A remote sensing image segmentation method and system based on neighborhood spectral similarity
By applying neighborhood spectral similarity analysis and segmentation thresholds, the problem of insufficient accuracy in remote sensing image segmentation was solved, and high-precision image segmentation and classification were achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID ECONOMIC TECH RES INST CO LTD
- Filing Date
- 2023-10-13
- Publication Date
- 2026-05-26
AI Technical Summary
Existing remote sensing image segmentation techniques fail to fully utilize the spectral similarity of adjacent ground features, resulting in insufficient segmentation accuracy.
By performing neighborhood spectral similarity analysis on the preprocessed remote sensing image data, the spectral reflectance distance, skewness, and kurtosis difference between each pixel and its surrounding neighboring pixels are calculated. Image segmentation is then performed using a segmentation threshold to achieve accurate image classification.
This significantly improves the segmentation accuracy of remote sensing images, ensuring the accuracy of subsequent image classification.
Smart Images
Figure CN117315253B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a remote sensing image segmentation method and system based on neighborhood spectral similarity, belonging to the field of image classification. Background Technology
[0002] Object-oriented image classification is a commonly used image classification method. It has advantages such as effectively avoiding the "salt and pepper" phenomenon caused by pixel-based classification methods and combining the texture and other features of objects for classification. Image segmentation is the fundamental step in object-oriented image classification; only with accurate image segmentation can subsequent image classification be accurate.
[0003] Adjacent land features are more similar, but abrupt changes occur at land class boundaries. Current image segmentation techniques, such as multi-scale segmentation algorithms, do not fully utilize the spectral similarity of adjacent land features for segmentation, which is an area where future technology needs to make breakthroughs. Summary of the Invention
[0004] To address the aforementioned problems, the present invention aims to provide a remote sensing image segmentation method and system based on neighborhood spectral similarity. This method, based on preprocessed remote sensing images, fully utilizes the spectral similarity and similarity indices of neighboring land features to segment the remote sensing images, thereby significantly improving segmentation accuracy.
[0005] To achieve the above objectives, the present invention adopts the following technical solution:
[0006] In a first aspect, the present invention provides a remote sensing image segmentation method based on neighborhood spectral similarity, comprising the following steps:
[0007] The acquired raw remote sensing images are preprocessed to obtain remote sensing image data containing the reflectance of ground objects in each band.
[0008] Based on remote sensing image data containing the reflectance of ground objects in each band, preset segmentation indicators are calculated, and remote sensing image data are classified based on preset segmentation thresholds to obtain remote sensing image classification results.
[0009] Furthermore, the preprocessing of the acquired raw remote sensing images includes radiometric correction, atmospheric correction, geometric correction, and cropping.
[0010] Furthermore, the process of calculating segmentation indices based on remote sensing image data containing reflectance of ground objects in each band, and classifying the remote sensing image data based on a preset segmentation threshold to obtain remote sensing image classification results includes:
[0011] Based on remote sensing image data containing the reflectance of ground objects in each band, a segmentation index based on neighborhood spectral similarity is calculated.
[0012] Determine the segmentation threshold based on the segmentation requirements;
[0013] By using a defined segmentation threshold, remote sensing image data is segmented to obtain remote sensing image classification results.
[0014] Furthermore, when calculating the segmentation index based on neighborhood spectral similarity using remote sensing image data containing reflectance of ground objects in each band, the following steps are included:
[0015] Starting from the first pixel in the remote sensing image, obtain the spectral reflectance distance S between each pixel and its neighboring pixels J.
[0016] Calculate the skewness W and kurtosis K of the reflectance of each pixel across all bands;
[0017] Based on the skewness W and kurtosis K of the reflectance of each pixel across all bands, a series of skewness differences ΔW and kurtosis ΔK between each central pixel and its neighboring pixels are calculated.
[0018] Furthermore, the step of segmenting remote sensing image data using determined segmentation thresholds to obtain remote sensing image segmentation results includes: merging center pixels and neighboring pixels that are smaller than each group of segmentation thresholds S', △W', △K' to achieve image segmentation.
[0019] Secondly, the present invention provides a remote sensing image segmentation system based on neighborhood spectral similarity, comprising:
[0020] The data preprocessing module preprocesses the acquired raw remote sensing images to obtain remote sensing image data containing the reflectance of ground objects in each band.
[0021] The image classification module is used to calculate segmentation indicators based on remote sensing image data containing the reflectance of ground objects in each band, and classify the remote sensing image data based on a preset segmentation threshold to obtain the remote sensing image classification result.
[0022] Furthermore, the image classification module includes:
[0023] The segmentation index calculation module is used to calculate the segmentation index based on the spectral similarity of the neighborhood based on image data containing the reflectance of ground objects in each band.
[0024] The segmentation threshold determination module is used to determine the segmentation threshold based on the segmentation requirements.
[0025] The segmentation module is used to segment remote sensing image data using a defined segmentation threshold to obtain remote sensing image classification results.
[0026] Furthermore, the segmentation index calculation module includes:
[0027] The reflectance calculation module is used to obtain the spectral reflectance distance S between each pixel and its neighboring pixels J, starting from the first pixel on the remote sensing image.
[0028] The skewness and kurtosis calculation module is used to calculate the skewness W and kurtosis K of the reflectance of each pixel across all bands.
[0029] The skewness and kurtosis difference calculation module is used to calculate a series of skewness differences ΔW and kurtosis differences ΔK between each central pixel and its neighboring pixels, based on the skewness W and kurtosis K of the reflectance of all bands of each pixel.
[0030] Thirdly, the present invention provides a computer-readable storage medium for storing one or more programs, said one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any of the methods described.
[0031] Fourthly, the present invention provides a computing device comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described.
[0032] The present invention, by adopting the above technical solutions, has the following advantages: The remote sensing image classification method based on neighborhood spectral similarity provided by the present invention, based on preprocessed remote sensing images, fully utilizes the spectral similarity and similarity indicators of neighboring land features to segment the remote sensing images, greatly improving the segmentation accuracy. Therefore, the present invention can be widely applied in the field of image classification. Attached Figure Description
[0033] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. In the drawings:
[0034] Figure 1 This is a flowchart of the remote sensing image segmentation method based on neighborhood spectral similarity provided in an embodiment of the present invention. Detailed Implementation
[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.
[0036] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0037] Based on the law of geographical similarity, everything is related; the closer they are, the more related they are, especially in images. Therefore, similar features in adjacent images generally have similar spectral characteristics. Accordingly, some embodiments of this invention provide a remote sensing image classification method based on neighborhood spectral similarity. Based on preprocessed remote sensing images, this method fully utilizes the spectral similarity and similarity indices of neighboring features to segment the remote sensing images, greatly improving segmentation accuracy.
[0038] Correspondingly, in other embodiments of the present invention, a remote sensing image classification system, device, and storage medium based on neighborhood spectral similarity are provided.
[0039] Example 1
[0040] like Figure 1 As shown, this embodiment provides a remote sensing image segmentation method based on neighborhood spectral similarity, including the following steps:
[0041] 1) Preprocess the acquired raw remote sensing images to obtain remote sensing image data containing the reflectance of ground objects in each band.
[0042] 2) Based on remote sensing image data containing the reflectance of ground objects in each band, the segmentation index is calculated, and the remote sensing image data is classified based on the preset segmentation threshold to obtain the remote sensing image classification result.
[0043] Preferably, in step 1) above, the preprocessing of the acquired original remote sensing image includes radiometric correction, atmospheric correction, geometric correction, cropping, etc.
[0044] Preferably, step 2) above includes the following steps:
[0045] 2.1) Based on remote sensing image data containing the reflectance of ground objects in each band, a segmentation index based on neighborhood spectral similarity is calculated;
[0046] 2.2) Determine the segmentation threshold based on the segmentation requirements;
[0047] 2.3) Using the determined segmentation threshold, the remote sensing image data is segmented to obtain the remote sensing image segmentation result.
[0048] Preferably, in step 2.1) above, when calculating the segmentation index based on neighborhood spectral similarity based on image data containing the reflectance of ground objects in each band, the calculation includes:
[0049] 2.1.1) Starting from the first pixel in the remote sensing image, obtain the spectral reflectance distance S between the first pixel and its neighboring pixels J.
[0050] In this embodiment, the 8-neighbor method is used when calculating the spectral reflectance distance of a pixel, as shown in Table 1 below.
[0051] Table 1. Neighboring pixels of a given pixel a
[0052] 1 2 3 4 a 5 6 7 8
[0053]
[0054] In the formula, S j Let a be the spectral distance between the center pixel and its neighboring pixel J. i J represents the spectral reflectance of the central pixel a in the i-band; i Let J be the spectral reflectance of the neighboring pixel J in the i-band.
[0055] 2.1.2) Calculate the skewness W and kurtosis K of the reflectance of each pixel across all bands.
[0056] The formulas for calculating the skewness W and kurtosis K of the reflectance of each pixel across all bands are as follows:
[0057]
[0058]
[0059] Among them, X i Let μ be the reflectance of pixel i in band i, μ be the mean, σ be the standard deviation, and n be the number of image bands.
[0060] 2.1.3) Based on the skewness W and kurtosis K of the reflectance of all bands of each pixel, calculate a series of skewness differences ΔW and kurtosis differences ΔK between each central pixel and its neighboring pixels.
[0061] The formulas for calculating the skewness difference ΔW and the kurtosis difference ΔK are as follows:
[0062] ΔW=Wa-Wi
[0063] ΔK=Ka-Ki
[0064] In the formula, Wa and Ka are the skewness and kurtosis of the reflectance of the central pixel a in all bands, respectively; Wi and Ki are the skewness and kurtosis of the reflectance of the neighboring pixel J of the central pixel a in band i, respectively.
[0065] Preferably, in step 2.2) above, when determining the segmentation threshold according to the segmentation requirements, multiple groups of central pixels and neighboring pixels that need to be segmented are manually selected from the remote sensing image, and the segmentation thresholds S', △W', and △K' are initialized.
[0066] Preferably, in step 2.3) above, the remote sensing image data is segmented using the determined segmentation thresholds to obtain the remote sensing image segmentation result, including merging the center pixels and neighboring pixels that are smaller than the segmentation thresholds S', ΔW', and ΔK' of each group to achieve image segmentation. Alternatively, the segmentation thresholds S', ΔW', and ΔK' of each group can be fine-tuned as needed to achieve the final segmentation.
[0067] Example 2
[0068] The above-described embodiment 1 provides a remote sensing image segmentation method based on neighborhood spectral similarity. Correspondingly, this embodiment provides a remote sensing image segmentation system based on neighborhood spectral similarity. The system provided in this embodiment can implement the remote sensing image segmentation method based on neighborhood spectral similarity of embodiment 1. The system can be implemented by software, hardware, or a combination of both. For example, the system may include integrated or separate functional modules or units to perform the corresponding steps in the methods of embodiment 1. Since the system in this embodiment is basically similar to the method embodiment, the description process in this embodiment is relatively simple. Relevant details can be found in the description of embodiment 1. The system embodiment provided in this embodiment is merely illustrative.
[0069] The remote sensing image segmentation system based on neighborhood spectral similarity provided in this embodiment includes:
[0070] The data preprocessing module preprocesses the acquired raw remote sensing images to obtain remote sensing image data containing the reflectance of ground objects in each band.
[0071] The image classification module is used to calculate segmentation indicators based on remote sensing image data containing the reflectance of ground objects in each band, and classify the remote sensing image data based on a preset segmentation threshold to obtain the remote sensing image classification result.
[0072] Preferably, the image classification module includes:
[0073] The segmentation index calculation module is used to calculate the segmentation index based on the spectral similarity of the neighborhood based on image data containing the reflectance of ground objects in each band.
[0074] The segmentation threshold determination module is used to determine the segmentation threshold based on the segmentation requirements.
[0075] The segmentation module is used to segment remote sensing image data using a defined segmentation threshold to obtain remote sensing image segmentation results.
[0076] Example 3
[0077] This embodiment provides a processing device corresponding to the remote sensing image segmentation method based on neighborhood spectral similarity provided in Embodiment 1. The processing device can be a client-side processing device, such as a mobile phone, laptop, tablet computer, desktop computer, etc., to execute the method of Embodiment 1.
[0078] The processing device includes a processor, a memory, a communication interface, and a bus. The processor, memory, and communication interface are connected via the bus to communicate with each other. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it executes the remote sensing image segmentation method based on neighborhood spectral similarity provided in Embodiment 1.
[0079] In some embodiments, the memory may be high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device.
[0080] In other embodiments, the processor can be a general-purpose processor of various types, such as a central processing unit (CPU) or a digital signal processor (DSP), and is not limited thereto.
[0081] Example 4
[0082] The remote sensing image segmentation method based on neighborhood spectral similarity in Embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the remote sensing image segmentation method based on neighborhood spectral similarity described in Embodiment 1 are loaded.
[0083] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.
[0084] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.
[0085] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0086] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0087] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0088] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A remote sensing image segmentation method based on neighborhood spectral similarity, characterized in that, Includes the following steps: The acquired raw remote sensing images are preprocessed to obtain remote sensing image data containing the reflectance of ground objects in each band. Based on remote sensing image data containing the reflectance of ground objects in each band, a preset segmentation index is calculated, and the remote sensing image data is segmented based on a preset segmentation threshold to obtain the remote sensing image segmentation result. The process involves calculating segmentation indices based on remote sensing image data containing reflectance of ground objects in various bands, and segmenting the remote sensing image data based on preset segmentation thresholds to obtain remote sensing image segmentation results. This includes: calculating segmentation indices based on neighborhood spectral similarity based on remote sensing image data containing reflectance of ground objects in various bands; determining segmentation thresholds according to segmentation requirements; and segmenting the remote sensing image data using the determined segmentation thresholds to obtain remote sensing image classification results. When calculating the segmentation index based on neighborhood spectral similarity using remote sensing image data containing reflectance of ground objects in each band, the process includes: starting from the first pixel in the remote sensing image, obtaining the relationship between each pixel and its surrounding neighboring pixels. Calculate the spectral reflectance distance S; calculate the skewness of reflectance across all bands for each pixel. and kurtosis ; Skewness based on reflectivity of all bands for each pixel and kurtosis Calculate a series of skewness differences between each central pixel and its neighboring pixels. and kurtosis ; The process of segmenting remote sensing image data using determined segmentation thresholds to obtain remote sensing image segmentation results includes merging center pixels and neighboring pixels that are smaller than each set of segmentation thresholds S', △W', and △K' to achieve image segmentation.
2. The remote sensing image segmentation method based on neighborhood spectral similarity as described in claim 1, characterized in that, The preprocessing of the acquired raw remote sensing images includes radiometric correction, atmospheric correction, geometric correction, and cropping.
3. A remote sensing image segmentation system based on neighborhood spectral similarity, characterized in that, include: The data preprocessing module preprocesses the acquired raw remote sensing images to obtain remote sensing image data containing the reflectance of ground objects in each band. The image classification module is used to calculate the segmentation index based on remote sensing image data containing the reflectance of ground objects in each band, and to segment the remote sensing image data based on the preset segmentation threshold to obtain the remote sensing image segmentation result. The image classification module includes: The segmentation index calculation module is used to calculate the segmentation index based on the spectral similarity of the neighborhood based on image data containing the reflectance of ground objects in each band. The segmentation threshold determination module is used to determine the segmentation threshold based on the segmentation requirements. The segmentation module is used to segment remote sensing image data using a determined segmentation threshold to obtain remote sensing image segmentation results. The segmentation index calculation module includes: The reflectance calculation module is used to obtain the reflectance of each pixel and its surrounding neighboring pixels, starting from the first pixel in the remote sensing image. Spectral reflectance distance S; The skewness and kurtosis calculation module is used to calculate the skewness of the reflectance of each pixel across all bands. and kurtosis ; The skewness and kurtosis difference calculation module is used to calculate the skewness based on the reflectance of each pixel across all bands. and kurtosis Calculate a series of skewness differences between each central pixel and its neighboring pixels. and kurtosis difference ; The process of segmenting remote sensing image data using determined segmentation thresholds to obtain remote sensing image segmentation results includes merging center pixels and neighboring pixels that are smaller than each set of segmentation thresholds S', △W', and △K' to achieve image segmentation.
4. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 2.
5. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 2.