Water and soil conservation monitoring method and device for power grid project

Abstract

The invention belongs to the technical field of spectral image processing, and particularly relates to a water and soil conservation monitoring method and device for power grid engineering, and the method comprises the steps: obtaining a plurality of spectral images of a target monitoring region at different construction stages; selecting feature points containing different types of typical ground features from the spectral image for extraction, and extracting target feature points of each wave band; carrying out ground feature classification on the spectral image of the target monitoring area, and determining different types of typical ground features contained in the spectral image; registering the plurality of spectral images by using the target feature points; performing time sequence analysis on the plurality of registered spectral images, and calculating dynamic change data of different types of typical ground features in the plurality of spectral images at different construction stages; and based on the dynamic change data, outputting water and soil environment influence evaluation results of different construction stages on the target monitoring area. The problem of spectral image drift in water and soil conservation analysis can be solved, and the accuracy of a water and soil conservation monitoring result is further improved.

Description

Technical Field

[0001] The invention belongs to the technical field of spectral image processing, and in particular relates to a soil and water conservation monitoring method and device for a power grid project. Background Art

[0002] The construction of power transmission lines in the Gobi region has a significant impact on the water and soil environment. First, the construction process requires excavation of land and construction of infrastructure, which will lead to the destruction of surface vegetation and increase the risk of water and wind erosion of the soil. In arid areas such as the Gobi, vegetation recovery is slow and soil is easily lost, leading to land degradation and desertification. Secondly, the use of heavy machinery and the activities of construction workers may disturb the originally fragile ecosystem and increase water resource consumption. Water resources in the Gobi region are already scarce, and the additional water demand during the construction process may make local water resources more tense, affecting the biodiversity and sustainability of water and soil resources in the region.

[0003] The soil and water loss monitoring technology based on remote sensing satellites uses the different advantages of meteorological satellites and resource satellites to obtain surface information in a large area or in areas that require high resolution. Combined with GPS positioning, it monitors the topography, climate, vegetation conditions and land use status of the construction project area in real time to analyze the soil and water loss situation. By comparing the vegetation coverage, monitoring range, earthwork excavation volume and prevention and control measures of remote sensing images, a soil and water loss monitoring database is finally established to achieve comprehensive analysis and effective monitoring of data information.

[0004] The Chinese invention patent with publication number CN113935861A discloses a method for soil and water conservation monitoring, which discloses monitoring the construction disturbance of the construction site at different times, and comparing the multispectral information of each acquisition time with the previous acquisition time, so as to obtain the overall situation of the soil and water disturbance of the construction site. However, there are still technical defects in this method of monitoring soil and water conservation. Due to the strong solar radiation, wind and sand and other environmental factors in the Gobi region, noise is prone to appear in the spectral data during the analysis process, such as image drift. Specifically, image drift refers to the inconsistency in spatial position between images collected in multiple phases or multiple times due to various reasons, that is, the same object appears misplaced or offset in different images. Drift will cause the image to be unable to be accurately superimposed, affecting the accuracy of applications such as time series analysis and object change monitoring. And these noises will gradually accumulate in the time series, affecting the accuracy of the analysis results. Therefore, a method is needed to solve the problem of spectral image drift in soil and water conservation analysis, thereby improving the accuracy of soil and water conservation monitoring results. Summary of the invention

[0005] The purpose of the present invention is to provide a method and device for soil and water conservation monitoring of power grid projects that can solve the problem of spectral image drift in soil and water conservation analysis and thereby improve the accuracy of soil and water conservation monitoring results in response to the above-mentioned problems existing in the prior art.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] In a first aspect, the present invention provides a method for monitoring soil and water conservation in a power grid project, the method comprising:

[0008] Obtain multiple spectral images of the target monitoring area at different construction stages;

[0009] Extracting characteristic points of typical land objects of different types contained in the plurality of spectral images, and extracting target characteristic points of each band;

[0010] Classifying the objects on the plurality of spectral images to determine the typical objects of different types contained in the spectral images;

[0011] Using the target feature points to register the plurality of spectral images;

[0012] Performing time series analysis on the registered multiple spectral images to calculate dynamic change data of typical objects of different types in the multiple spectral images at different construction stages;

[0013] According to the dynamic change data of different types of typical landforms at different construction stages, the water and soil environmental impact assessment results of different construction stages on the target monitoring area are output.

[0014] Optionally, extracting characteristic points of typical land objects of different types from the plurality of spectral images to extract target characteristic points of each band specifically includes:

[0015] In the spectral image, bands corresponding to typical land objects of different types are selected;

[0016] According to the bands corresponding to different types of typical land objects, for each pixel of the spectral image, a spectral index value is calculated to generate a spectral index image containing map features of the typical land objects;

[0017] Segmenting the spectral index image based on a set threshold value, and dividing the pixels of the spectral index image into spectral feature areas corresponding to different types of typical land objects according to the spectral index values;

[0018] Corner point detection is performed on the spectral feature area, and corner points are detected at different scales, so as to extract the target feature points.

[0019] Optionally, calculating the spectral index value for each pixel of the spectral image according to the bands corresponding to different types of typical land objects specifically includes:

[0020] The spectral index value of the spectral image is calculated by the near-infrared band and the red light band to obtain the normalized vegetation index corresponding to the vegetation area in different types of typical objects, which is specifically calculated by the following formula:

[0021]

[0022] Wherein, NDVI is the normalized vegetation index, NIR is the reflectance value of the near infrared band, and RED is the reflectance value of the red light band;

[0023] The spectral index value of the spectral image is calculated by the green light band and the near infrared band to obtain the normalized water index corresponding to the water area in different types of typical landforms, which is specifically calculated by the following formula:

[0024]

[0025] Wherein, NDWI is the normalized water index, NIR is the reflectance value of the near infrared band, and GREEN is the reflectance value of the green light band;

[0026] The spectral index value of the spectral image is calculated by the red light band and the green light band to obtain the normalized bare soil index corresponding to the bare soil area in different types of typical land features, which is specifically calculated by the following formula:

[0027]

[0028] Wherein, NDSI is the normalized bare soil index, RED is the reflectance value of the red light band, and GREEN is the reflectance value of the green light band.

[0029] Optionally, registering the plurality of spectral images by using the target feature points specifically includes:

[0030] Generate feature descriptors for target feature points of each spectral image respectively;

[0031] By calculating the similarity of feature descriptors, finding potential matching points in the reference image that match the target feature points;

[0032] Calculating a geometric transformation matrix between the spectral image and the reference image based on the target feature points and the potential matching points;

[0033] Performing geometric transformation on the spectral image according to the geometric transformation matrix to align it to the coordinate system of the reference image;

[0034] The aligned spectral image is superimposed with the reference image, the target feature point is aligned with the potential matching point, and the registration of multiple spectral images with the reference image is completed.

[0035] Optionally, performing time series analysis on the registered multiple spectral images to calculate dynamic change data of typical objects of different types in the multiple spectral images at different construction stages specifically includes:

[0036] The spectral index of any two adjacent spectral images at different phases is calculated respectively, and the specific calculation is performed by the following formula:

[0037]

[0038]

[0039] Among them, NDVI t1 is the normalized vegetation index at the t1 phase, NIR t1 is the reflectance value of the near infrared band at the t1 phase, RED t1 is the reflectance value of the red light band at the t1 phase, NDVI t2 is the normalized vegetation index at the t2 phase, NIR t2 is the reflectance value of the near infrared band at the t2 phase, RED t2 is the reflectance value of the red light band at the t2 phase, NDVI t1 Represents the normalized water index at the t2 phase; GREEN t1 is the reflectance value of the green light band at the t1 phase, NDVI t2 is the normalized water index at the t2 phase, GREEN t2 is the reflectance value of the green light band at the t2 phase, NIR t2 is the reflectance value of the near-infrared band at the t2 phase; NDSI t1 is the normalized bare soil index, NDSI, at the t1 phase t2 is the normalized bare soil index of the t2 phase, and the t1 phase and the t2 phase are the acquisition phases corresponding to any two adjacent spectral images;

[0040] For any two adjacent spectral image spectral indices, the change rate of different phases is calculated, which is specifically calculated by the following formula:

[0041]

[0042] Among them, ΔNDVI is the change rate of the normalized vegetation index, ΔNDWI is the change rate of the normalized vegetation index, ΔNDSI is the change rate of the normalized vegetation index, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images;

[0043] The change rates of the normalized vegetation index at different phases were aligned with different construction stages, the change rates of the normalized water index at different phases were aligned with different construction stages, and the change rates of the normalized bare soil index at different phases were aligned with different construction stages to obtain the dynamic change data of different types of typical landforms at different construction stages.

[0044] Optionally, the outputting of the water and soil environmental impact assessment results of different construction stages on the target monitoring area according to the dynamic change data of different types of typical landforms at different construction stages specifically includes:

[0045] According to the change rate of different phases, the normalized change index is calculated, which is specifically calculated by the following formula:

[0046]

[0047] Among them, DND is the normalized difference index;

[0048] It is determined whether the normalized change index is less than or equal to a preset threshold value. If it is determined that the normalized change index is less than or equal to the preset threshold value, it is determined that the soil erosion exceeds the preset standard.

[0049] Optionally, after performing corner point detection on the spectral feature region and detecting corner points at different scales to extract the feature points, the method further includes:

[0050] For the extracted target feature points, consistency verification is performed in multiple bands to determine whether the feature points have stable spectral characteristics in different bands; and the spectral images are registered using the feature points with stable spectral characteristics in different bands.

[0051] In a second aspect, the present invention provides a soil and water conservation monitoring device for a power grid project, the device comprising an acquisition module, an extraction module, a classification module, a matching module, a processing module and an output module, wherein:

[0052] The acquisition module is used to acquire multiple spectral images of the target monitoring area at different construction stages;

[0053] The extraction module is used to extract characteristic points of typical objects of different types contained in the plurality of spectral images, and extract target characteristic points of each band;

[0054] The classification module is used to classify the objects in the plurality of spectral images and determine the typical objects of different types contained in the spectral images;

[0055] The matching module is used to register the plurality of spectral images using the target feature points;

[0056] The processing module is used to perform time series analysis on the registered multiple spectral images, and calculate dynamic change data of typical objects of different types in the multiple spectral images at different construction stages;

[0057] The output module is used to output the water and soil environmental impact assessment results of different construction stages on the target monitoring area based on the dynamic change data of different types of typical landforms at different construction stages.

[0058] Optionally, the extraction module is used to select bands corresponding to different types of typical landforms in the spectral image; calculate spectral index values ​​for each pixel of the spectral image according to the bands corresponding to different types of typical landforms, and generate a spectral index image containing map features of typical landforms; segment the spectral index image based on a set threshold, and divide the pixels of the spectral index image into spectral feature areas corresponding to different types of typical landforms according to the spectral index values; perform corner point detection on the spectral feature areas, and detect corner points at different scales, so as to extract the target feature points.

[0059] Optionally, the processing module is used to calculate the spectral index value of the spectral image through the near-infrared band and the red light band to obtain the normalized vegetation index corresponding to the vegetation area in different types of typical landforms, which is specifically calculated by the following formula:

[0060]

[0061] Wherein, NDVI is the normalized vegetation index, NIR is the reflectance value of the near infrared band, and RED is the reflectance value of the red light band;

[0062] The processing module is used to calculate the spectral index value of the spectral image through the green light band and the near infrared band to obtain the normalized water index corresponding to the water area in different types of typical landforms, which is specifically calculated by the following formula:

[0063]

[0064] Wherein, NDWI is the normalized water index, NIR is the reflectance value of the near infrared band, and GREEN is the reflectance value of the green light band;

[0065] The processing module is used to calculate the spectral index value of the spectral image through the red light band and the green light band to obtain the normalized bare soil index corresponding to the bare soil area in different types of typical land features, which is specifically calculated by the following formula:

[0066]

[0067] Wherein, NDSI is the normalized bare soil index, RED is the reflectance value of the red light band, and GREEN is the reflectance value of the green light band.

[0068] Optionally, the matching module is used to generate feature descriptors for target feature points of the spectral image; find potential matching points in the reference image that match the target feature points by calculating the similarity of the feature descriptors; calculate the geometric transformation matrix between the spectral image and the reference image based on the target feature points and the potential matching points; perform geometric transformation on the spectral image according to the geometric transformation matrix to align it to the coordinate system of the reference image; superimpose the aligned spectral image and the reference image to align the target feature points with the potential matching points, and complete the registration of multiple spectral images with the reference image.

[0069] Optionally, the processing module is used to respectively calculate the spectral indexes of any two adjacent spectral images at different phases, specifically by the following formula:

[0070]

[0071]

[0072] Among them, NDVI t1 is the normalized vegetation index at the t1 phase, NIR t1 is the reflectance value of the near infrared band at the t1 phase, RED t1 is the reflectance value of the red light band at the t1 phase, NDVI t2 is the normalized vegetation index at the t2 phase, NIR t2 is the reflectance value of the near infrared band at the t2 phase, RED t2 is the reflectance value of the red light band at the t2 phase, NDWI t1 Represents the normalized water index at the t2 phase; GREEN t1 is the reflectance value of the green light band at the t1 phase, NDWI t2 is the normalized water index at the t2 phase, GREEN t2 is the reflectance value of the green light band at the t2 phase, NIR t2 is the reflectivity value of the near-infrared band at the t2 phase; BDSIt1 is the normalized bare soil index, NDSI, at the t1 phase t2 is the normalized bare soil index of the t2 phase, and the t1 phase and the t2 phase are the acquisition phases corresponding to any two adjacent spectral images;

[0073] The processing module is used to calculate the change rate of the spectral index of any two adjacent spectral images in different time phases, which is specifically calculated by the following formula:

[0074]

[0075] Among them, ΔNDVI is the change rate of the normalized vegetation index, ΔNDWI is the change rate of the normalized vegetation index, ΔNDSI is the change rate of the normalized vegetation index, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images;

[0076] The processing module is used to align the change rate of the normalized vegetation index at different phases with different construction stages, align the change rate of the normalized water body index at different phases with different construction stages, and align the change rate of the normalized bare soil index at different phases with different construction stages, so as to obtain dynamic change data of different types of typical landforms at different construction stages.

[0077] Optionally, the processing module is used to calculate a normalized change index according to the change rates of different phases, specifically by the following formula:

[0078]

[0079] Among them, DND is the normalized difference index;

[0080] The output module is used to determine whether the normalized change index is less than or equal to a preset threshold. If it is determined that the normalized change index is less than or equal to the preset threshold, it is determined that the soil erosion exceeds the preset standard.

[0081] Optionally, the matching module is used to perform consistency verification on the extracted target feature points in multiple bands to determine whether the feature points have stable spectral characteristics in different bands; and to align the spectral image using the feature points with stable spectral characteristics in different bands.

[0082] In a third aspect, the present invention provides an electronic device comprising a processor, a memory, a user interface and a network interface, wherein the memory is used to store computer programs and data, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the computer program stored in the memory so that the electronic device performs the aforementioned method.

[0083] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the aforementioned method when executed by a processor.

[0084] Compared with the prior art, the present invention has the following beneficial effects:

[0085] 1. The soil and water conservation monitoring method for power grid projects of the present invention extracts target feature points of typical landforms of different types, and performs coordinate alignment and spatial correction on images of different time phases through precise matching and geometric transformation of feature points, which not only ensures the consistency of images of different time phases in the same coordinate system, so that the images remain stably aligned in space, but also corrects the image drift problem caused by environmental factors such as wind, sand, and solar radiation, so that the dynamic changes of landforms in each time phase are more real and reliable; when the aligned images are used for time series analysis, the changes of landforms in each time phase can truly reflect the impact of construction, thereby improving the accuracy of the assessment of the impact of soil and water environment in the construction area.

[0086] 2. The soil and water conservation monitoring method for power grid projects described in the present invention calculates spectral index values ​​for specific bands of different types of land features, generates spectral index images with significant land feature characteristics, and divides the images into spectral feature regions of each land feature through threshold segmentation, and then performs multi-scale corner point detection on these spectral feature regions to extract target feature points with significant spatial characteristics. This process can ensure that the selected feature points have stability and discrimination in different phases and different bands, effectively enhances the accuracy and representativeness of the land feature points, and provides a reliable spatial anchor point for subsequent image registration and change detection, thereby improving the accuracy and consistency of the analysis.

[0087] 3. The soil and water conservation monitoring method for power grid projects described in the present invention calculates the spectral index and its change rate of two adjacent time phases, aligns the change of each spectral index with the different construction stages, thereby dynamically capturing the changes in local landform types during construction, effectively revealing the responses and change trends of vegetation, water bodies and bare soil in different construction stages, and providing a quantitative analysis of the impact of different construction stages on soil and water loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 The present invention provides a flow chart of the soil and water conservation monitoring method for power grid projects.

[0089] Figure 2 This is a structural block diagram of the soil and water conservation monitoring device for the power grid project described in the present invention.

[0090] Figure 3 This is a structural block diagram of the soil and water conservation monitoring equipment for the power grid project described in the present invention.

[0091] In the above figure, there is an acquisition module 201; an extraction module 202; a classification module 203; a matching module 204; a processing module 205; an output module 206; a processor 01; a communication bus 302; a user interface 303; a network interface 304; and a memory 305. DETAILED DESCRIPTION

[0092] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0093] The construction of power transmission lines in the Gobi region has a significant impact on the soil and water environment. Land excavation and infrastructure construction during the construction process will destroy the surface vegetation and increase the risk of soil erosion and wind erosion. Due to the drought in the region, vegetation recovery is slow and soil is very easy to lose, further leading to land degradation and desertification. In addition, the interference of heavy machinery and personnel activities during construction may aggravate water resource consumption, affect the already scarce water resources, and thus threaten the sustainability of regional biodiversity and soil and water resources. Soil and water loss monitoring technology based on remote sensing satellites can obtain surface information in real time over a large range or high-resolution area, and combine GPS positioning to monitor the terrain, climate, vegetation and land use changes in the construction project area, helping to establish a soil and water loss monitoring database. However, the lag of remote sensing satellites makes it difficult to grasp the soil and water disturbance in a timely manner. Therefore, multi-spectrometers are used for real-time monitoring to identify changes in soil, vegetation and water bodies based on the spectral reflectance characteristics of different objects. However, the environmental conditions in the Gobi make multi-spectrometer images susceptible to noise interference such as drift, and further improvement of methods is needed to improve the accuracy of soil and water conservation analysis and monitoring.

[0094] See also Figure 1 A method for monitoring soil and water conservation in a power grid project includes the following steps S110-S160:

[0095] S110, using a multi-spectrometer to obtain multiple spectral images of the target monitoring area at different construction stages.

[0096] The multi-spectrometer is connected to the server in communication, and the communication method can be any method such as wired or wireless. The server includes but is not limited to electronic devices such as mobile phones, tablet computers, wearable devices, PCs (Personal Computers), and can also be a background server that runs a soil and water conservation monitoring method for power grid projects. The server can be implemented as an independent server or a server cluster composed of multiple servers. In order to collect the soil and water conservation conditions of the target monitoring area of ​​the power grid project, before collecting the multi-spectral image of the target monitoring area through the multi-spectrometer, the multi-spectrometer needs to be mounted on the flight equipment first, and the flight equipment can fly according to the preset flight trajectory to achieve aerial photography of the target monitoring area and multi-spectral imaging of the target monitoring area.

[0097] The principle of analyzing soil and water disturbance through multispectral images is that different land objects (such as soil, vegetation, water bodies, etc.) have specific reflection and absorption characteristics under different wavelengths of electromagnetic waves. For example, healthy vegetation has a higher reflectivity in the near-infrared band, while soil has a higher reflectivity in the red light band, and water bodies have a strong absorption of near-infrared and red light. Multispectral remote sensing technology captures spectral information of different surface features by imaging these bands, which can be used to analyze changes in soil and water.

[0098] Determine the key objects that need to be monitored (such as vegetation, bare soil, and water bodies), and select spectral bands suitable for the identification of these objects (such as red light, green light, near infrared, etc.) to ensure the pertinence of monitoring needs. Select a suitable multi-spectrometer according to the monitoring needs to ensure that the selected equipment can cover the key bands. For example, commonly used bands are visible light (red, green, and blue) and near-infrared bands, which are helpful for the identification of vegetation, water bodies, and soil. The near infrared band (NIR) generally refers to the electromagnetic spectrum band with a wavelength range of 700 nanometers (nm) to 1400 nanometers (nm), which is adjacent to the red light band of visible light (about 620-700nm). This band is widely used in remote sensing and spectral analysis, and is particularly suitable for the identification and monitoring of objects such as vegetation, water bodies, and soil.

[0099] S120 , extracting feature points of typical land objects of different types from each spectral image, and extracting target feature points of each band.

[0100] For different types of land objects (such as vegetation, water bodies, bare soil, etc.), select bands that reflect the characteristics of these land objects to enhance the discrimination in the subsequent spectral index calculation. For vegetation, the near-infrared band (NIR) and the red band (Red) are usually selected. Vegetation has a higher reflectivity in the near-infrared band and a lower reflectivity in the red band. This difference helps to detect vegetation through indices such as NDVI (Normalized Difference Vegetation Index). Water bodies have strong absorption of the near-infrared and red bands and a relatively high reflectivity in the green band. Therefore, green and near-infrared bands, such as NDWI (Normalized Difference Water Index), are often used to identify water bodies. Bare soil has a higher reflectivity in the red and green bands, and the NDSI (Normalized Difference Bare Soil Index) can be used to identify bare soil areas.

[0101] According to the above band combination, the spectral index is calculated pixel by pixel to distinguish the types of ground objects.

[0102] The spectral index value of the spectral image is calculated by the near-infrared band and the red light band to obtain the normalized vegetation index corresponding to the vegetation area in different types of typical objects. The specific calculation is based on the following formula:

[0103]

[0104] Among them, NDVI is the normalized vegetation index, NIR is the reflectance value of the near infrared band, and RED is the reflectance value of the red light band.

[0105] The spectral reflectance characteristics of vegetation in different bands vary significantly, especially in the red and near-infrared bands. Healthy vegetation has a higher reflectivity in the near-infrared band and a lower reflectivity in the red band. Therefore, the "greenness" of vegetation is reflected by calculating the ratio of the difference and sum of near-infrared and red light.

[0106] The spectral index values ​​of the spectral images are calculated by the green band and the near-infrared band to obtain the normalized water index corresponding to the water area in different types of typical landforms. The specific calculation is done by the following formula:

[0107]

[0108] Among them, NDWI is the normalized water index, NIR is the reflectance value of the near-infrared band, and GREEN is the reflectance value of the green light band.

[0109] Water bodies have strong absorption in the near-infrared band, but relatively high reflectivity in the green band. Therefore, the combination of green light and near-infrared can enhance the contrast between water areas and other landforms, which is particularly suitable for separating water bodies from vegetation or bare soil.

[0110] The spectral index values ​​of the spectral image are calculated by the red light band and the green light band to obtain the bare soil index corresponding to the bare soil area in different types of typical landforms. The specific calculation is based on the following formula:

[0111]

[0112] Wherein, NDSI is the normalized bare soil index, RED is the reflectance value of the red light band, and GREEN is the reflectance value of the green light band.

[0113] The reflectivity of bare soil in the red and green bands is high, and the reflectivity in the red band is usually slightly higher than that in the green band. Therefore, the combination of red and green light can amplify the reflectivity characteristics of bare soil, thereby distinguishing it from vegetation and water bodies.

[0114] S130, using each spectral image to classify the objects in the target monitoring area, and determining the different types of typical objects contained in the spectral image.

[0115] NDVI values ​​range from -1 to 1. High values ​​(close to 1) usually represent healthy vegetation, and low values ​​(close to 0 or negative values) represent water bodies, bare soil, or other non-vegetated areas. Areas with NDWI values ​​greater than 0 usually represent water bodies, and areas less than 0 represent non-water bodies. Areas with higher NDSI values ​​can represent bare soil or sand, while negative areas are generally vegetation or water bodies. According to the numerical range of each spectral index, appropriate thresholds are set for typical features, and each index image is divided into different spectral characteristic areas. For example: NDVI values ​​above 0.3 can be regarded as vegetation areas, while those below 0.1 may be non-vegetated areas. NDWI values ​​greater than 0.2 are regarded as water areas, while those less than 0 are regarded as non-water areas. NDSI values ​​greater than 0.15 can be judged as bare soil, and negative values ​​indicate non-bare soil. Based on the threshold segmentation results, a distribution map is generated for each feature type, marking the regional locations of various features.

[0116] Furthermore, points with significant features in the feature area of ​​the object are extracted through corner point detection, and these points can be used as feature points in subsequent registration and change detection.

[0117] A multi-scale pyramid is built for the spectral feature area, and the image is scaled layer by layer at each layer to generate multiple resolution levels. A corner detection algorithm (such as Harris corner detection, Shi-Tomasi corner detection, or FAST algorithm) is applied to identify corners in each layer. Stable corners are screened out at each scale level to ensure that these corners are consistent in multi-scale images. A feature descriptor is generated for each corner point, which can encode the spectral and spatial information around the corner point for subsequent feature point matching and change analysis.

[0118] Since multi-scale detection may lead to duplicate corner points, it is necessary to deduplicate corner points at different scales and retain only one corner point with a similar position. The deduplicated corner points are screened to remove noise points and false detection points to ensure the stability and accuracy of the feature points. The screened corner points are used as the final target feature point set to provide high-quality feature point data for subsequent image registration, time series analysis, and change detection.

[0119] In one possible implementation, corner point detection is performed on the spectral feature area, and corner points are detected at different scales to extract target feature points; for the extracted target feature points, consistency verification is performed in multiple bands to determine whether the feature points have stable spectral features in different bands; if it is determined that the feature points have stable spectral features in different bands, the target feature points are used to align multiple spectral images.

[0120] Specifically, the extracted target feature points are verified for consistency in different bands to ensure that they have stable spectral characteristics in different bands and to exclude inconsistent points caused by noise or local characteristics. The consistency verification method includes: for each target feature point, obtaining the reflectance value or spectral index value of the point in different bands (such as red light, green light, near infrared, etc.), and calculating the spectral similarity of the feature point between each band. Commonly used similarity measurement methods include: calculating the vector angle between the target feature points in each band, a small angle indicates a high spectral similarity; for any two target feature points, the spectral vector x = (x1, x2, ..., x n ) and y=(y1,y2,...,y n ), where n is the number of bands, and the vector angle is calculated as:

[0121]

[0122] In the above formula, SAD represents the angle between the spectral vectors of any two target feature points in each band, and the output result is the angle between the two vectors (usually in radians). If SAD is close to 0, it means that the two spectral vectors are almost in the same direction and their spectral characteristics are very similar. If SAD is close to 90 degrees or greater, it means that the directions of the two spectral vectors are quite different and their spectral characteristics are obviously different.

[0123] The angle threshold is set according to the actual situation. For example, the SAD is set to be less than a certain angle threshold (such as 10 degrees) as the standard for spectral feature stability. Feature points with consistent spectral features in different bands are retained, and those with unstable spectra are removed to ensure the reliability of feature points.

[0124] S140, registering multiple spectral images using target feature points.

[0125] In a possible implementation, multiple spectral images are registered using target feature points, specifically including: generating feature descriptors for the target feature points of each spectral image respectively; finding potential matching points in the reference image that match the target feature points by calculating the similarity of the feature descriptors; calculating the geometric transformation matrix between the spectral image and the reference image based on the target feature points and the potential matching points; performing a geometric transformation on the spectral image according to the geometric transformation matrix to align it to the coordinate system of the reference image; superimposing the aligned spectral image and the reference image to align the target feature points and the potential matching points; and achieving registration of multiple spectral images with the reference image.

[0126] Specifically, a feature descriptor is generated for each target feature point. A descriptor is a vector containing texture or spectral information around a feature point, which is used to represent the local feature information of the feature point. Descriptors can be generated using algorithms such as SIFT, SURF, or ORB. For example, the SIFT descriptor calculates the gradient direction of the area around the feature point to generate a 128-dimensional vector containing position and direction information, and normalizes the descriptor to reduce the impact of illumination and scale changes, ensuring the consistency of the descriptor in images of different phases.

[0127] By calculating the similarity of feature descriptors, the most similar potential matching points in the reference image are found for each target feature point. Common similarity calculation methods include Euclidean distance, cosine similarity or Hamming distance. The K-nearest neighbor algorithm (KNN) is used to find the nearest neighbor points of the target feature points in the reference image to form preliminary matching point pairs. A distance threshold or ratio threshold can be set to ensure that the point pairs with the smallest distance have a higher matching degree.

[0128] According to the image acquisition conditions and required accuracy, select a suitable geometric transformation model. Common transformation models include: rigid transformation model, which only includes translation and rotation and is suitable for stable image pairs; similarity transformation model, which includes scaling, translation and rotation and is suitable for images with slight changes in scale; affine transformation model, which supports rotation, scaling, translation and tilt and is suitable for images with slight deformation; perspective transformation model, which is suitable for situations with large changes in viewing angle and can handle perspective deformation of images.

[0129] Using the selected feature point pairs, the geometric transformation matrix of the image is solved according to the selected transformation model. For example, in affine transformation, the coordinates of the point pairs are fitted using the least squares method to obtain the parameters of the affine matrix.

[0130] According to the calculated geometric transformation matrix, the spectral image to be registered is geometrically transformed to align it to the coordinate system of the reference image. After the geometric transformation, the pixel values ​​of the transformed image are resampled. Commonly used interpolation methods include: nearest neighbor interpolation, which is suitable for classified images and retains the original pixel values; bilinear interpolation, which is suitable for continuous image data and makes the transformed image smoother; cubic spline interpolation, which retains details and reduces blur in high-precision image registration.

[0131] Superimpose the aligned spectral image with the reference image, check the alignment of the target feature points and potential matching points, and ensure that the features of the objects in different phases are consistent. Calculate the mean square error of the registration point pairs to evaluate the registration accuracy. When the mean square error is small, it means that the registration effect is good. Manually check the registration results to confirm whether the key objects are accurately aligned. For areas with large errors, manually adjust the feature points or transformation matrix for correction. Repeat the above steps for each phase of the spectral image and align it with the reference image. By aligning each image one by one, ensure that all images are aligned with the reference image in the same coordinate system.

[0132] S150, performing time series analysis on the registered multiple spectral images, and calculating dynamic change data of typical ground objects of different types in the multiple spectral images at different construction stages.

[0133] Based on the spectral indexes (NDVI, NDWI, NDSI) of different phases calculated in step S120, the ground feature of each phase is obtained, and the red, near-infrared and green band data are extracted from the spectral image. The spectral indexes of any two adjacent spectral images in different phases are calculated respectively, and are specifically calculated by the following formula:

[0134]

[0135] Among them, NDVI t1 is the normalized vegetation index at the t1 phase, NIR t1 is the reflectance value of the near infrared band at the t1 phase, RED t1 is the reflectance value of the red light band at the t1 phase, NDVI t2 is the normalized vegetation index at the t2 phase, NIR t2 is the reflectance value of the near infrared band at the t2 phase, RED t2 is the reflectance value of the red light band at the t2 phase, NDWI t1 Represents the normalized water index at the t2 phase; GREEN t1 is the reflectance value of the green light band at the t1 phase, NDWI t2 is the normalized water index at the t2 phase, GREEN t2is the reflectance value of the green light band at the t2 phase, NIR t2 is the reflectivity value of the near-infrared band at the t2 phase; BDSI t1 is the normalized bare soil index, NDSI, at the t1 phase t2 is the normalized bare soil index of the t2 phase, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images.

[0136] Then, the change rate of the spectral index of adjacent phases is calculated to quantify the dynamic changes of the ground objects between different phases. For any two adjacent spectral image spectral indices, the change rate of different phases is calculated, which is specifically calculated by the following formula:

[0137]

[0138] Among them, ΔNDVI is the change rate of the normalized vegetation index, ΔNDWI is the change rate of the normalized vegetation index, ΔNDSI is the change rate of the normalized vegetation index, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images.

[0139] A positive change rate indicates that the index increased in t2 compared to t1; a negative change rate indicates that the index decreased in t2 compared to t1; the absolute value of the change rate reflects the intensity of the change. The change rate of NDVI reflects the dynamic changes of vegetation, the change rate of NDWI reflects the dynamic changes of water bodies, and the change rate of NDSI reflects the changes in bare soil areas.

[0140] Finally, the change rate of the normalized change index of different phases is aligned with the construction stage, which is convenient for analyzing the impact of construction activities on typical landforms. According to the construction progress, the collected phases are divided into different construction stages (such as before construction, early construction, mid-construction, and after construction). The change rate of the normalized change index of each phase is aligned with the corresponding construction stage, specifically: the change rate of the normalized vegetation index of different phases is aligned with different construction stages, the change rate of the normalized water index of different phases is aligned with different construction stages, and the change rate of the normalized bare soil index of different phases is aligned with different construction stages, so as to establish the association between landform changes and construction activities. Record the change rate of the normalized change index under each construction stage and other data to form a dynamic change data table of different phases and construction stages.

[0141] S160, based on the dynamic change data of typical landforms of different types, output the water and soil environmental impact assessment results of the construction on the target monitoring area. Then use the normalized change index (DND) to comprehensively reflect the dynamic changes of landform types and provide a unified index for landform changes in different phases. Calculate the normalized change index based on the change rate of different phases, specifically using the following formula:

[0142]

[0143] Among them, DND is the normalized difference change index. The DND value combines the change rates of vegetation, water bodies and bare soil, and is used to reflect the dynamic change trend of the overall landform. A positive DND value usually indicates an increase in vegetation or an expansion of water bodies, while a decrease in bare soil. A negative DND value indicates a decrease in vegetation, a decrease in water bodies, and an increase in bare soil, which may reflect problems such as soil erosion or vegetation degradation. When it is judged that the normalized difference change index is less than or equal to the preset threshold, it is determined that the soil erosion exceeds the preset standard.

[0144] Embodiment 2:

[0145] See also Figure 2 A soil and water conservation monitoring device for a power grid project includes an acquisition module 201, an extraction module 202, a classification module 203, a matching module 204, a processing module 205 and an output module 206, wherein the acquisition module 201 is used to acquire a plurality of spectral images of a multi-spectrometer at different construction stages for a target monitoring area; the extraction module 202 is used to respectively select characteristic points of typical objects of different types from each spectral image to extract target characteristic points of each band; the classification module 203 is used to classify objects in the target monitoring area using each spectral image and determine the different types of typical objects contained in the spectral image; the matching module 204 is used to register the plurality of spectral images using the target characteristic points; the processing module 205 is used to perform time series analysis on the registered plurality of spectral images and calculate the dynamic change data of the typical objects of different types in the plurality of spectral images at different construction stages; the output module 206 is used to output the water and soil environmental impact assessment result of the construction on the target monitoring area according to the dynamic change data of the typical objects of different types.

[0146] In a possible implementation, the extraction module 202 is used to perform corner point detection on the spectral feature region, detect corner points at different scales, and thus extract target feature points. The classification module 203 is used to select bands corresponding to different types of typical landforms, calculate the spectral index value for each pixel of the spectral image according to the bands corresponding to different types of typical landforms, and generate a spectral index image containing map features of typical landforms; and to segment the spectral index image based on a set threshold value, and divide the pixels of the spectral index image into spectral feature regions corresponding to different types of typical landforms.

[0147] In a possible implementation, the processing module 205 is used to calculate the spectral index value of the spectral image through the near-infrared band and the red light band to obtain the normalized vegetation index corresponding to the vegetation area in different types of typical land objects, which is specifically calculated by the following formula:

[0148]

[0149] Among them, NDVI is the normalized vegetation index, NIR is the reflectance value of the near infrared band, and RED is the reflectance value of the red light band.

[0150] The processing module 205 is used to calculate the spectral index value of the spectral image through the green light band and the near infrared band to obtain the normalized water index corresponding to the water area in different types of typical landforms, which is specifically calculated by the following formula:

[0151]

[0152] Among them, NDWI is the normalized water index, NIR is the reflectance value of the near-infrared band, and GREEN is the reflectance value of the green light band.

[0153] The processing module 205 is used to calculate the spectral index value of the spectral image through the red light band and the green light band to obtain the normalized bare soil index corresponding to the bare soil area in different types of typical land features, which is specifically calculated by the following formula:

[0154]

[0155] Wherein, NDSI is the normalized bare soil index, RED is the reflectance value of the red light band, and GREEN is the reflectance value of the green light band.

[0156] In one possible implementation, the matching module 204 is used to generate feature descriptors for the target feature points of each spectral image, find potential matching points in the reference image that match the target feature points by calculating the similarity of the feature descriptors, calculate the geometric transformation matrix between the spectral image and the reference image based on the target feature points and the potential matching points, perform geometric transformation on the spectral image according to the geometric transformation matrix, align it to the coordinate system of the reference image, superimpose the aligned spectral image and the reference image, align them with the target feature points and the potential matching points, and thus complete the registration of the spectral image and the reference image.

[0157] In a possible implementation, the processing module 205 is used to respectively calculate the spectral indexes of any two adjacent spectral images at different phases, specifically by the following formula:

[0158]

[0159]

[0160] Among them, NDVI t1 is the normalized vegetation index at the t1 phase, NIR t1is the reflectance value of the near infrared band at the t1 phase, RED t1 is the reflectance value of the red light band at the t1 phase, NDVI t2 is the normalized vegetation index at the t2 phase, NIR t2 is the reflectance value of the near infrared band at the t2 phase, RED t2 is the reflectance value of the red light band at the t2 phase, NDWI t1 Represents the normalized water index at the t2 phase; GREEN t1 is the reflectance value of the green light band at the t1 phase, NDWI t2 is the normalized water index at the t2 phase, GREEN t2 is the reflectance value of the green light band at the t2 phase, NIR t2 is the reflectivity value of the near-infrared band at the t2 phase; BDSI t1 is the normalized bare soil index, NDSI, at the t1 phase t2 is the normalized bare soil index of the t2 phase, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images.

[0161] The processing module 205 is used to calculate the change rate of the spectral index of any two adjacent spectral images in different time phases, which is specifically calculated by the following formula:

[0162]

[0163] Among them, ΔNDVI is the change rate of the normalized vegetation index, ΔNDWI is the change rate of the normalized vegetation index, ΔNDSI is the change rate of the normalized vegetation index, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images.

[0164] Processing module 205 is used to align the change rate of the normalized vegetation index at different phases with different construction stages, align the change rate of the normalized water index at different phases with different construction stages, and align the change rate of the normalized bare soil index at different phases with different construction stages, so as to obtain dynamic change data of different types of typical landforms at different construction stages.

[0165] In a possible implementation, the processing module 205 is used to calculate the normalized change index according to the change rates of different phases, specifically by the following formula:

[0166]

[0167] Among them, DND is the normalized difference index.

[0168] The output module 206 is used to determine whether the normalized change index is less than or equal to a preset threshold. If it is determined that the normalized change index is less than or equal to the preset threshold, it is determined that the soil erosion exceeds the preset standard.

[0169] In a possible implementation, the matching module 204 is used to perform consistency verification on the extracted target feature points in multiple bands to determine whether the feature points have stable spectral characteristics in different bands; and to perform subsequent alignment using the target feature points with stable spectral characteristics in different bands.

[0170] It should be noted that: when the device provided in the above embodiment realizes its function, it only uses the division of the above-mentioned functional modules as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0171] Embodiment 3:

[0172] Reference Figure 3 , an electronic device, including at least one processor 301, at least one communication bus 302, a user interface 303, a network interface 304, and at least one memory 305. Among them, the communication bus 302 is used to realize the connection and communication between these components. The user interface 303 may include a display screen and a camera. The user interface 303 may include a standard wired interface and a wireless interface (such as a WIFI interface). The network interface 304 may include a standard wired interface and a wireless interface. The processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the electronic device by running or executing a computer program stored in the memory 305, and calling data stored in the memory 305. As an implementation method, the processor 301 can be implemented in at least one hardware form of a digital signal processing (DSP), a field programmable gate array (FPGA), and a programmable logic array (PLA). The processor 301 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU mainly processes the operating system, the user interface, and the application program; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; and the modem is used to process wireless communication. The memory 305 includes a program storage area and a data storage area, wherein the program storage area stores a computer program for implementing an operating system, a computer program for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), a computer program for implementing the method described in Example 1, etc.; the data storage area stores data involved in the process of implementing the method described in Example 1.

[0173] Embodiment 4:

[0174] A computer-readable storage medium stores a computer program, wherein the computer program implements the method described in Example 1 when executed by a processor.

[0175] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0176] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0177] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0178] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0179] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A soil and water conservation monitoring method for a power grid project, characterized in that: The method comprises: Obtain multiple spectral images of the target monitoring area at different construction stages; Extracting characteristic points of typical land objects of different types contained in the plurality of spectral images, and extracting target characteristic points of each band; Classifying the objects on the plurality of spectral images to determine the typical objects of different types contained in the spectral images; Using the target feature points to register the plurality of spectral images; Performing time series analysis on the registered multiple spectral images to calculate dynamic change data of typical objects of different types in the multiple spectral images at different construction stages; According to the dynamic change data of different types of typical landforms at different construction stages, the water and soil environmental impact assessment results of different construction stages on the target monitoring area are output.

2. The method for monitoring soil and water conservation of a power grid project according to claim 1, characterized in that: The step of extracting characteristic points of typical objects of different types from the plurality of spectral images to extract target characteristic points of each band specifically includes: In the spectral image, bands corresponding to typical land objects of different types are selected; According to the bands corresponding to different types of typical land objects, for each pixel of the spectral image, a spectral index value is calculated to generate a spectral index image containing map features of the typical land objects; Segmenting the spectral index image based on a set threshold value, and dividing the pixels of the spectral index image into spectral feature areas corresponding to different types of typical land objects according to the spectral index values; Corner point detection is performed on the spectral feature area, and corner points are detected at different scales, so as to extract the target feature points.

3. The method for monitoring soil and water conservation of a power grid project according to claim 2, characterized in that: The calculating of the spectral index value for each pixel of the spectral image according to the bands corresponding to the typical objects of different types specifically includes: The spectral index value of the spectral image is calculated by the near-infrared band and the red light band to obtain the normalized vegetation index corresponding to the vegetation area in different types of typical objects, which is specifically calculated by the following formula: Wherein, NDVI is the normalized vegetation index, NIR is the reflectance value of the near infrared band, and RED is the reflectance value of the red light band; The spectral index value of the spectral image is calculated by the green light band and the near infrared band to obtain the normalized water index corresponding to the water area in different types of typical landforms, which is specifically calculated by the following formula: Wherein, NDWI is the normalized water index, NIR is the reflectance value of the near infrared band, and GREEN is the reflectance value of the green light band; The spectral index value of the spectral image is calculated by the red light band and the green light band to obtain the bare soil index corresponding to the bare soil area in different types of typical land features, which is specifically calculated by the following formula: Wherein, NDSI is the normalized bare soil index, RED is the reflectance value of the red light band, and GREEN is the reflectance value of the green light band.

4. The method for monitoring soil and water conservation of a power grid project according to claim 1, characterized in that: The registering of the plurality of spectral images by using the target feature points specifically includes: Generate feature descriptors for target feature points of each spectral image respectively; By calculating the similarity of feature descriptors, finding potential matching points in the reference image that match the target feature points; Calculating a geometric transformation matrix between the spectral image and the reference image based on the target feature points and the potential matching points; Performing geometric transformation on the spectral image according to the geometric transformation matrix to align it to the coordinate system of the reference image; The aligned spectral image is superimposed with the reference image, the target feature point is aligned with the potential matching point, and the registration of multiple spectral images with the reference image is completed.

5. The method for monitoring soil and water conservation in a power grid project according to claim 1, characterized in that: The performing of time series analysis on the registered multiple spectral images to calculate dynamic change data of typical objects of different types in the multiple spectral images at different construction stages specifically includes: The spectral index of any two adjacent spectral images at different phases is calculated respectively, and the specific calculation is performed by the following formula: Among them, NDVI t1 is the normalized vegetation index at the t1 phase, NIR t1 is the reflectance value of the near infrared band at the t1 phase, RED t1 is the reflectance value of the red light band at the t1 phase, NDVI t2 is the normalized vegetation index at the t2 phase, NIR t2 is the reflectance value of the near infrared band at the t2 phase, RED t2 is the reflectance value of the red light band at the t2 phase, NDWI t1 Represents the normalized water index at the t2 phase; GREEN t1 is the reflectance value of the green light band at the t1 phase, NDWI t2 is the normalized water index at the t2 phase, GREEN t2 is the reflectance value of the green light band at the t2 phase, NIR t2 is the reflectance value of the near-infrared band at the t2 phase; NDSI t1 is the normalized bare soil index, NDSI, at the t1 phase t2 is the normalized bare soil index of the t2 phase, and the t1 phase and the t2 phase are the acquisition phases corresponding to any two adjacent spectral images; For any two adjacent spectral image spectral indices, the change rate of different phases is calculated, which is specifically calculated by the following formula: Among them, △NDVI is the change rate of the normalized vegetation index, △NDWI is the change rate of the normalized vegetation index, △NDSI is the change rate of the normalized vegetation index, and the t1 phase and t2 phase are the acquisition phases corresponding to any two adjacent spectral images; The change rates of the normalized vegetation index at different phases were aligned with different construction stages, the change rates of the normalized water index at different phases were aligned with different construction stages, and the change rates of the normalized bare soil index at different phases were aligned with different construction stages to obtain the dynamic change data of different types of typical landforms at different construction stages.

6. The method for monitoring soil and water conservation in a power grid project according to claim 5, characterized in that: The output of the water and soil environmental impact assessment results of different construction stages on the target monitoring area according to the dynamic change data of different types of typical landforms at different construction stages specifically includes: According to the change rate of different phases, the normalized change index is calculated, which is specifically calculated by the following formula: Among them, DND is the normalized difference index; It is determined whether the normalized change index is less than or equal to a preset threshold value. If it is determined that the normalized change index is less than or equal to the preset threshold value, it is determined that the soil erosion exceeds the preset standard.

7. The method for monitoring soil and water conservation in a power grid project according to claim 2, characterized in that: After performing corner point detection on the spectral feature region and detecting corner points at different scales to extract the feature points, the method further includes: For the extracted target feature points, consistency verification is performed in multiple bands to determine whether the feature points have stable spectral characteristics in different bands; and the spectral images are registered using the feature points with stable spectral characteristics in different bands.

8. A soil and water conservation monitoring device for a power grid project, characterized in that: The device comprises an acquisition module (201), an extraction module (202), a classification module (203), a matching module (204), a processing module (205) and an output module (206), wherein: The acquisition module (201) is used to acquire multiple spectral images of the target monitoring area at different construction stages; The extraction module (202) is used to extract characteristic points of typical land objects of different types contained in the plurality of spectral images, and extract target characteristic points of each band; The classification module (203) is used to classify the ground objects of the plurality of spectral images, and determine the different types of typical ground objects contained in the spectral images; The matching module (204) is used to register the plurality of spectral images using the target feature points; The processing module (205) is used to perform time series analysis on the multiple registered spectral images, and calculate dynamic change data of typical ground objects of different types in the multiple spectral images at different construction stages; The output module (206) is used to output the water and soil environmental impact assessment results of different construction stages on the target monitoring area based on the dynamic change data of different types of typical landforms at different construction stages.

9. An electronic device, characterized in that: The electronic device comprises a processor (301), a communication bus (302), a memory (305), a user interface (303), a network interface (304) and a memory (305), wherein the memory (305) is used to store computer programs and data, the user interface (303) and the network interface (304) are both used to communicate with other devices, and the processor (301) is used to execute the computer program stored in the memory (305) so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is performed.

Images