A rock debris slope extraction method and system based on hyperspectral remote sensing images
By using multi-dimensional processing of hyperspectral remote sensing images, the subjectivity and interference problems of traditional remote sensing image recognition methods in identifying debris slopes are solved, achieving efficient and accurate debris slope extraction under complex geological conditions, which is suitable for railway engineering geological route selection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA RAILWAY ENG CONSULTING GRP CO LTD
- Filing Date
- 2026-02-09
- Publication Date
- 2026-05-29
AI Technical Summary
Traditional remote sensing image recognition methods are highly subjective when identifying debris slopes, making it difficult to accurately distinguish between snow and vegetation interference, resulting in low recognition efficiency and poor accuracy, especially in complex geological conditions in high-altitude areas where there is snow cover in winter and rapid vegetation growth in summer.
Hyperspectral remote sensing images were used for linear stretching, tone analysis, band analysis, and normalized vegetation index calculation. Through multi-dimensional spectral feature collaborative screening, interfering ground features were eliminated, and the extent of the debris slope was accurately extracted.
It enables automated identification of rock debris slopes under complex geological conditions, improves identification accuracy and efficiency, reduces computational complexity, and is suitable for railway engineering geological route selection.
Smart Images
Figure CN122116133A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of railway engineering, and more specifically, to a method and system for extracting debris slopes based on hyperspectral remote sensing images. Background Technology
[0002] Rock debris slopes are an extreme and highly hazardous geological hazard. Under the influence of gravity, they flow continuously along steep slopes, causing blockages and damage to roadbeds, bridges, tunnel entrances, and other engineering projects. Therefore, during the selection of geological engineering routes, it is generally necessary to avoid this hazardous geological area or select a smaller area of rock debris slopes. Traditional remote sensing image recognition methods use RGB images and human visual interpretation to identify the characteristics of rock debris slopes, which is highly subjective and relies on the experience and knowledge of geological engineers. Summary of the Invention
[0003] The purpose of this invention is to provide a method and system for extracting debris slopes based on hyperspectral remote sensing images, so as to improve the above-mentioned problems.
[0004] To achieve the above objectives, the embodiments of this application provide the following technical solutions: On one hand, embodiments of this application provide a method for extracting debris slopes based on hyperspectral remote sensing images, the method comprising: Acquire hyperspectral remote sensing images; A first spectral image is obtained by performing linear stretching on the hyperspectral remote sensing image. Based on the first spectral image, a hue analysis process is performed to obtain a first classification map, which includes regions that conform to the hue characteristics of the rock debris slope. The hyperspectral remote sensing image is processed using band analysis to obtain a second spectral image; A second classification map is obtained by performing calculations based on the first spectral image and the second spectral image. The second classification map includes regions that conform to the lithological brightness of the rock debris slope. The hyperspectral remote sensing image is processed to obtain a third classification map, which represents the normalized vegetation index corresponding to each pixel. The first classification map, the second classification map, and the third classification map are processed to obtain the range of the rock debris slope.
[0005] Secondly, this application provides a system for extracting debris slopes based on hyperspectral remote sensing images, the system comprising: The acquisition module is used to acquire hyperspectral remote sensing images; The first processing module is used to perform linear stretching processing on the hyperspectral remote sensing image to obtain a first spectral image; The second processing module is used to perform tone analysis processing based on the first spectral image to obtain a first classification map, the first classification map including areas that conform to the tone characteristics of the rock debris slope. The third processing module is used to process the hyperspectral remote sensing image using band analysis to obtain a second spectral image; The first calculation module is used to perform calculations based on the first spectral image and the second spectral image to obtain a second classification map, the second classification map including regions that conform to the lithological brightness of the rock debris slope; The second calculation module is used to process the hyperspectral remote sensing image to obtain a third classification map, which represents the normalized vegetation index corresponding to each pixel. The fourth processing module is used to process the first classification map, the second classification map and the third classification map to obtain the range of the rock debris slope.
[0006] The beneficial effects of this invention are: This invention acquires hyperspectral remote sensing images and sequentially performs linear stretching, tone analysis, and band analysis to obtain a first classification map containing the tone characteristics of the rock debris slope, a second classification map containing the lithological brightness of the rock debris slope, and a third classification map representing the normalized vegetation index. These classification maps are then comprehensively processed to accurately extract the extent of the rock debris slope. This invention is particularly suitable for complex scenarios such as railway engineering geological route selection, effectively solving the problems of low efficiency and high subjectivity in traditional manual visual interpretation. Through multi-dimensional spectral feature collaborative screening and automatic removal of interfering features, it achieves automated identification with near-expert judgment accuracy without significantly increasing computational complexity, providing efficient and reliable technical support for geological hazard assessment and engineering route selection.
[0007] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description
[0008] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0009] Figure 1 This is a schematic flowchart of the method for extracting rock debris slopes based on hyperspectral remote sensing images as described in an embodiment of the present invention.
[0010] Figure 2 This refers to the RGB image described in the embodiments of the present invention.
[0011] Figure 3 This is the first classification diagram described in the embodiments of the present invention.
[0012] Figure 4 This is the second spectral image described in this embodiment of the invention.
[0013] Figure 5 This is a standard spectral curve of carbonate rock as described in an embodiment of the present invention.
[0014] Figure 6 This is the second classification diagram described in the embodiments of the present invention.
[0015] Figure 7 This is the third classification diagram described in the embodiments of the present invention.
[0016] Figure 8 This is a spectral curve of snow as described in an embodiment of the present invention.
[0017] Figure 9 This is the fourth classification diagram described in the embodiments of the present invention.
[0018] Figure 10 This is the fifth classification diagram described in the embodiments of the present invention.
[0019] Figure 11 The rock debris slope range is as described in the embodiments of the present invention. Detailed Implementation
[0020] 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0021] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.
[0022] Due to the snow cover in winter and the rapid growth of vegetation in summer in high-altitude areas, the color of snow is similar to that of rock debris slopes, and vegetation obscures the spectral characteristics of rock debris slopes. This makes it easy for traditional rock debris slope identification methods to misidentify snow as rock debris slopes, and they cannot remove vegetation interference, ultimately resulting in low accuracy of extraction results.
[0023] Example 1 This embodiment provides a method for extracting debris slopes based on hyperspectral remote sensing images. This embodiment can perform remote sensing image processing and analysis on a section of road traversing the Hengduan Mountains. Given the rugged terrain and complex geological conditions along the route, with some sections covered by snow in winter and experiencing rapid vegetation growth in summer, this technical solution solves the problem of poor accuracy in debris slope identification, providing survey data for route selection and geological engineering.
[0024] See Figure 1 The figure shows that this method includes steps S1-S7, which specifically include: Step S1: Acquire hyperspectral remote sensing images; In this step, the hyperspectral remote sensing image needs to be preprocessed before subsequent processing. The main preprocessing includes radiometric correction, atmospheric correction, geometric correction, error value removal, and negative value removal. The processing steps and order of radiometric correction, atmospheric correction, and geometric correction can vary depending on the specific situation. After radiometric correction, atmospheric correction, and geometric correction are completed, topographic radiometric correction is performed first, followed by error value removal and negative value removal to obtain the preprocessed hyperspectral remote sensing image.
[0025] Step S2: Perform linear stretching processing on the hyperspectral remote sensing image to obtain a first spectral image; Step S2 further includes steps S21, S22, S23, S24, S25, S26, and S27, specifically as follows: Step S21: Process the hyperspectral remote sensing image to obtain an RGB image; Before performing linear stretching on the hyperspectral remote sensing image in this step, the red, green, and blue bands of the preprocessed hyperspectral remote sensing image are exported, such as... Figure 2As shown, the wavelength of the red light band is about 615~650nm, the wavelength of the green light band is about 495~530nm, and the wavelength of the blue light band is about 450~480nm.
[0026] Step S22: Extract the first band value, second band value, and third band value of each pixel in the RGB image; In this step, the first band value represents the band value of the red light band, the second band value represents the band value of the green light band, and the third band value represents the band value of the blue light band.
[0027] Step S23: Add the first band value, the second band value and the third band value together to obtain the sum of the band values; Step S24: Sort the sum of the band values in ascending order to obtain a band value sequence; Step S25: Truncate the band value sequence from both ends to obtain a first truncated region and a second truncated region. The range of the first truncated region and the second truncated region is two percent of the band value sequence. By truncating the two percent regions at both ends of the band value sequence, outliers that do not represent the true surface characteristics of the target area are eliminated, avoiding interference from extreme values on subsequent tone analysis and lithology identification. This ensures that the first band values participating in subsequent processing are all valid data reflecting the true surface conditions, thus improving the anti-interference capability of the extraction process.
[0028] Step S26: Crop the image based on the first truncated region and the second truncated region to obtain the truncated image; Step S27: Linearly stretch the truncated image to obtain the first spectral image.
[0029] In this step, the truncated image is linearly stretched as follows: Where, round means rounding to the nearest integer; min means taking the minimum value; Indicates the value of the first band; Indicates the value of the second band; Indicates the value of the third band; This indicates the output value after stretching; This represents the sum of the band values; This represents the output value after stretching the first band; This represents the output value after stretching the second band; This represents the output value after stretching the third band; This represents the maximum band value corresponding to the first truncated region; This represents the minimum band value corresponding to the second truncation region. Since the first band value is very small in some dark areas and very large in some bright areas in the preprocessed hyperspectral remote sensing image, the preprocessed hyperspectral remote sensing image is not bright. After removing these low and high outliers, the resulting outlier-removed image is linearly stretched to normalize the first band value, the second band value, and the third band value to the standard range of 0~255. This ensures that subsequent operations such as tone screening and lithological brightness calculation based on the first spectral image adopt a unified standard, thereby improving the stability and reliability of the classification results.
[0030] Step S3: Perform tone analysis processing based on the first spectral image to obtain a first classification map, which includes regions that conform to the tone characteristics of the rock debris slope. In this step, for the rock debris slope, after multi-sample statistical analysis, its hue brightness is relatively high, appearing as light gray or grayish-white. Pixel values that simultaneously meet the following three conditions are extracted from the first spectral image: The red light band value is between 200 and 255; the green light band value is between 205 and 255; and the blue light band value is between 190 and 255. like Figure 3 As shown, the first classification map is obtained through the above extraction. By using quantitative color filtering rules, the core features of the rock debris slope are accurately captured. This not only eliminates most irrelevant interference items, but also provides a clear candidate range for subsequent multi-dimensional analysis such as lithology verification and vegetation removal. This improves the overall process efficiency and lays a key foundation for the accuracy of the final extraction results.
[0031] Step S4: Process the hyperspectral remote sensing image using band analysis to obtain a second spectral image; In this step, such as Figure 4 As shown, based on the hyperspectral remote sensing image, absorption and reflection characteristics are observed, characteristic bands are derived, and the second spectral image is obtained. By stripping away redundant information and amplifying lithological spectral differences, a high-quality and highly targeted spectral data foundation is provided for subsequent lithological brightness identification, reducing computational complexity.
[0032] Step S5: Perform calculations based on the first spectral image and the second spectral image to obtain a second classification map, which includes regions that conform to the lithological brightness of the rock debris slope. Step S5 further includes steps S51, S52, S53, and S54, which are specifically as follows: Step S51: Obtain geological data for the target area; Step S52: Obtain the main lithology within the target area based on the geological data of the target area; Step S53: Extract the standard spectral curves of the main lithologies within the target area; In this step, the standard spectral curve can be obtained through on-site measurement, manual acquisition from spectral images, or querying from a spectral library, such as... Figure 5 The figure shows the standard spectral curve of carbonate rocks.
[0033] Step S54: Calculate based on the standard spectral curve, the first spectral image, and the second spectral image to obtain the second classification image.
[0034] In this step, the spectral curve vector of the standard spectral curve is set as Suppose there is a pixel B, and the spectral curve vector of pixel B in the second spectral image is... The band values of pixel B in the first spectral image for red, green, and blue light are respectively .
[0035] The evaluation value D is calculated based on the above parameters, and the calculation formula is as follows: in, Represents the inverse cosine function; This represents the band value corresponding to the red light in pixel B in the first spectral image; This represents the band value corresponding to the green light in pixel B in the first spectral image; This represents the band value of blue light corresponding to pixel B in the first spectral image; n represents the total number of channels; Representing vectors Channel values; Representing vectors Channel values; Representing vectors No. The channel value; Representing vectors No. The channel value; This represents the brightness weight, and its value is determined based on the image and extraction conditions, typically ranging from 0 to 1.
[0036] The above calculation is performed on each pixel in the first and second spectral images, selecting a D value between 0.05 and 0.08. Figure 6As shown, a second classification map can be obtained. By integrating hue information and lithological spectral characteristics, and using geological data as a benchmark, target areas that meet the lithological and brightness standards are selected through D-value quantification. This not only solves the limitations of single feature recognition, but also achieves deep integration of geological background and image data, providing high-quality verification results for subsequent multi-dimensional screening.
[0037] Step S6: Perform calculation processing on the hyperspectral remote sensing image to obtain a third classification map, wherein the third classification map represents the normalized vegetation index corresponding to each pixel; In this step, vegetation cover is assessed by calculating the difference in reflectance between the near-infrared band and the red band, thus eliminating vegetation interference for the extraction of rock debris slopes.
[0038] Step S6 further includes steps S61, S62, S63, and S64, which are specifically as follows: Step S61: Extract the near-infrared band reflectance value and the red band reflectance value of the hyperspectral remote sensing image; Step S62: Calculate the difference between the near-infrared band reflectance value and the red band reflectance value to obtain the first calculation item; Step S63: Calculate the second calculation item by adding the near-infrared band reflectance value and the red band reflectance value; Step S64: Compare the first calculation item with the second calculation item to obtain the third classification map.
[0039] In this step, NDVI is short for Normalized Difference Vegetation Index, a vegetation detection index widely used in remote sensing. Its calculation formula is as follows: in, Indicates the normalized vegetation index; Indicates the reflectance value in the near-infrared band; This indicates the reflectance value in the red light band.
[0040] Extract the calculation results for NDVI values between 0 and 0.2, such as... Figure 7 As shown, the third classification map is obtained. The vegetation identification latitude is supplemented by NDVI calculation, eliminating the interference of vegetation cover on the identification process and further improving the efficiency of subsequent processing. With mature indicators and standardized calculation, the objectivity and stability of the results are guaranteed, which is a key link to achieve multi-dimensional and accurate screening of rock debris slopes.
[0041] Step S7: Process the first classification map, the second classification map and the third classification map to obtain the range of the rock debris slope.
[0042] Step S7 further includes steps S71, S72, S73, and S74, which are specifically as follows: Step S71: Obtain the standard spectral curve of snow; In this step, the standard spectral curve of snow can be obtained through on-site measurement, manual acquisition from spectral images, or querying from a spectral library, such as... Figure 8 The figure shows the standard spectral curve of snow; Step S72: Process the standard spectral curve of the snow with the hyperspectral remote sensing image to obtain the fourth classification image; In this step, such as Figure 9 As shown, supervised classification or spectral matching methods are used to filter out the independent categories of snow pixels in hyperspectral remote sensing images. The classification results are exported as a fourth classification map with the same resolution as the hyperspectral remote sensing image, which is used for the final removal of the debris slope range. This not only solves the core problem of visual confusion between snow and debris slope, but also provides a highly reliable template for subsequent differential removal. At the same time, it adapts to the seasonal needs of high-altitude and cold scenes, and is a key step in achieving accurate extraction of debris slopes in complex scenes.
[0043] Step S73: Perform intersection processing on the first classification map, the second classification map, and the third classification map to obtain the fifth classification map; In this step, such as Figure 10 As shown, the pixels of the fifth classification image must simultaneously meet three core conditions: light gray or grayish-white hue, high brightness of the lithology, and no vegetation cover. Through intersection calculation, an image that simultaneously meets the three conditions is obtained, laying the foundation for the final snow removal range.
[0044] Step S74: Perform a difference operation between the fourth classification map and the fifth classification map to obtain the range of the rock debris slope.
[0045] In this step, such as Figure 11 As shown, snow interference is removed through subtraction, and interference stripping and feature set are finally completed. All core features of the debris slope are aggregated, and a high-purity and high-precision debris slope range is finally output, realizing the complete transformation from hyperspectral raw data to the precise range of the debris slope.
[0046] Example 2 This embodiment provides a system for extracting debris slopes based on hyperspectral remote sensing images. The system includes an acquisition module, a first processing module, a second processing module, a third processing module, a first calculation module, a second calculation module, and a fourth processing module, specifically: The acquisition module is used to acquire hyperspectral remote sensing images; The first processing module is used to perform linear stretching processing on the hyperspectral remote sensing image to obtain a first spectral image; The second processing module is used to perform tone analysis processing based on the first spectral image to obtain a first classification map, the first classification map including areas that conform to the tone characteristics of the rock debris slope. The third processing module is used to process the hyperspectral remote sensing image using band analysis to obtain a second spectral image; The first calculation module is used to perform calculations based on the first spectral image and the second spectral image to obtain a second classification map, the second classification map including regions that conform to the lithological brightness of the rock debris slope; The second calculation module is used to process the hyperspectral remote sensing image to obtain a third classification map, which represents the normalized vegetation index corresponding to each pixel. The fourth processing module is used to process the first classification map, the second classification map and the third classification map to obtain the range of the rock debris slope.
[0047] In one specific embodiment of this disclosure, the first processing module further includes a first processing unit, a first extraction unit, a second processing unit, a third processing unit, a fourth processing unit, and a fifth processing unit, specifically: The first processing unit is used to process the hyperspectral remote sensing image to obtain an RGB image; The first extraction unit is used to extract the first band value, the second band value and the third band value of each pixel in the RGB image; The first calculation unit is used to add the first band value, the second band value and the third band value to obtain the sum of the band values; The second processing unit is used to sort the sum of the band values in ascending order to obtain a band value sequence. The third processing unit is used to truncate the band value sequence from both ends to obtain a first truncation region and a second truncation region, the range of which is two percent of the band value sequence. The fourth processing unit is used to truncate the image based on the first truncation region and the second truncation region to obtain the truncated image; The fifth processing unit is used to linearly stretch the truncated image to obtain a first spectral image.
[0048] In one specific embodiment of this disclosure, the first calculation module further includes a first acquisition unit, a sixth processing unit, a second extraction unit, and a first calculation unit, specifically as follows: The first acquisition unit is used to acquire geological data of the target area; The sixth processing unit is used to obtain the main lithology within the target area based on the geological data of the target area; The second extraction unit is used to extract the standard spectral curves of the main lithologies within the target area; The second calculation unit is used to perform calculations based on the standard spectral curve, the first spectral image, and the second spectral image to obtain a second classification map.
[0049] In one specific embodiment of this disclosure, the second calculation module further includes a third extraction unit, a second calculation unit, a third calculation unit, and a fourth calculation unit, specifically: The third extraction unit is used to extract the near-infrared band reflectance value and the red band reflectance value of the hyperspectral remote sensing image; The third calculation unit is used to calculate the difference between the near-infrared band reflectance value and the red band reflectance value to obtain the first calculation item. The fourth calculation unit is used to calculate the second calculation item by adding the near-infrared band reflectance value and the red band reflectance value; The fifth calculation unit is used to compare the first calculation item with the second calculation item to obtain the third classification diagram.
[0050] In one specific embodiment of this disclosure, the fourth processing module further includes a second acquisition unit, a seventh processing unit, an eighth processing unit, and a ninth processing unit, specifically: The second acquisition unit is used to acquire the standard spectral curve of snow. The seventh processing unit is used to process the standard spectral curve of the snow and the hyperspectral remote sensing image to obtain the fourth classification map; The eighth processing unit is used to perform intersection processing on the first classification map, the second classification map, and the third classification map to obtain the fifth classification map; The ninth processing unit is used to perform a difference operation between the fourth classification map and the fifth classification map to obtain the range of the rock debris slope.
[0051] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.
[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
[0053] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.
Claims
1. A method for extracting debris slopes based on hyperspectral remote sensing images, characterized in that, include: Acquire hyperspectral remote sensing images; A first spectral image is obtained by performing linear stretching on the hyperspectral remote sensing image. Based on the first spectral image, a hue analysis process is performed to obtain a first classification map, which includes regions that conform to the hue characteristics of the rock debris slope. The hyperspectral remote sensing image is processed using band analysis to obtain a second spectral image; A second classification map is obtained by performing calculations based on the first spectral image and the second spectral image. The second classification map includes regions that conform to the lithological brightness of the rock debris slope. The hyperspectral remote sensing image is processed to obtain a third classification map, which represents the normalized vegetation index corresponding to each pixel. The first classification map, the second classification map, and the third classification map are processed to obtain the range of the rock debris slope.
2. The method for extracting debris slopes based on hyperspectral remote sensing images according to claim 1, wherein the hyperspectral remote sensing image is linearly stretched to obtain a first spectral image, characterized in that, include: The hyperspectral remote sensing image is processed to obtain an RGB image; Extract the first band value, second band value, and third band value of each pixel in the RGB image; The sum of the first band value, the second band value, and the third band value is calculated by adding them together. The sum of the band values is sorted in ascending order to obtain a band value sequence; The band value sequence is truncated from both ends to obtain a first truncated region and a second truncated region. The range of the first truncated region and the second truncated region is two percent of the band value sequence. The image is truncated based on the first truncated region and the second truncated region to obtain the truncated image; The truncated image is linearly stretched to obtain a first spectral image.
3. The method for extracting debris slopes based on hyperspectral remote sensing images according to claim 1, wherein a second classification map is obtained by performing calculations based on the first spectral image and the second spectral image, characterized in that, include: Obtain geological data for the target area; Based on the geological data of the target area, the main lithology within the target area is determined; Extract standard spectral curves of the main lithologies within the target area; The second classification map is obtained by calculating based on the standard spectral curve, the first spectral image, and the second spectral image.
4. The method for extracting debris slopes based on hyperspectral remote sensing images according to claim 1, wherein the hyperspectral remote sensing image is processed to obtain a third classification map, characterized in that, include: Extract the near-infrared and red reflectance values from the hyperspectral remote sensing image; The first calculation term is obtained by calculating the difference between the near-infrared band reflectance value and the red band reflectance value. The second calculation term is obtained by adding the near-infrared band reflectance value and the red band reflectance value; The first calculation item is compared with the second calculation item to obtain the third classification map.
5. The method for extracting debris slopes based on hyperspectral remote sensing images according to claim 1, wherein the first classification map, the second classification map, and the third classification map are processed to obtain the extent of the debris slope, characterized in that, include: Obtain the standard spectral curve of snow; The fourth classification image is obtained by processing the standard spectral curve of the snow with the hyperspectral remote sensing image. The intersection of the first classification map, the second classification map, and the third classification map is processed to obtain the fifth classification map; The extent of the debris slope is obtained by subtracting the fourth and fifth classification maps.
6. A system for extracting debris slopes based on hyperspectral remote sensing images, characterized in that, include: The acquisition module is used to acquire hyperspectral remote sensing images; The first processing module is used to perform linear stretching processing on the hyperspectral remote sensing image to obtain a first spectral image; The second processing module is used to perform tone analysis processing based on the first spectral image to obtain a first classification map, the first classification map including areas that conform to the tone characteristics of the rock debris slope. The third processing module is used to process the hyperspectral remote sensing image using band analysis to obtain a second spectral image; The first calculation module is used to perform calculations based on the first spectral image and the second spectral image to obtain a second classification map, the second classification map including regions that conform to the lithological brightness of the rock debris slope; The second calculation module is used to process the hyperspectral remote sensing image to obtain a third classification map, which represents the normalized vegetation index corresponding to each pixel. The fourth processing module is used to process the first classification map, the second classification map and the third classification map to obtain the range of the rock debris slope.
7. The system for extracting rock debris slopes based on hyperspectral remote sensing images according to claim 6, wherein the first processing module is characterized in that, include: The first processing unit is used to process the hyperspectral remote sensing image to obtain an RGB image; The first extraction unit is used to extract the first band value, the second band value and the third band value of each pixel in the RGB image; The first calculation unit is used to add the first band value, the second band value and the third band value to obtain the sum of the band values; The second processing unit is used to sort the sum of the band values in ascending order to obtain a band value sequence. The third processing unit is used to truncate the band value sequence from both ends to obtain a first truncation region and a second truncation region, the range of which is two percent of the band value sequence. The fourth processing unit is used to truncate the image based on the first truncation region and the second truncation region to obtain the truncated image; The fifth processing unit is used to linearly stretch the truncated image to obtain a first spectral image.
8. The system for extracting debris slopes based on hyperspectral remote sensing images according to claim 6, wherein the first calculation module is characterized in that, include: The first acquisition unit is used to acquire geological data of the target area; The sixth processing unit is used to obtain the main lithology within the target area based on the geological data of the target area; The second extraction unit is used to extract the standard spectral curves of the main lithologies within the target area; The second calculation unit is used to perform calculations based on the standard spectral curve, the first spectral image, and the second spectral image to obtain a second classification map.
9. The system for extracting debris slopes based on hyperspectral remote sensing images according to claim 6, wherein the second calculation module is characterized in that, include: The third extraction unit is used to extract the near-infrared band reflectance value and the red band reflectance value of the hyperspectral remote sensing image; The third calculation unit is used to calculate the difference between the near-infrared band reflectance value and the red band reflectance value to obtain the first calculation item. The fourth calculation unit is used to calculate the second calculation item by adding the near-infrared band reflectance value and the red band reflectance value; The fifth calculation unit is used to compare the first calculation item with the second calculation item to obtain the third classification diagram.
10. The system for extracting rock debris slopes based on hyperspectral remote sensing images according to claim 6, wherein the fourth processing module is characterized in that, include: The second acquisition unit is used to acquire the standard spectral curve of snow. The seventh processing unit is used to process the standard spectral curve of the snow and the hyperspectral remote sensing image to obtain the fourth classification map; The eighth processing unit is used to perform intersection processing on the first classification map, the second classification map, and the third classification map to obtain the fifth classification map; The ninth processing unit is used to perform a difference operation between the fourth classification map and the fifth classification map to obtain the range of the rock debris slope.