Remote sensing image classification method, device, computer equipment and readable storage medium

Abstract

The present invention provides a remote sensing image classification method, device, computer equipment and readable storage medium, the remote sensing image classification method comprising: performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks; extracting the image feature value of each image block according to a preset algorithm; performing clustering and merging processing on the multiple image blocks according to the image feature value to obtain a segmented cluster image, and obtaining the corresponding number of clusters; when it is determined according to the number of clusters and preset rules that the segmented cluster image does not need to be further segmented, the segmented cluster image is used to perform similarity calculation with a pre-stored standard classification image to obtain a corresponding classification result; otherwise, returning to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels. The remote sensing image classification method of the present invention can reduce the difficulty of remote sensing image classification and reduce the requirements for hardware in the classification process.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a remote sensing image classification method, device, computer equipment and readable storage medium. Background Art

[0002] The existing classification of remote sensing images generally relies on artificial intelligence technology. However, the classification of remote sensing images through artificial intelligence technology requires a large amount of sample data for the training of classification models. Due to the strict requirements of data confidentiality in my country's surveying and mapping industry, it is relatively difficult to obtain a large amount of sample data, and training sample data takes a lot of time, which makes the classification of remote sensing images more difficult. In addition, the classification of remote sensing images that relies on artificial intelligence technology also has very high hardware requirements and is difficult to implement. Summary of the invention

[0003] In view of the above problems, the present invention provides a remote sensing image classification method, apparatus, computer equipment and readable storage medium to reduce the difficulty of remote sensing image classification and reduce the requirements for hardware during the classification process.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] A remote sensing image classification method, comprising:

[0006] Obtaining a preset initial number of superpixels as the current number of superpixels, and performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks;

[0007] Extracting image feature values ​​of each image block according to a preset algorithm;

[0008] Performing clustering and merging processing on the plurality of image blocks according to the image feature values ​​to obtain a segmented cluster image, and acquiring a corresponding number of clusters;

[0009] When it is determined according to the number of clusters and the preset rules that the segmented cluster image does not need to be further segmented, similarity calculation is performed using the segmented cluster image and a pre-stored standard classification image to obtain a corresponding classification result;

[0010] When it is determined that the segmented cluster image needs to be further segmented, the next preset number of superpixels is obtained in sequence as the current number of superpixels, and the process returns to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

[0011] Preferably, in the remote sensing image classification method, extracting the image feature value of each image block according to a preset algorithm comprises:

[0012] Extracting RGB color feature values ​​of the image block;

[0013] After grayscale processing of the image block, extracting texture feature values;

[0014] The RGB color feature value and the texture feature value are used to perform a preset weight addition algorithm to obtain the image feature value.

[0015] Preferably, in the remote sensing image classification method, the formula of the preset weight addition algorithm includes:

[0016] O=i(R+G+B)+jGray; Gray=(R×30+G×59+B×11) / 100;

[0017] In the formula, O is the image feature value of the image block, R is the average red light brightness level of the image block, G is the average green light brightness level of the image block, B is the average blue light brightness level of the image block, Gray is the texture feature value of the image block, and i, j are weight values.

[0018] Preferably, in the remote sensing image classification method, the preset rules include:

[0019] A receiver operating characteristic curve is constructed by the number of superpixels and the corresponding number of clusters; the classification accuracy Q(n) is calculated by the receiver operating characteristic curve, where n is the number of segmentation clustering times; if Q(n)>λ, it is determined that the segmented cluster image needs to be further segmented; if Q(n+1)-Q(n)≤λ, it is determined that the segmented cluster image does not need to be further segmented, and the n-th layer of segmented cluster images is used to perform similarity calculation, where λ is a preset classification accuracy threshold.

[0020] Preferably, in the remote sensing image classification method, the step of calculating similarity between the segmented cluster image and a pre-stored standard classification image to obtain a corresponding classification result comprises:

[0021] Calculating the target cluster center of the image feature values ​​of the segmented cluster image, and obtaining the standard cluster centers of the image feature values ​​of all pre-stored standard classification images;

[0022] Calculating the Euclidean distance between the target cluster center and all the standard cluster centers;

[0023] The category of the standard classification image corresponding to the minimum Euclidean distance value is determined as the classification result.

[0024] Preferably, the remote sensing image classification method further comprises:

[0025] The remote sensing image to be classified and the standard classification image are sorted according to the similarity according to the Euclidean distance value, and the corresponding sorting result is output.

[0026] Preferably, in the remote sensing image classification method, the standard classification image is stored in a standard classification image library, and the method further comprises:

[0027] The segmented clustered images, the cluster centers of the corresponding image feature values ​​and the corresponding classification results are associated and stored in the standard classification image library.

[0028] The present invention also provides a remote sensing image classification device, comprising:

[0029] A superpixel segmentation module is used to obtain a preset initial number of superpixels as the current number of superpixels, and perform superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks;

[0030] A feature value extraction module, used to extract the image feature value of each image block according to a preset algorithm;

[0031] A clustering merging module, used for performing clustering merging processing on the plurality of image blocks according to the image feature values, obtaining a segmented cluster image, and acquiring a corresponding number of clusters;

[0032] A classification result acquisition module, for obtaining a corresponding classification result by using the segmented cluster image and a pre-stored standard classification image to perform similarity calculation when determining that the segmented cluster image does not need to be further segmented according to the number of clusters and a preset rule;

[0033] The segmentation return module is used to determine that when the segmented cluster image needs to be further segmented, sequentially obtain the next preset number of superpixels as the current number of superpixels, and return to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

[0034] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program runs on the processor, the remote sensing image classification method is executed.

[0035] The present invention also provides a readable storage medium storing a computer program, wherein the computer program executes the remote sensing image classification method when running on a processor.

[0036] The present invention provides a remote sensing image classification method, which comprises: obtaining a preset initial superpixel number as the current superpixel number, performing superpixel segmentation processing on the remote sensing image to be classified according to the current superpixel number to obtain multiple image blocks; extracting image feature values ​​of each of the image blocks according to a preset algorithm; performing clustering and merging processing on the multiple image blocks according to the image feature values ​​to obtain a segmented cluster image, and obtaining a corresponding number of clusters; when it is determined according to the number of clusters and a preset rule that the segmented cluster image does not need to be further segmented, using the segmented cluster image and a pre-stored standard classification image to perform similarity calculation to obtain a corresponding classification result; when it is determined that the segmented cluster image needs to be further segmented, sequentially obtaining the next preset number of superpixels as the current number of superpixels, and returning to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels. The remote sensing image classification method of the present invention performs segmentation and merging of remote sensing images to be classified by successively increasing the number of superpixels to obtain an optimally layered segmented cluster image, and finally uses the segmented cluster image to perform similarity calculation with a pre-stored standard classification image to obtain a corresponding classification effect. It does not require a large amount of training sample data, and only requires a small number of standard classification images to complete the classification process of the remote sensing images to be classified, thereby reducing the difficulty of remote sensing image classification and reducing the requirements for hardware in the classification process.

[0037] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope of protection of the present invention. In each of the drawings, similar components are numbered similarly.

[0039] Figure 1 is a flow chart of a remote sensing image classification method provided by Embodiment 1 of the present invention;

[0040] Figure 2 is a flow chart of extracting image feature values ​​provided by Embodiment 2 of the present invention;

[0041] Figure 3 is a flow chart of similarity calculation provided by Embodiment 3 of the present invention;

[0042] Figure 4 is a flowchart of another similarity calculation provided by Embodiment 3 of the present invention;

[0043] Figure 5is a flow chart of a remote sensing image classification method provided by Embodiment 4 of the present invention;

[0044] Figure 6 It is a structural schematic diagram of a remote sensing image classification device provided in Example 5 of the present invention. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0046] The components of the embodiments of the present invention generally described and shown in the drawings herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] Hereinafter, the terms "including", "having" and their cognates, which may be used in various embodiments of the present invention, are intended only to indicate specific features, numbers, steps, operations, elements, components or combinations of the foregoing items, and should not be understood as first excluding the existence of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing items or adding the possibility of one or more features, numbers, steps, operations, elements, components or combinations of the foregoing items.

[0048] Furthermore, the terms “first”, “second”, “third”, etc. are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0049] Unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meanings as those generally understood by those skilled in the art to which the various embodiments of the present invention belong. The terms (such as those defined in generally used dictionaries) will be interpreted as having the same meanings as the contextual meanings in the relevant technical field and will not be interpreted as having idealized meanings or overly formal meanings unless clearly defined in the various embodiments of the present invention.

[0050] Example 1

[0051] Figure 1 : is a flowchart of a remote sensing image classification method provided by Example 1 of the present invention, the method comprising the following steps:

[0052] Step S11: obtaining a preset initial number of superpixels as the current number of superpixels, and performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks.

[0053] In an embodiment of the present invention, according to the principle of segmentation and merging of remote sensing images, a layered multiple segmentation and merging design can be performed on the remote sensing images. For example, superpixel segmentation of the first layer image is performed according to a preset initial number of superpixels, and then clustering and merging are performed according to the color features and texture features of each image block after segmentation, that is, image blocks with color features and texture features within a certain similarity range are merged together to finally form a first layer image. Similarly, the number of superpixels is continued to be increased to perform segmentation and merging of the first layer image to form a second layer image, and segmentation and merging are continued to the Nth layer image, so as to use the standard images of the same layer for similarity comparison to obtain the corresponding categories.

[0054] In an embodiment of the present invention, a plurality of superpixel numbers that increase in sequence are stored in a computer device. For example, the initial superpixel number is 100, and superpixel numbers 500, 1000, 2000, 4000, etc. are stored in sequence, which is not limited here. An application program for performing superpixel segmentation processing using the current number of superpixels is also stored in the computer device. For example, after determining the current number of superpixels and obtaining the remote sensing image to be classified, the current number of superpixels and the remote sensing image to be classified can be input into the application program to obtain a corresponding plurality of image blocks.

[0055] Step S12: extracting the image feature value of each image block according to a preset algorithm.

[0056] In the embodiment of the present invention, after obtaining multiple image blocks, a preset algorithm is used to extract image feature values ​​for each image block, wherein the image feature values ​​include color feature values ​​and texture feature values, etc., which are not limited here. An application based on a preset algorithm may be provided in the computer device, and after obtaining multiple image blocks, the image blocks may be sequentially input into the application to extract corresponding image feature values.

[0057] Step S13: performing clustering and merging processing on the plurality of image blocks according to the image feature values ​​to obtain a segmented cluster image and acquire a corresponding number of clusters.

[0058] In the embodiment of the present invention, image blocks with the same image feature values ​​or those within a certain range can be clustered and merged to obtain a segmented cluster image of the current segmentation and merging level. In addition, during the merging process, the corresponding number of clusters, that is, the number of merged image blocks, will be obtained. The above-mentioned clustering and merging process can be implemented using an application. For example, a clustering and merging application is provided in a computer device, and the clustering range of the application can be set. The image feature values ​​of each image block can be input into the application to obtain the clustering results of the image blocks, and clustering and merging processing is performed according to the clustering results.

[0059] Step S14: when it is determined according to the number of clusters and preset rules that the segmented cluster image does not need to be further segmented, similarity calculation is performed using the segmented cluster image and a pre-stored standard classification image to obtain a corresponding classification result.

[0060] In the embodiment of the present invention, the more superpixels there are, the better the effect of the final segmentation and merging of the segmented clustered image. However, after the number of superpixels reaches a certain number, the segmentation and merging effect does not change significantly, and the upper limit of the number of superpixels for each remote sensing image to be classified is different. Therefore, a preset rule can be set. When the number of clusters no longer changes significantly, the segmentation of the remote sensing image to be classified can be stopped, and the similarity calculation can be performed between the current segmented clustered image and the pre-stored standard classification image.

[0061] In an embodiment of the present invention, the above-mentioned pre-stored standard classification image is a remote sensing image with classification categories manually marked in advance. For example, in the national geographic conditions monitoring project technology, it is stipulated that 0110 represents aquatic crops, 0120 represents drought-tolerant crops, 0390 represents natural grass, 0550 represents low houses, 0950 represents rock surface, etc., 8 major categories and 88 third-level categories, all of the standard classification images are respectively calculated for similarity with the remote sensing image to be classified, the standard classification image most similar to the remote sensing image to be classified is obtained, and the corresponding category is determined as the classification result of the remote sensing image to be classified.

[0062] In an embodiment of the present invention, the above-mentioned pre-stored standard classification images are stored in a standard classification image library. When similarity calculation is required, the computer device can retrieve them from the standard classification image library and perform similarity calculations with the current segmentation clustering image in sequence through the application program.

[0063] Step S15: When it is determined that the segmented cluster image needs to be further segmented, the next preset number of superpixels is obtained in sequence as the current number of superpixels, and the process returns to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

[0064] In an embodiment of the present invention, when it is determined that the segmented cluster image needs to continue to be segmented, that is, the optimal number of superpixels has not been reached and the effect of segmentation and merging has not reached the optimal level, the next stored number of superpixels can be obtained, and the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels is returned.

[0065] In the embodiment of the present invention, the preset rules include:

[0066] A receiver operating characteristic curve is constructed by the number of superpixels and the corresponding number of clusters; the classification accuracy Q(n) is calculated by the receiver operating characteristic curve, where n is the number of segmentation clustering times; if Q(n)>λ, it is determined that the segmented cluster image needs to be further segmented; if Q(n+1)-Q(n)≤λ, it is determined that the segmented cluster image does not need to be further segmented, and the n-th layer of segmented cluster images is used to perform similarity calculation, where λ is a preset classification accuracy threshold.

[0067] In the embodiment of the present invention, the remote sensing image to be classified is segmented and merged by successively increasing the number of superpixels to obtain an optimally layered segmented cluster image, and finally the segmented cluster image is used to calculate the similarity with the pre-stored standard classification image to obtain the corresponding classification effect. Without a large amount of training sample data, only a small number of standard classification images are needed to complete the classification process of the remote sensing image to be classified, thereby reducing the difficulty of remote sensing image classification and reducing the requirements for hardware in the classification process.

[0068] Example 2

[0069] Figure 2 This is a flowchart of extracting image feature values ​​provided by Embodiment 2 of the present invention, comprising the following steps:

[0070] Step S21: extracting the RGB color feature value of the image block.

[0071] Step S22: After grayscale processing of the image block, texture feature values ​​are extracted.

[0072] Step S23: performing a preset weight addition algorithm operation using the RGB color feature value and the texture feature value to obtain the image feature value.

[0073] In the embodiment of the present invention, the formula of the preset weight addition algorithm includes:

[0074] O=i(R+G+B)+jGray; Gray=(R×30+G×59+B×11) / 100;

[0075] In the formula, O is the image feature value of the image block, R is the average red light brightness level of the image block, G is the average green light brightness level of the image block, B is the average blue light brightness level of the image block, Gray is the texture feature value of the image block, and i, j are weight values.

[0076] Example 3

[0077] Figure 3 3 is a flowchart of similarity calculation provided by Example 3 of the present invention, comprising the following steps:

[0078] Step S31: Calculate the target cluster center of the image feature values ​​of the segmented cluster image, and obtain the standard cluster centers of the image feature values ​​of all pre-stored standard classification images.

[0079] In an embodiment of the present invention, an application for calculating the target cluster center may be provided in the computer device. After obtaining the optimal segmented cluster image, the corresponding image feature values ​​may be input into the application to calculate the corresponding target cluster center.

[0080] Step S32: Calculate the Euclidean distance between the target cluster center and all the standard cluster centers.

[0081] Step S33: Determine the category of the standard classification image corresponding to the minimum Euclidean distance value as the classification result.

[0082] In the embodiment of the present invention, the Euclidean distance is used to measure the similarity between the target cluster center and the standard cluster center, that is, the similarity between the segmented cluster image and the standard classification image is measured by the Euclidean distance. An application for calculating the Euclidean distance value can also be set in the computer device. After inputting the target cluster center, the standard cluster center can be input in turn to calculate a plurality of corresponding Euclidean distance values.

[0083] Figure 4 is another similarity calculation flowchart provided by Embodiment 3 of the present invention, and further includes the following steps:

[0084] Step S34: sorting the similarities between the remote sensing images to be classified and the standard classification images according to the Euclidean distance values, and outputting the corresponding sorting results.

[0085] In the embodiment of the present invention, the remote sensing images to be classified and the standard classification images may be sorted according to the Euclidean distance values ​​to generate a ranking result for manual inspection and measurement of classification accuracy.

[0086] Example 4

[0087] Figure 54 is a flowchart of a remote sensing image classification method provided by embodiment 4 of the present invention, the method comprising the following steps:

[0088] Step S51: obtaining a preset initial number of superpixels as the current number of superpixels, and performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks.

[0089] This step is consistent with the above step S11 and will not be described again here.

[0090] Step S52: extracting the image feature value of each image block according to a preset algorithm.

[0091] This step is consistent with the above step S12 and will not be repeated here.

[0092] Step S53: performing clustering and merging processing on the plurality of image blocks according to the image feature values ​​to obtain a segmented cluster image and acquire a corresponding number of clusters.

[0093] This step is consistent with the above step S13 and will not be repeated here.

[0094] Step S54: when it is determined according to the number of clusters and the preset rule that the segmented cluster image does not need to be further segmented, similarity calculation is performed using the segmented cluster image and a pre-stored standard classification image to obtain a corresponding classification result.

[0095] This step is consistent with the above step S14 and will not be repeated here.

[0096] Step S55: When it is determined that the segmented cluster image needs to be further segmented, the next preset number of superpixels is obtained in sequence as the current number of superpixels, and the process returns to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

[0097] This step is consistent with the above step S15 and will not be repeated here.

[0098] Step S56: the segmented clustered images, the cluster centers of the corresponding image feature values ​​and the corresponding classification results are associated and stored in the standard classification image library.

[0099] In an embodiment of the present invention, after obtaining the correct classification result, the segmented cluster image, the cluster center of the corresponding image feature value and the corresponding classification result can be associated and stored in a standard classification image library as a new standard classification image to improve the accuracy of subsequent classification.

[0100] Example 5

[0101] Figure 6 It is a structural schematic diagram of a remote sensing image classification device provided in Example 5 of the present invention.

[0102] The remote sensing image classification device 600 comprises:

[0103] The superpixel segmentation module 610 is used to obtain a preset initial number of superpixels as the current number of superpixels, and perform superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks;

[0104] A feature value extraction module 620 is used to extract the image feature value of each image block according to a preset algorithm;

[0105] A clustering merging module 630 is used to perform clustering merging processing on the plurality of image blocks according to the image feature values ​​to obtain a segmented cluster image and obtain a corresponding number of clusters;

[0106] The classification result acquisition module 640 is used to calculate the similarity between the segmented cluster image and the pre-stored standard classification image to obtain the corresponding classification result when it is determined that the segmented cluster image does not need to be further segmented according to the number of clusters and the preset rules;

[0107] The segmentation return module 650 is used to determine that when the segmented cluster image needs to be further segmented, sequentially obtain the next preset number of superpixels as the current number of superpixels, and return to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

[0108] In the embodiment of the present invention, a more detailed functional description of each of the above modules can be referred to the corresponding parts of the aforementioned embodiment, which will not be repeated here.

[0109] In addition, the present invention also provides a computer device, which may include a smart phone, a tablet computer, a car computer, a smart wearable device, etc. The computer device includes a memory and a processor, the memory can be used to store a computer program, and the processor runs the computer program to enable the computer device to perform the functions of each module in the above method or the above remote sensing image classification device.

[0110] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the computer device (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0111] This embodiment also provides a computer storage medium for storing a computer program used in the above-mentioned computer device.

[0112] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and structure diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in an alternative implementation, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the structure diagram and / or the flow chart, and the combination of boxes in the structure diagram and / or the flow chart, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0113] In addition, the functional modules or units in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0114] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0115] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A remote sensing image classification method, characterized in that: include: Obtaining a preset initial number of superpixels as the current number of superpixels, and performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks; Extracting image feature values ​​of each image block according to a preset algorithm; Performing clustering and merging processing on the plurality of image blocks according to the image feature values ​​to obtain a segmented cluster image, and acquiring a corresponding number of clusters; When it is determined according to the number of clusters and the preset rules that the segmented cluster image does not need to be further segmented, similarity calculation is performed using the segmented cluster image and a pre-stored standard classification image to obtain a corresponding classification result; The preset rules include: A receiver operating characteristic curve is constructed by the number of superpixels and the corresponding number of clusters; the classification accuracy Q(n) is calculated by the receiver operating characteristic curve, where n is the number of segmentation clustering times; if Q(n)>λ, it is determined that the segmentation cluster image needs to be further segmented; if Q(n+1)-Q(n)≤λ, it is determined that the segmentation cluster image does not need to be further segmented, and the n-th layer segmentation cluster image is used to perform similarity calculation, where λ is a preset classification accuracy threshold; The method of using the segmented cluster image to perform similarity calculation with a pre-stored standard classification image to obtain a corresponding classification result includes: Calculating the target cluster center of the image feature values ​​of the segmented cluster image, and obtaining the standard cluster centers of the image feature values ​​of all pre-stored standard classification images; Calculating the Euclidean distance between the target cluster center and all the standard cluster centers; Determine the category of the standard classification image corresponding to the minimum Euclidean distance value as the classification result; When it is determined that the segmented cluster image needs to be further segmented, the next preset number of superpixels is obtained as the current number of superpixels according to the successively increasing number of superpixels, and the process returns to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

2. The remote sensing image classification method according to claim 1, characterized in that: The extracting the image feature value of each image block according to a preset algorithm comprises: Extracting RGB color feature values ​​of the image block; After grayscale processing of the image block, extracting texture feature values; The RGB color feature value and the texture feature value are used to perform a preset weight addition algorithm to obtain the image feature value.

3. The remote sensing image classification method according to claim 2, characterized in that: The formulas of the preset weighted addition algorithm include: O=i(R+G+B)+jGray; Gray=(R×30+G×59+B×11) / 100; In the formula, O is the image feature value of the image block, R is the average red light brightness level of the image block, G is the average green light brightness level of the image block, B is the average blue light brightness level of the image block, Gray is the texture feature value of the image block, and i, j are weight values.

4. The remote sensing image classification method according to claim 1, characterized in that: Also includes: The remote sensing image to be classified and the standard classification image are sorted according to the similarity according to the Euclidean distance value, and the corresponding sorting result is output.

5. The remote sensing image classification method according to claim 1, characterized in that: The standard classification image is stored in a standard classification image library, and the method further comprises: The segmented clustered images, the cluster centers of the corresponding image feature values ​​and the corresponding classification results are associated and stored in the standard classification image library.

6. A remote sensing image classification device, characterized in that: include: A superpixel segmentation module is used to obtain a preset initial number of superpixels as the current number of superpixels, and perform superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels to obtain multiple image blocks; A feature value extraction module, used to extract the image feature value of each image block according to a preset algorithm; A clustering merging module, used for performing clustering merging processing on the plurality of image blocks according to the image feature values, obtaining a segmented cluster image, and acquiring a corresponding number of clusters; A classification result acquisition module, for obtaining a corresponding classification result by using the segmented cluster image and a pre-stored standard classification image to perform similarity calculation when determining that the segmented cluster image does not need to be further segmented according to the number of clusters and a preset rule; The preset rules include: A receiver operating characteristic curve is constructed by the number of superpixels and the corresponding number of clusters; the classification accuracy Q(n) is calculated by the receiver operating characteristic curve, where n is the number of segmentation clustering times; if Q(n)>λ, it is determined that the segmentation cluster image needs to be further segmented; if Q(n+1)-Q(n)≤λ, it is determined that the segmentation cluster image does not need to be further segmented, and the n-th layer segmentation cluster image is used to perform similarity calculation, where λ is a preset classification accuracy threshold; The classification result acquisition module is also used for: Calculating the target cluster center of the image feature values ​​of the segmented cluster image, and obtaining the standard cluster centers of the image feature values ​​of all pre-stored standard classification images; Calculating the Euclidean distance between the target cluster center and all the standard cluster centers; Determine the category of the standard classification image corresponding to the minimum Euclidean distance value as the classification result; The segmentation return module is used to determine that when the segmented cluster image needs to be further segmented, obtain the next preset number of superpixels as the current number of superpixels according to the successively increasing number of superpixels, and return to the step of performing superpixel segmentation processing on the remote sensing image to be classified according to the current number of superpixels.

7. A computer device, characterized in that: The invention comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is run on the processor, the remote sensing image classification method according to any one of claims 1 to 5 is executed.

8. A readable storage medium, characterized in that: The computer program stores a computer program, which executes the remote sensing image classification method according to any one of claims 1 to 5 when running on a processor.

Images