less than a preset threshold (4 in this example), the current vector point is deleted.

[0138] Figure 30 is a schematic diagram of another optional plot boundary regularization processing result according to an embodiment of the present application. As shown inFigure 27 The polygon boundary regularization processing result (equivalent to the third processing result) shown after the polygon boundary vector point translation processing is subjected to the polygon boundary vector point thinning processing, and the result is as shown in the following figure. Figure 30 The polygon boundary regularization processing result (equivalent to the second regularization result) shown. It can be easily found that through the polygon boundary vector point thinning processing, the boundary of the vector polygon can be more regularized and smoother, and the visual effect is better.

[0139] The technical scheme provided by the present application can remove topological errors, merge boundary gaps, and correct polygon sawteeth, etc. in the intelligent interpretation result of the ground feature classification of the remote sensing image, thereby improving the visualization effect and usability of the vector polygon.

[0140] It should be noted that the technical scheme provided by the present application can be deployed in an intelligent earth (AI Earth), a proprietary cloud project, and a public cloud project to generate and display the ground feature classification polygon corresponding to the remote sensing image.

[0141] In an optional embodiment, a graphical user interface is provided through a target cloud platform, and the content displayed by the graphical user interface at least partially contains a remote sensing image polygon regularization scene. The remote sensing image processing method further includes the following method steps:

[0142] Step S941, in response to a first touch operation acting on the graphical user interface, selecting a remote sensing image from a storage space provided by the target cloud platform;

[0143] Step S942, in response to a second touch operation acting on the graphical user interface, selecting an artificial intelligence interpretation tab from a plurality of tabs provided by the target cloud platform;

[0144] Step S943, in response to a third touch operation acting on the graphical user interface, selecting a polygon regularization function from the artificial intelligence interpretation tab;

[0145] Step S944, in response to a fourth touch operation acting on the graphical user interface, determining a target polygon regularization processing option from a plurality of candidate polygon regularization processing options provided by the polygon regularization function;

[0146] Step S945, displaying a target regularization result corresponding to the target polygon regularization processing option in the graphical user interface.

[0147] In the optional embodiments described above, the target cloud platform can be an Artificial Intelligence Earth (AI Earth), a proprietary cloud project, and a public cloud project. A graphical user interface can be provided through the target cloud platform. The graphical user interface can display a remote sensing image polygon regularization scenario.

[0148] In the optional embodiments described above, the user can perform a first touch operation on the graphical user interface. The user can touch a "image selection" button or a "to-be-regularized image" button in the graphical user interface, or perform a specified setting touch gesture in a preset image selection touch area, to achieve the following process: selecting a remote sensing image from a storage space provided by the target cloud platform.

[0149] In the optional embodiments described above, the user can also perform a second touch operation on the graphical user interface. The user can touch a "label selection" button, an "AI interpretation" button, or a "confirm" button in the graphical user interface, or point, box, or long-press among a plurality of tab pages displayed in the graphical user interface, to achieve the following process: selecting an artificial intelligence interpretation tab page from the plurality of tab pages provided by the target cloud platform.

[0150] Specifically, the artificial intelligence interpretation tab page selected from the plurality of tab pages provided by the target cloud platform can be a feature classification intelligent interpretation tab page, a feature recognition intelligent interpretation tab page, a feature change detection intelligent interpretation tab page, and a feature multi-classification change intelligent interpretation tab page, etc.

[0151] In the optional embodiments described above, the user can also perform a third touch operation on the graphical user interface. The user can touch a "function selection" button, a "confirm" button, or a "polygon regularization" button in the graphical user interface, to achieve the following process: selecting a polygon regularization function from the artificial intelligence interpretation tab page.

[0152] In the optional embodiments described above, the user can also perform a fourth touch operation on the graphical user interface. The user can touch a "processing option" button or a "confirm" button in the graphical user interface, or point, box, or long-press among a plurality of candidate polygon regularization processing options displayed in the graphical user interface, to achieve the following process: determining a target polygon regularization processing option from a plurality of candidate polygon regularization processing options provided by the polygon regularization function.

[0153] In the optional embodiments described above, after the target polygon regularization processing option is determined in response to the fourth touch operation described above, the selected remote sensing image can be automatically subjected to image regularization processing according to the target polygon regularization processing option, to obtain a target regularization result. Further, the target regularization result corresponding to the target polygon regularization processing option can be displayed to the user in the graphical user interface.

[0154] In particular, the first touch operation, the second touch operation, the third touch operation and the fourth touch operation can be operations of contacting the display screen of the terminal device with a finger and touching the terminal device. The touch operation can include single-point touch and multi-point touch, wherein the touch operation of each touch point can include clicking, long pressing, double pressing, swiping, etc. The first touch operation, the second touch operation, the third touch operation and the fourth touch operation can also be touch operations realized by a mouse, a keyboard or the like input device.

[0155] In summary, the embodiment of the present application focuses on providing a method for regularizing a graph patch corresponding to an intelligent interpretation result of a remote sensing image ground object classification, which can remove topological errors, merge boundary gaps and correct graph patch sawteeth, etc., thereby improving the visualization effect and usability of the graph patch.

[0156] In the above operating environment, the present application provides a remote sensing image processing method as shown in Figure 31 Figure 31 is a flowchart of another remote sensing image processing method according to the embodiment of the present application, as shown in Figure 31

[0157] Step S311: performing ground object classification interpretation on the image features of the target regional type remote sensing image to obtain target regional type raster data;

[0158] Step S312: performing data format conversion on the target regional type raster data to obtain target regional type vector data;

[0159] Step S313: performing graph patch regularization processing based on the target regional type raster data and the target regional type vector data to obtain a target regional type regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on the graph patch raster in the target regional type raster data and the graph patch vector in the target regional type vector data to adjust the graph patch visualization result.

[0160] In the embodiment of the present application, the image features of the target regional type remote sensing image can be color features, shape features, etc. The ground object classification interpretation on the image features of the target regional type remote sensing image can be intelligent interpretation of the target regional type ground object classification by using a deep learning algorithm. The above target regional type raster data can be a target regional type ground object classification result in raster format.

[0161] ​​The target region type raster data in the grid format is subjected to data format conversion to obtain the target region type vector data. The target region type vector data can be a target region type feature classification result in a vector data format. The second processing result can be displayed in the form of a vector polygon.

[0162] To make the visualization of the target region type feature classification result more concise and natural, the polygon regularization processing can be performed based on the target region type raster data and the target region type vector data to obtain a target region type regularization result. The polygon regularization processing can be used to regularize the polygon raster in the target region type raster data and the polygon vector in the target region type vector data to adjust the polygon visualization result.

[0163] In the embodiment of the present application, the image features of the target region type remote sensing image are subjected to feature classification interpretation to obtain the target region type raster data, the target region type raster data is subjected to data format conversion to obtain the target region type vector data, and the polygon regularization processing is performed based on the target region type raster data and the target region type vector data to obtain the target region type regularization result. The polygon regularization processing is used to regularize the polygon raster in the target region type raster data and the polygon vector in the target region type vector data to adjust the polygon visualization result.

[0164] It is easy to note that, by the embodiment of the present application, the image features of the target region type remote sensing image are subjected to feature classification interpretation, data format conversion and polygon regularization processing to obtain the polygon regularization result corresponding to the target region type remote sensing image, so that the feature classification interpretation result of the target region type remote sensing image is subjected to regularization processing to optimize the polygon visualization result, thereby achieving the technical effect of improving the regularization degree and the visualization effect of the polygon extracted in the feature classification process of the target region type remote sensing image, and further solving the technical problem that the extraction result of the polygon extracted by directly calling the contour detection algorithm in the related art has poor regularization degree and poor visualization effect.

[0165] It should be noted that the embodiments of the present application can be applied to, but are not limited to, the actual application scenarios of target regional type remote sensing image ground object classification, the actual application scenarios of target regional type remote sensing image change detection, and the actual application scenarios of target regional type remote sensing image ground object recognition. For example, it can also be applied to the scenes involving target regional type remote sensing images in the following technical fields: meteorological field (for example, cloud layer extraction, weather forecast, weather warning, etc.); natural resources and ecological environment field (for example, weather forecast, change detection, ecological red line change detection, multi-classification change detection, ground object classification, greenhouse extraction, road network extraction, building extraction, building change detection (satellite, unmanned aerial vehicle), etc.); water conservancy field (for example, water area change detection, greenhouse extraction, water body extraction (optical, radar), forest extraction, net cage culture extraction, sand mining field extraction, riverfront house extraction, river dam extraction, photovoltaic power plant extraction, etc.); agriculture and forestry field (for example, crop extraction (wheat, rice, potato, etc.), unmanned aerial vehicle crop recognition (corn, flue-cured tobacco, Job's tears, etc.), land parcel recognition, growth monitoring (index calculation), agricultural yield estimation, pest monitoring, planting suggestion pushing, etc.); secondary disaster field (for example, disaster monitoring, disaster warning, etc.); life service (travel, takeout, logistics) field (for example, travel path planning, travel suggestion pushing, personnel mobilization, price adjustment, etc.); urban planning field (for example, road network extraction (satellite, unmanned aerial vehicle), building extraction, building change detection (satellite, unmanned aerial vehicle), fire fighting, etc.).

[0166] In an optional embodiment, a graphical user interface is provided through the target cloud platform, and the content displayed by the graphical user interface at least partially contains the polygon regularization scene of the target regional type remote sensing image. The remote sensing image processing method further includes the following method steps:

[0167] Step S3141, in response to the first touch operation acting on the graphical user interface, the target regional type remote sensing image is selected from the storage space provided by the target cloud platform;

[0168] Step S3142, in response to the second touch operation acting on the graphical user interface, the artificial intelligence interpretation tab page is selected from the plurality of tab pages provided by the target cloud platform;

[0169] Step S3143, in response to the third touch operation acting on the graphical user interface, the polygon regularization function is selected from the artificial intelligence interpretation tab page;

[0170] Step S3144, in response to the fourth touch operation acting on the graphical user interface, the target polygon regularization processing option is determined from the plurality of candidate polygon regularization processing options provided by the polygon regularization function;

[0171] Step S3145, the target regional type regularization result corresponding to the target polygon regularization processing option is displayed in the graphical user interface.

[0172] In the optional embodiments described above, the target cloud platform can be an Artificial Intelligence Earth (AI Earth), a proprietary cloud project, and a public cloud project. A graphical user interface can be provided through the target cloud platform. The graphical user interface can display a polygon regularization scene of the target regional type remote sensing image.

[0173] In the optional embodiments described above, the user can perform a first touch operation on the graphical user interface. The user can touch a "image selection" button or a "to-be-regularized image" button in the graphical user interface, or perform a specified setting touch gesture in a preset image selection touch area, to achieve the following process: selecting the target regional type remote sensing image from a storage space provided by the target cloud platform.

[0174] In the optional embodiments described above, the user can also perform a second touch operation on the graphical user interface. The user can touch a "label selection" button, an "AI interpretation" button, or a "confirm" button in the graphical user interface, or point, box, or long-press among a plurality of tabs displayed in the graphical user interface, to achieve the following process: selecting an artificial intelligence interpretation tab from the plurality of tabs provided by the target cloud platform.

[0175] In the optional embodiments described above, the user can also perform a third touch operation on the graphical user interface. The user can touch a "function selection" button, a "confirm" button, or a "polygon regularization" button in the graphical user interface, to achieve the following process: selecting a polygon regularization function from the artificial intelligence interpretation tab.

[0176] In the optional embodiments described above, the user can also perform a fourth touch operation on the graphical user interface. The user can touch a "processing option" button or a "confirm" button in the graphical user interface, or point, box, or long-press among a plurality of candidate polygon regularization processing options displayed in the graphical user interface, to achieve the following process: determining a target polygon regularization processing option from the plurality of candidate polygon regularization processing options provided by the polygon regularization function.

[0177] In the optional embodiments described above, after the target polygon regularization processing option is determined in response to the fourth touch operation described above, the selected target regional type remote sensing image can be automatically subjected to image regularization processing according to the target polygon regularization processing option, to obtain a target regularization result. Further, the target regional type regularization result corresponding to the target polygon regularization processing option is displayed to the user in the graphical user interface.

[0178] In particular, the first touch operation, the second touch operation, the third touch operation and the fourth touch operation can be operations of a user touching a display screen of the terminal device with a finger and touching the terminal device. The touch operation can include single-point touch and multi-point touch, wherein the touch operation of each touch point can include clicking, long pressing, double pressing, swiping, etc. The first touch operation, the second touch operation, the third touch operation and the fourth touch operation can also be touch operations realized by a mouse, a keyboard or the like input device.

[0179] In summary, the embodiment of the present application focuses on providing a method for regularizing a graph patch corresponding to an intelligent interpretation result of a target regional type remote sensing image land feature classification, which can remove topological errors, merge boundary gaps and correct graph sawteeth of the graph patch, thereby improving the visualization effect and usability of the graph patch.

[0180] In the above operating environment, the present application provides a remote sensing image processing method as shown in Figure 32 Figure 32 is a flowchart of another remote sensing image processing method according to the embodiment of the present application, which is shown in Figure 32 The remote sensing image processing method comprises the following steps:

[0181] Step S321: performing land feature classification interpretation on the image features of the target water area type remote sensing image to obtain target water area type raster data;

[0182] Step S322: performing data format conversion on the target water area type raster data to obtain target water area type vector data;

[0183] Step S323: performing graph patch regularization processing based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on the graph patch raster in the target water area type raster data and the graph patch vector in the target water area type vector data to adjust the graph patch visualization result.

[0184] In the embodiment of the present application, the image features of the target water area type remote sensing image can be color features, shape features, etc. The land feature classification interpretation on the image features of the target water area type remote sensing image can be intelligent interpretation of the target water area type land feature classification by using a deep learning algorithm. The target water area type raster data can be a target water area type land feature classification result in raster format.

[0185] ​The target water area type raster data in the grid format is subjected to data format conversion to obtain the target water area type vector data. The target water area type vector data can be a target water area type feature classification result in a vector data format. The second processing result can be displayed in the form of a vector polygon.

[0186] To make the visualization of the target water area type feature classification result more concise and natural, the polygon regularization processing can be performed based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result. The polygon regularization processing can be used to regularize the polygon raster in the target water area type raster data and the polygon vector in the target water area type vector data to adjust the polygon visualization result.

[0187] In the embodiment of the present application, the image features of the target water area type remote sensing image are subjected to feature classification interpretation to obtain target water area type raster data, the target water area type raster data is subjected to data format conversion to obtain target water area type vector data, and the polygon regularization processing is performed based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result. The polygon regularization processing is used to regularize the polygon raster in the target water area type raster data and the polygon vector in the target water area type vector data to adjust the polygon visualization result.

[0188] It is easy to note that, by the embodiment of the present application, the image features of the target water area type remote sensing image are subjected to feature classification interpretation, data format conversion and polygon regularization processing to obtain the polygon regularization result corresponding to the target water area type remote sensing image, which achieves the purpose of regularizing the feature classification interpretation result of the target water area type remote sensing image to optimize the polygon visualization result, thereby realizing the technical effect of improving the regularization degree and the visualization effect of the polygon extracted in the feature classification process of the target water area type remote sensing image, and further solving the technical problem that the extraction result of the polygon extracted by directly calling the contour detection algorithm in the related art has poor regularization degree and poor visualization effect.

[0189] It should be noted that the embodiments of the present application can be applied to, but are not limited to, the actual application scenarios of target water type remote sensing image feature classification, the actual application scenarios of target water type remote sensing image change detection, and the actual application scenarios of target water type remote sensing image feature recognition. For example, it can also be applied to the following technical fields related to target water type remote sensing images: meteorological field (for example, cloud extraction, weather forecast, weather warning, etc.); natural resources and ecological environment field (for example, weather forecast, change detection, ecological red line change detection, multi-classification change detection, feature classification, greenhouse extraction, road network extraction, building extraction, building change detection (satellite, unmanned aerial vehicle), etc.); water conservancy field (for example, water area change detection, greenhouse extraction, water body extraction (optical, radar), forest extraction, net cage culture extraction, sand mining field extraction, riverfront house extraction, river dam extraction, photovoltaic power plant extraction, etc.); agriculture and forestry field (for example, crop extraction (wheat, rice, potato, etc.), unmanned aerial vehicle crop recognition (corn, flue-cured tobacco, adlay, etc.), land identification, growth monitoring (index calculation), agricultural yield estimation, pest monitoring, planting suggestion pushing, etc.); secondary disaster field (for example, disaster monitoring, disaster warning, etc.); life service (travel, takeout, logistics) field (for example, travel path planning, travel suggestion pushing, personnel mobilization, price adjustment, etc.); urban planning field (for example, road network extraction (satellite, unmanned aerial vehicle), building extraction, building change detection (satellite, unmanned aerial vehicle), fire fighting, etc.).

[0190] In summary, the focus of the embodiments of the present application is to provide a method for regularizing a graph patch corresponding to an intelligent interpretation result of target water type remote sensing image feature classification, which can perform topological error removal, boundary gap merging and graph patch jaggy correction on the graph patch, thereby improving the visualization effect and usability of the graph patch.

[0191] It should be noted that the technical solutions provided by the present application can be deployed in an artificial intelligence earth (AI Earth), a proprietary cloud project and a public cloud project to generate and display feature classification graph patches corresponding to target water type remote sensing images.

[0192] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the action sequence described, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0193] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software on a general hardware platform as necessary, and of course can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product in essence or in the form of a part that contributes to the prior art, and the computer software product is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk), and includes a plurality of instructions for causing an end device (which can be a mobile phone, a computer, a server, or a network device) to execute the method described in each embodiment of the present application.

[0194] Embodiment 2

[0195] According to the embodiments of the present application, a device embodiment for implementing the above remote sensing image processing method is also provided, Figure 33 is a structural schematic diagram of a remote sensing image processing device according to an embodiment of the present application, as Figure 33 shown, the device comprises an interpretation module 3301, a conversion module 3302, and a processing module 3303, wherein,

[0196] The interpretation module 3301 is configured to interpret the image features of the remote sensing image to obtain a first processing result.

[0197] The conversion module 3302 is configured to perform data format conversion on the first processing result to obtain a second processing result.

[0198] The processing module 3303 is configured to perform graph patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on the graph patch raster in the first processing result and the graph patch vector in the second processing result to adjust the graph patch visualization result.

[0199] Optionally, the processing module 3303 is further configured to perform graph patch topology regularization processing based on the first processing result to obtain a first regularization result, perform graph patch boundary regularization processing based on the second processing result to obtain a second regularization result, and determine the target regularization result by using the first regularization result and the second regularization result.

[0200] Optionally, the processing module 3303 is further configured to obtain a target raster graph patch from the first processing result, wherein the target raster graph patch comprises a first raster graph patch and a second raster graph patch, the first raster graph patch is a raster graph patch with a pixel width of a preset value, and the second raster graph patch is a diagonal point graph patch belonging to the same category; and perform graph patch topology regularization processing on the target raster graph patch and an adjacent raster graph patch of the target raster graph patch to obtain the first regularization result.

[0201] Optionally, the processing module 3303 is further configured to: determine the first category of the first grid patch and the second category of the adjacent grid patches; and based on the first and second categories, perform patch topology merging processing on the target grid patch and the adjacent grid patches to obtain a first regularization result.

[0202] Optionally, the above processing module 3303 is further configured to: perform a translation process on the boundary vector points of different categories of patches in the second processing result to obtain a third processing result; and perform a thinning process on the boundary vector points of patches based on the third processing result to obtain a second regularization result.

[0203] Optionally, the processing module 3303 is further configured to: obtain a first vector point and a second vector point from the second processing result, wherein the first vector point is the boundary vector point of the first category of patch and the second vector point is the boundary vector point of the second category of patch; perform patch boundary vector point translation processing based on the relative position between the first vector point and the second vector point to obtain a third processing result, so that the boundary of the first category of patch and the boundary of the second category of patch fit together.

[0204] Optionally, the processing module 3303 is further configured to: obtain a third vector point from the third processing result, wherein the third vector point is a boundary vector point formed by the boundary of the first category patch and the boundary of the second category patch being attached to each other; obtain a fourth vector point and a fifth vector point, wherein the fourth vector point is the vector point adjacent to the third vector point and the fifth vector point is the vector point adjacent to the third vector point; in response to the vector point positions of the fourth vector point and the fifth vector point satisfying a first preset condition, and / or in response to the vector point shapes of the fourth vector point and the fifth vector point satisfying a second preset condition, delete the third vector point from the second processing result to obtain a second regularization result.

[0205] Optionally, Figure 34 This is a schematic diagram of the structure of an optional remote sensing image processing device according to an embodiment of the present invention, such as... Figure 34 As shown, the device includes Figure 33 In addition to all the modules shown, it also includes: a display module 3304, used to select a remote sensing image from the storage space provided by the target cloud platform in response to a first touch operation on the graphical user interface; to select an artificial intelligence interpretation tab from multiple tabs provided by the target cloud platform in response to a second touch operation on the graphical user interface; to select a patch normalization function from the artificial intelligence interpretation tab in response to a third touch operation on the graphical user interface; to determine a target patch normalization processing option from multiple candidate patch normalization processing options provided by the patch normalization function in response to a fourth touch operation on the graphical user interface; and to display the target normalization result corresponding to the target patch normalization processing option within the graphical user interface.

[0206] It should be noted that the above interpretation module 3301, conversion module 3302 and processing module 3303 correspond to steps S91 to S93 in Embodiment 1, and the three modules have the same instances and application scenarios as the corresponding steps, but are not limited to the above embodiment one. It should be noted that the above modules can run in the computer terminal 10 provided in Embodiment 1 as part of the device.

[0207] In the embodiment of the present application, the image features of the remote sensing image are interpreted to obtain a first processing result, and the first processing result is subjected to data format conversion to obtain a second processing result, and further based on the first processing result and the second processing result, a graph spot regularization processing is performed to obtain a target regularization result, wherein the graph spot regularization processing is used to perform graph spot regularization on the graph spot raster in the first processing result and the graph spot vector in the second processing result to adjust the graph spot visualization result.

[0208] It is easy to note that, through the embodiment of the present application, the image features of the remote sensing image are interpreted, data format conversion and graph spot regularization processing are performed to obtain the graph spot regularization result corresponding to the remote sensing image, which achieves the purpose of regularizing the ground object classification interpretation result of the remote sensing image to optimize the graph spot visualization result, thereby realizing the technical effect of improving the regularization degree and visualization effect of the graph spot extracted in the ground object classification process of the remote sensing image, and further solving the technical problem that the extraction result of the method of directly calling the contour detection algorithm to extract the graph spot has poor regularization degree and poor visualization effect.

[0209] According to the embodiment of the present application, another device embodiment for implementing the above-mentioned remote sensing image processing method is also provided, Figure 35 is a structural schematic diagram of another remote sensing image processing device according to the embodiment of the present application, as Figure 35 shown, the device comprises an interpretation module 3501, a conversion module 3502 and a processing module 3503, wherein,

[0210] The interpretation module 3501 is used for ground object classification interpretation of image features of a target regional type remote sensing image to obtain target regional type raster data;

[0211] The conversion module 3502 is used for data format conversion of the target regional type raster data to obtain target regional type vector data;

[0212] The processing module 3503 is used for graph spot regularization processing based on the target regional type raster data and the target regional type vector data to obtain a target regional type regularization result, wherein the graph spot regularization processing is used to perform graph spot regularization on the graph spot raster in the target regional type raster data and the graph spot vector in the target regional type vector data to adjust the graph spot visualization result.

[0213] Optionally, Figure 36 is a structural schematic diagram of an optional remote sensing image processing device according to an embodiment of the present application, as shown in the figure, in addition to all the modules shown in the figure, the device further comprises: a display module 3504, configured to: in response to a first touch operation on the graphical user interface, select a target regional type remote sensing image from a storage space provided by a target cloud platform; in response to a second touch operation on the graphical user interface, select an artificial intelligence interpretation tab page from a plurality of tab pages provided by the target cloud platform; in response to a third touch operation on the graphical user interface, select a patch regularization function from the artificial intelligence interpretation tab page; in response to a fourth touch operation on the graphical user interface, determine a target patch regularization processing option from a plurality of candidate patch regularization processing options provided by the patch regularization function; and display a target regional type regularization result corresponding to the target patch regularization processing option in the graphical user interface. Figure 36 Figure 35

[0214] It should be noted here that the above-mentioned interpretation module 3501, conversion module 3502 and processing module 3503 correspond to steps S311 to S313 in Embodiment 1, and the three modules have the same instances and application scenarios as the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above-mentioned modules as part of the device can run in the computer terminal 10 provided in Embodiment 1.

[0215] In the embodiment of the present application, by performing feature classification interpretation on the image features of the target regional type remote sensing image, target regional type raster data is obtained, and data format conversion is performed on the target regional type raster data to obtain target regional type vector data, and further based on the target regional type raster data and the target regional type vector data, patch regularization processing is performed to obtain the target regional type regularization result, wherein the patch regularization processing is used to perform patch regularization on the patch raster in the target regional type raster data and the patch vector in the target regional type vector data to adjust the patch visualization result.

[0216] It is easy to note that, by performing feature classification interpretation, data format conversion and patch regularization processing on the image features of the target regional type remote sensing image, the patch regularization result corresponding to the target regional type remote sensing image is obtained, which achieves the purpose of performing regularization processing on the feature classification interpretation result of the target regional type remote sensing image to optimize the patch visualization result, thereby realizing the technical effect of improving the regularization degree and visualization effect of the patches extracted in the feature classification process of the target regional type remote sensing image, and further solving the technical problem that the extraction result of the method of directly calling the contour detection algorithm to extract the patches has poor regularization degree and poor visualization effect.

[0217] ​​According to the embodiment of the present application, a device for implementing the above-mentioned remote sensing image processing method is also provided, Figure 37 is a structural schematic diagram of another remote sensing image processing device according to the embodiment of the present application, as shown in the figure, the device comprises an interpretation module 3701, a conversion module 3702 and a processing module 3703, wherein, Figure 37

[0218] The interpretation module 3701 is used for performing ground object classification interpretation on the image features of the target water area type remote sensing image to obtain target water area type raster data.

[0219] The conversion module 3702 is used for performing data format conversion on the target water area type raster data to obtain target water area type vector data.

[0220] The processing module 3703 is used for performing graph spot regularization processing based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result, wherein the graph spot regularization processing is used for performing graph spot regularization on the graph spot raster in the target water area type raster data and the graph spot vector in the target water area type vector data to adjust the graph spot visualization result.

[0221] It should be noted that the above-mentioned interpretation module 3701, conversion module 3702 and processing module 3703 correspond to steps S321 to S323 in the embodiment 1, the three modules have the same instances and application scenarios as the corresponding steps, but are not limited to the contents disclosed in the above-mentioned embodiment 1. It should be noted that the above-mentioned modules as a part of the device can run in the computer terminal 10 provided in the embodiment 1.

[0222] In the embodiment of the present application, by performing ground object classification interpretation on the image features of the target water area type remote sensing image, target water area type raster data is obtained, data format conversion is performed on the target water area type raster data to obtain target water area type vector data, and further based on the target water area type raster data and the target water area type vector data, graph spot regularization processing is performed to obtain a target water area type regularization result, wherein the graph spot regularization processing is used for performing graph spot regularization on the graph spot raster in the target water area type raster data and the graph spot vector in the target water area type vector data to adjust the graph spot visualization result.

[0223] ​It is easily noticed that, by the embodiment of the present application, the image features of the target water area type remote sensing image are classified and interpreted, data format conversion and map patch regularization processing are performed, the map patch regularization result corresponding to the target water area type remote sensing image is obtained, the purpose of regularizing the ground object classification and interpretation result of the target water area type remote sensing image to optimize the map patch visualization result is achieved, thereby realizing the technical effect of improving the regularization degree and visualization effect of the extracted map patch in the ground object classification process of the target water area type remote sensing image, and further solving the technical problem that the extraction result of the method of directly calling the contour detection algorithm to extract the map patch is poor in regularization degree and visualization effect.

[0224] It should be noted that the preferred embodiments of the present embodiment can refer to the related description in Embodiment 1, which will not be repeated here.

[0225] Embodiment 3

[0226] According to the embodiment of the present application, an embodiment of an electronic device is also provided, which can be any one of the computing devices in the computing device group. The electronic device comprises a processor and a memory, wherein:

[0227] The memory is connected with the processor, and is used to provide the processor with instructions for processing the following processing steps: interpreting the image features of the remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing map patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the map patch regularization processing is used to perform map patch regularization on the map patch raster in the first processing result and the map patch vector in the second processing result to adjust the map patch visualization result.

[0228] In the embodiment of the present application, by interpreting the image features of the remote sensing image to obtain a first processing result, performing data format conversion on the first processing result to obtain a second processing result, and further performing map patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the map patch regularization processing is used to perform map patch regularization on the map patch raster in the first processing result and the map patch vector in the second processing result to adjust the map patch visualization result.

[0229] It is easily noticed that, by the embodiment of the present application, the image features of the target water area type remote sensing image are classified and interpreted, data format conversion and map patch regularization processing are performed, the map patch regularization result corresponding to the target water area type remote sensing image is obtained, the purpose of regularizing the ground object classification and interpretation result of the target water area type remote sensing image to optimize the map patch visualization result is achieved, thereby realizing the technical effect of improving the regularization degree and visualization effect of the extracted map patch in the ground object classification process of the target water area type remote sensing image, and further solving the technical problem that the extraction result of the method of directly calling the contour detection algorithm to extract the map patch is poor in regularization degree and visualization effect.

[0230] It should be noted that the preferred embodiments of the present embodiment can refer to the related description in Embodiment 1, which will not be repeated here.

[0231] Embodiment 4

[0232] The embodiments of the present application can provide a computer terminal, which can be any computer terminal device in a computer terminal group. Alternatively, in the present embodiment, the above-mentioned computer terminal can also be replaced by a terminal device such as a mobile terminal.

[0233] Alternatively, in the present embodiment, the above-mentioned computer terminal can be located in at least one network device of a plurality of network devices of a computer network.

[0234] In the present embodiment, the above-mentioned computer terminal can execute program codes of the following steps in the remote sensing image processing method: interpreting the image features of the remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; performing map patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the map patch regularization processing is used to perform map patch regularization on the map patch raster in the first processing result and the map patch vector in the second processing result to adjust the map patch visualization result.

[0235] Alternatively, Figure 38 is another structural block diagram of a computer terminal according to an embodiment of the present application, as shown in Figure 38 The computer terminal can include one or more (only one is shown in the figure) processors 122, memories 124, and peripheral interfaces 126.

[0236] The memory can be used to store software programs and modules, such as program instructions / modules corresponding to the remote sensing image processing method and device in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implements the above-mentioned remote sensing image processing method. The memory can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory can further include a memory remotely arranged with respect to the processor, which can be connected to the computer terminal through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0237] The processor can call information and application programs stored in the memory through the transmission device to execute the following steps: interpreting image features of the remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing polygon regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the polygon regularization processing is used to perform polygon regularization on polygon grids in the first processing result and polygon vectors in the second processing result to adjust a polygon visualization result.

[0238] Optionally, the processor can further execute program codes of the following steps: performing polygon topology regularization processing based on the first processing result to obtain a first regularization result; performing polygon boundary regularization processing based on the second processing result to obtain a second regularization result; and determining the target regularization result by using the first regularization result and the second regularization result.

[0239] Optionally, the processor can further execute program codes of the following steps: obtaining a target grid polygon from the first processing result, wherein the target grid polygon includes a first grid polygon and a second grid polygon, the first grid polygon is a grid polygon with a pixel width of a preset value, and the second grid polygon is a diagonal point polygon belonging to the same category; and performing polygon topology regularization processing on the target grid polygon and an adjacent grid polygon of the target grid polygon to obtain the first regularization result.

[0240] Optionally, the processor can further execute program codes of the following steps: determining a first category of the first grid polygon and a second category of the adjacent grid polygon; and performing polygon topology merging processing on the target grid polygon and the adjacent grid polygon based on the first category and the second category to obtain the first regularization result.

[0241] Optionally, the processor can further execute program codes of the following steps: performing polygon boundary vector point translation processing on boundary vector points of different category polygons in the second processing result to obtain a third processing result; and performing polygon boundary vector point thinning processing based on the third processing result to obtain the second regularization result.

[0242] Optionally, the processor can further execute program codes of the following steps: obtaining a first vector point and a second vector point from the second processing result, wherein the first vector point is a boundary vector point of a first category polygon, and the second vector point is a boundary vector point of a second category polygon; performing polygon boundary vector point translation processing based on a relative position between the first vector point and the second vector point to obtain the third processing result, so that boundaries of the first category polygon and the second category polygon are mutually fitted.

[0243] Optionally, the processor can further execute program codes of the following steps: obtaining a third vector point from the third processing result, wherein the third vector point is a boundary vector point formed after the boundary of the first category plot and the boundary of the second category plot are mutually attached; obtaining a fourth vector point and a fifth vector point, wherein the fourth vector point is a previous vector point adjacent to the third vector point, and the fifth vector point is a next vector point adjacent to the third vector point; in response to the vector point positions of the fourth vector point and the fifth vector point satisfying a first preset condition, and / or in response to the vector point shapes of the fourth vector point and the fifth vector point satisfying a second preset condition, deleting the third vector point from the second processing result to obtain a second regularized result.

[0244] Optionally, the processor can further execute program codes of the following steps: in response to a first touch operation on the graphical user interface, selecting a remote sensing image from a storage space provided by the target cloud platform; in response to a second touch operation on the graphical user interface, selecting an artificial intelligence interpretation tab page from a plurality of tab pages provided by the target cloud platform; in response to a third touch operation on the graphical user interface, selecting a plot regularization function from the artificial intelligence interpretation tab page; in response to a fourth touch operation on the graphical user interface, determining a target plot regularization processing option from a plurality of candidate plot regularization processing options provided by the plot regularization function; and displaying a target regularized result corresponding to the target plot regularization processing option in the graphical user interface.

[0245] The processor can call information and application programs stored in the memory through the transmission device to execute the following steps: performing feature classification interpretation on the image features of the target regional type remote sensing image to obtain target regional type raster data; performing data format conversion on the target regional type raster data to obtain target regional type vector data; and performing plot regularization processing on the target regional type raster data and the target regional type vector data to obtain a target regional type regularized result, wherein the plot regularization processing is used to regularize the plot raster in the target regional type raster data and the plot vector in the target regional type vector data to adjust the plot visualization result.

[0246] Optionally, the processor can further execute program codes of the following steps: selecting a target regional type remote sensing image from a storage space provided by the target cloud platform in response to a first touch operation acting on the graphical user interface; selecting an artificial intelligence interpretation tab page from a plurality of tab pages provided by the target cloud platform in response to a second touch operation acting on the graphical user interface; selecting a plot regularization function from the artificial intelligence interpretation tab page in response to a third touch operation acting on the graphical user interface; determining a target plot regularization processing option from a plurality of candidate plot regularization processing options provided by the plot regularization function in response to a fourth touch operation acting on the graphical user interface; and displaying a target regional type regularization result corresponding to the target plot regularization processing option in the graphical user interface.

[0247] The processor can call information and application programs stored in the memory through the transmission device to execute the following steps: performing feature classification interpretation on the image features of the target water type remote sensing image to obtain target water type raster data; performing data format conversion on the target water type raster data to obtain target water type vector data; and performing plot regularization processing based on the target water type raster data and the target water type vector data to obtain a target water type regularization result, wherein the plot regularization processing is used to perform plot regularization on plot grids in the target water type raster data and plot vectors in the target water type vector data to adjust the plot visualization result.

[0248] In the embodiment of the present application, the image features of the remote sensing image are interpreted to obtain a first processing result, and the first processing result is subjected to data format conversion to obtain a second processing result, and further based on the first processing result and the second processing result, plot regularization processing is performed to obtain a target regularization result, wherein the plot regularization processing is used to perform plot regularization on plot grids in the first processing result and plot vectors in the second processing result to adjust the plot visualization result.

[0249] It is easy to note that, by interpreting the image features of the remote sensing image, performing data format conversion and plot regularization processing, the plot regularization result corresponding to the remote sensing image is obtained, which achieves the purpose of regularization processing on the feature classification interpretation result of the remote sensing image to optimize the plot visualization result, thereby realizing the technical effect of improving the regularization degree and visualization effect of the plot extracted in the feature classification process of the remote sensing image, and further solving the technical problem that the method of directly calling a contour detection algorithm to extract the plot has poor regularization degree and visualization effect of the extraction result in the related art.

[0250] Those skilled in the art can understand that, Figure 38The structure shown is only schematic, and the computer terminal can also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, or other terminal device. Figure 38 The above electronic device is not limited in structure. For example, the computer terminal can further include more or fewer components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure. Figure 38 Figure 38

[0251] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiments can be completed by instructing the hardware related to the terminal device by a program, and the program can be stored in a computer readable storage medium, which can include a flash disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc.

[0252] According to the embodiments of the present application, an embodiment of a computer readable storage medium is further provided. Optionally, in the present embodiment, the above-mentioned computer readable storage medium can be used to save the program code executed by the remote sensing image processing method provided in Embodiment 1.

[0253] Optionally, in the present embodiment, the above-mentioned computer readable storage medium can be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.

[0254] Optionally, in the present embodiment, the computer readable storage medium is configured to store program code for performing the following steps: interpreting the image features of the remote sensing image to obtain a first processing result; performing data format conversion on the first processing result to obtain a second processing result; and performing map patch regularization processing based on the first processing result and the second processing result to obtain a target regularization result, wherein the map patch regularization processing is used to perform map patch regularization on the map patch raster in the first processing result and the map patch vector in the second processing result to adjust the map patch visualization result.

[0255] Optionally, in the present embodiment, the computer readable storage medium is configured to store program code for performing the following steps: performing map patch topology regularization processing based on the first processing result to obtain a first regularization result; performing map patch boundary regularization processing based on the second processing result to obtain a second regularization result; and determining the target regularization result by using the first regularization result and the second regularization result.

[0256] ​​Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: obtaining a target grid polygon from the first processing result, wherein the target grid polygon comprises a first grid polygon and a second grid polygon, the first grid polygon is a grid polygon with a preset pixel width, and the second grid polygon is a diagonal point polygon belonging to the same category; and performing polygon topology regularization processing on the target grid polygon and an adjacent grid polygon of the target grid polygon to obtain a first regularization result.

[0257] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: determining a first category of the first grid polygon and a second category of the adjacent grid polygon; and performing polygon topology merging processing on the target grid polygon and the adjacent grid polygon based on the first category and the second category to obtain the first regularization result.

[0258] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: performing polygon boundary vector point translation processing on boundary vector points of different category polygons in the second processing result to obtain a third processing result; and performing polygon boundary vector point thinning processing based on the third processing result to obtain a second regularization result.

[0259] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: obtaining a first vector point and a second vector point from the second processing result, wherein the first vector point is a boundary vector point of a first category polygon, and the second vector point is a boundary vector point of a second category polygon; and performing polygon boundary vector point translation processing based on a relative position between the first vector point and the second vector point to obtain a third processing result, so that the boundary of the first category polygon and the boundary of the second category polygon are mutually fitted.

[0260] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: obtaining a third vector point from the third processing result, wherein the third vector point is a boundary vector point formed after the boundary of the first category polygon and the boundary of the second category polygon are mutually fitted; obtaining a fourth vector point and a fifth vector point, wherein the fourth vector point is a previous vector point adjacent to the third vector point, and the fifth vector point is a next vector point adjacent to the third vector point; and in response to the vector point position of the fourth vector point and the fifth vector point satisfying a first preset condition, and / or in response to the vector point shape of the fourth vector point and the fifth vector point satisfying a second preset condition, deleting the third vector point from the second processing result to obtain the second regularization result.

[0261] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: selecting a remote sensing image from a storage space provided by a target cloud platform in response to a first touch operation on the graphical user interface; selecting an artificial intelligence interpretation tab page from a plurality of tab pages provided by the target cloud platform in response to a second touch operation on the graphical user interface; selecting a polygon regularization function from the artificial intelligence interpretation tab page in response to a third touch operation on the graphical user interface; determining a target polygon regularization processing option from a plurality of candidate polygon regularization processing options provided by the polygon regularization function in response to a fourth touch operation on the graphical user interface; and displaying a target regularization result corresponding to the target polygon regularization processing option in the graphical user interface.

[0262] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: performing ground object classification interpretation on image features of a target regional type remote sensing image to obtain target regional type raster data; performing data format conversion on the target regional type raster data to obtain target regional type vector data; and performing polygon regularization processing on the target regional type raster data and the target regional type vector data to obtain a target regional type regularization result, wherein the polygon regularization processing is used to perform polygon regularization on polygon raster in the target regional type raster data and polygon vector in the target regional type vector data to adjust a polygon visualization result.

[0263] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: selecting a target regional type remote sensing image from a storage space provided by a target cloud platform in response to a first touch operation on the graphical user interface; selecting an artificial intelligence interpretation tab page from a plurality of tab pages provided by the target cloud platform in response to a second touch operation on the graphical user interface; selecting a polygon regularization function from the artificial intelligence interpretation tab page in response to a third touch operation on the graphical user interface; determining a target polygon regularization processing option from a plurality of candidate polygon regularization processing options provided by the polygon regularization function in response to a fourth touch operation on the graphical user interface; and displaying a target regional type regularization result corresponding to the target polygon regularization processing option in the graphical user interface.

[0264] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: performing ground object classification interpretation on the image features of the target water area type remote sensing image to obtain target water area type raster data; performing data format conversion on the target water area type raster data to obtain target water area type vector data; and performing graph patch regularization processing based on the target water area type raster data and the target water area type vector data to obtain a target water area type regularization result, wherein the graph patch regularization processing is used to perform graph patch regularization on the graph patch raster in the target water area type raster data and the graph patch vector in the target water area type vector data to adjust the graph patch visualization result.

[0265] The above-mentioned serial numbers of the embodiments of the application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0266] In the above-mentioned embodiments of the application, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0267] In the several embodiments of the present application, it should be understood that the disclosed technology can be implemented in other ways. Of course, the embodiment described above is only a schematic, for example, the division of units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, unit or module, and can be electrical or other forms.

[0268] 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. According to actual needs, some or all of the units can be selected to achieve the purpose of the embodiment.

[0269] In addition, each functional unit in each embodiment of the application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0270] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0271] The above is only the preferred embodiment of the present application, it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A remote sensing image processing method, characterized in that, include: The image features of the remote sensing image are interpreted to obtain the first processing result; The first processing result is converted into a data format to obtain the second processing result; The target grid patch is obtained from the first processing result, wherein the target grid patch includes: a first grid patch and a second grid patch, the first grid patch is a grid patch with a pixel width of a preset value, and the second grid patch is a diagonal point patch belonging to the same category; The first grid patch is merged with its adjacent grid patches to obtain the merged result; In response to the fact that the category of the second grid patch is different from the category of the surrounding point grid patches of the second grid patch, the surrounding point grid patches are replaced with patches of the same category as the second grid patch, and a replacement result is obtained; Based on the merging result and the replacement result, a first regularization result is obtained; Based on the second processing result, the boundary regularization of the patch is performed to obtain the second regularization result; The target regularization result is determined using the first regularization result and the second regularization result, wherein the target regularization result is used to characterize the visualization regularization result of the vector patch corresponding to the remote sensing image.

2. The remote sensing image processing method according to claim 1, characterized in that, Based on the second processing result, the boundary regularization process of the patch is performed to obtain the second regularization result, which includes: The boundary vector points of different categories of patches in the second processing result are translated to obtain the third processing result; Based on the third processing result, the boundary vector points of the patch are thinned to obtain the second regularization result.

3. The remote sensing image processing method according to claim 2, characterized in that, The boundary vector points of different categories of patches in the second processing result are translated to obtain the third processing result, which includes: Obtain a first vector point and a second vector point from the second processing result, wherein the first vector point is the boundary vector point of the first category of patch, and the second vector point is the boundary vector point of the second category of patch; The third processing result is obtained by translating the boundary vector points of the patch based on the relative position between the first vector point and the second vector point, so that the boundary of the first type of patch and the boundary of the second type of patch fit together.

4. The remote sensing image processing method according to claim 3, characterized in that, Based on the third processing result, the boundary vector points of the patch are thinned to obtain the second regularization result, including: A third vector point is obtained from the third processing result, wherein the third vector point is the boundary vector point formed after the boundary of the first category patch and the boundary of the second category patch are attached to each other; Obtain the fourth vector point and the fifth vector point, wherein the fourth vector point is the vector point preceding the third vector point, and the fifth vector point is the vector point following the third vector point; In response to the vector point positions of the fourth vector point and the fifth vector point satisfying a first preset condition, and / or in response to the vector point shapes of the fourth vector point and the fifth vector point satisfying a second preset condition, the third vector point is deleted from the second processing result to obtain the second regularization result.

5. The remote sensing image processing method according to claim 1, characterized in that, A graphical user interface is provided through a target cloud platform, wherein the content displayed by the graphical user interface at least partially includes a normalized scene of remote sensing image patches, and the remote sensing image processing method further includes: In response to a first touch operation applied to the graphical user interface, the remote sensing image is selected from the storage space provided by the target cloud platform; In response to a second touch operation applied to the graphical user interface, an AI-interpreted tab is selected from multiple tabs provided by the target cloud platform; In response to a third touch operation applied to the graphical user interface, the image normalization function is selected from the AI ​​interpretation tab. In response to a fourth touch operation applied to the graphical user interface, a target patch normalization processing option is determined from multiple candidate patch normalization processing options provided by the patch normalization function; The target regularization result corresponding to the target patch regularization processing option is displayed within the graphical user interface.

6. A remote sensing image processing method, characterized in that, include: Land cover classification and interpretation are performed on the image features of remote sensing images of the target region type to obtain raster data of the target region type; The target region type raster data is converted to a different data format to obtain target region type vector data. Obtain target region type raster patches from the target region type raster data, wherein the target region type raster patches include: a first region type raster patch and a second region type raster patch, the first region type raster patch is a raster patch with a pixel width of a preset value, and the second region type raster patch is a diagonal point patch belonging to the same category; The first regional type raster patch is merged with the adjacent regional type raster patches of the first regional type raster patch to obtain the merged result; In response to the fact that the category of the second regional type raster patch is different from the category of the surrounding point regional type raster patches of the second regional type raster patch, the surrounding point regional type raster patches are replaced with patches of the same category as the second regional type raster patch, and a replacement result is obtained; Based on the merging results and the replacement results, the first regional type normalization result is obtained; Based on the target region type vector data, the map patch boundary regularization process is performed to obtain the second region type regularization result; The target region type is determined by using the first region type normalization result and the second region type normalization result, wherein the target region type normalization result is used to characterize the vector patch visualization normalization result corresponding to the remote sensing image of the target region type.

7. The remote sensing image processing method according to claim 6, characterized in that, A graphical user interface is provided through a target cloud platform, wherein the content displayed by the graphical user interface at least partially includes a patch normalization scene of a remote sensing image of a target region type, and the remote sensing image processing method further includes: In response to a first touch operation applied to the graphical user interface, a remote sensing image of the target region type is selected from the storage space provided by the target cloud platform. In response to a second touch operation applied to the graphical user interface, an AI-interpreted tab is selected from multiple tabs provided by the target cloud platform; In response to a third touch operation applied to the graphical user interface, the image normalization function is selected from the AI ​​interpretation tab. In response to a fourth touch operation applied to the graphical user interface, a target patch normalization processing option is determined from multiple candidate patch normalization processing options provided by the patch normalization function; The graphical user interface displays the target region type normalization result corresponding to the target patch normalization processing option.

8. A remote sensing image processing method, characterized in that, include: Land cover classification and interpretation are performed on the image features of remote sensing images of target water body types to obtain raster data of target water body types; The target water area type raster data is converted to obtain target water area type vector data; Obtain target water type raster patches from the target water type raster data, wherein the target water type raster patches include: a first water type raster patch and a second water type raster patch, the first water type raster patch being a raster patch with a preset pixel width, and the second water type raster patch being a diagonal point patch belonging to the same category; The first water area type raster patch is merged with the adjacent water area type raster patches of the first water area type raster patch to obtain the merged result; In response to the fact that the category of the second water type raster patch is different from the category of the surrounding point water type raster patches of the second water type raster patch, the surrounding point water type raster patches are replaced with patches of the same category as the second water type raster patch, and a replacement result is obtained; Based on the merging results and the replacement results, the first water area type regularization result is obtained; Based on the target water area type vector data, the map boundary regularization process is performed to obtain the second water area type regularization result; The target water type is determined by using the first water type normalization result and the second water type normalization result, wherein the target water type normalization result is used to characterize the vector patch visualization normalization result corresponding to the remote sensing image of the target water type.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the remote sensing image processing method according to any one of claims 1 to 8.

10. An electronic device, characterized in that, include: processor; as well as A memory, connected to the processor, for providing the processor with instructions to perform the following processing steps: The image features of the remote sensing image are interpreted to obtain the first processing result; The first processing result is converted into a data format to obtain the second processing result; The target grid patch is obtained from the first processing result, wherein the target grid patch includes: a first grid patch and a second grid patch, the first grid patch is a grid patch with a pixel width of a preset value, and the second grid patch is a diagonal point patch belonging to the same category; The first grid patch is merged with its adjacent grid patches to obtain the merged result; In response to the fact that the category of the second grid patch is different from the category of the surrounding point grid patches of the second grid patch, the surrounding point grid patches are replaced with patches of the same category as the second grid patch, and a replacement result is obtained; Based on the merging result and the replacement result, a first regularization result is obtained; Based on the second processing result, the boundary regularization of the patch is performed to obtain the second regularization result; The target regularization result is determined using the first regularization result and the second regularization result, wherein the target regularization result is used to characterize the visualization regularization result of the vector patch corresponding to the remote sensing image.

Images