Hydropower station biological control method and device based on disease and pest recognition

Abstract

The invention relates to the technical field of hydropower station ecological protection, in particular to a hydropower station biological control method and device based on disease and pest recognition, and the method comprises the following steps: S1, collecting an initial image set of the equipment surface and surrounding environment of a hydropower station through a multispectral camera device; s2, performing multi-band fusion enhancement processing on the initial image set to generate a fusion enhanced image; s3, constructing an insect pest distribution vector diagram; s4, determining a target disease and pest type and a corresponding biological natural enemy type; s5, calculating a putting coordinate and a putting quantity, and forming a target prevention and control scheme; and S6, executing the prevention and control scheme through a directional launching device carried by the unmanned aerial vehicle. According to the invention, through linkage of insect pest image identification and spatial distribution analysis with natural enemy release control, accurate identification and efficient biological control of hydropower station insect pests are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological protection of hydropower stations, and in particular to a method and device for biological control of hydropower stations based on pest identification. Background Art

[0002] As the scale of operation of large-scale hydropower stations continues to expand, the operating environment of their equipment has gradually exposed the potential impact of ecological interference factors on equipment performance; among them, insect pests, as a typical biological destructive factor, have gradually become a key hidden danger affecting the operational stability of hydropower stations and the life of equipment; on the one hand, insect bodies and insect swarms are prone to gather on the surface of equipment, line grooves, sensor elements and other areas, resulting in reduced equipment heat dissipation performance, signal interference, insulation aging and even corrosion and perforation; on the other hand, some aquatic insects and wetland pests will show explosive growth under specific humid and hot climatic conditions, which makes manual identification difficult and response delayed, resulting in untimely prevention and control and high costs.

[0003] Existing pest control methods primarily rely on manual inspections and standardized pesticide spraying, which struggle to meet the comprehensive safety, regional diversity, and environmental friendliness requirements of hydropower stations. Technically, there is currently a lack of a systematic approach that can automatically identify pests, locate clusters, determine pest species, and precisely release their corresponding natural enemies. This is particularly true in environments with a complex mix of insects, small, and dispersed targets, making it difficult for existing solutions to achieve high-precision identification and precise intervention. Therefore, a hydropower station biological control method and device based on pest identification is urgently needed to address these issues. Summary of the Invention

[0004] Based on the above objectives, the present invention provides a method and device for biological control in hydropower stations based on pest and disease identification.

[0005] A method for biological control of a hydropower station based on pest and disease identification comprises the following steps: S1: An initial set of images of the surface of the hydropower station equipment and its surrounding environment is collected using a multispectral camera device, where the multispectral image includes visible light bands and near-infrared bands; S2: Perform multi-band fusion enhancement processing on the initial image set, calculate the color contrast index based on the visible light band; calculate the topological connectivity index based on the near-infrared band, and generate a fusion enhanced image; S3: Extract the pigment deposition area of ​​the insect target based on the color contrast index, and divide the insect swarm core area and diffusion edge area into two parts based on the topological connectivity index to form a pest distribution vector map. S4: Match the pest distribution vector map with the pre-built hydropower station pest and disease feature database to determine the target pest and disease species and the corresponding biological natural enemy types; S5: Determine the priority level of natural enemy deployment based on the spatial weight relationship between the core area of ​​insect swarm aggregation and the diffusion edge area. Calculate the deployment coordinates and quantity based on the pest density distribution to form a targeted control plan. S6: Implement the prevention plan through the directional delivery device carried by the drone.

[0006] Optionally, the S1 specifically includes: S11: A fixed or mobile data acquisition platform equipped with a multispectral imaging module is deployed inside and outside the hydropower station. The acquisition range is set to cover the main engine surface, transmission facilities, turbine blade casing, and the surrounding environment within 10 meters. S12: Start the multispectral camera device, collect image data including visible light band and near-infrared band in sequence according to preset band parameters, and align the images of different bands in time sequence, control the single exposure time within 30 ms, and maintain the frame rate at no less than 25 fps; S13: Mark the multispectral image acquisition results according to the shooting timestamp and the device identification number to form an initial image set with spatial location information.

[0007] Optionally, the S2 specifically includes: S21: performing pixel-level brightness normalization processing on the visible light band images in the initial image set; S22: Calculate the color contrast index for the normalized visible light image. The formula is: ,in, represents the color contrast index, represents the total number of pixels in the image, Respectively The grayscale value of each pixel in the red, green and blue channels; S23: Binarize the near-infrared band image and use the Otsu automatic threshold segmentation algorithm to distinguish the insect body area from the background area. Then, establish the insect body connectivity map and calculate the topological connectivity index. The formula is: ,in, represents the topological connectivity index, represents the pixel area of ​​the largest connected insect body region, It represents the total area of ​​all insect body regions; S24: Perform weighted fusion of the color contrast index and the topological connectivity index to generate a fused enhanced image.

[0008] Optionally, the S3 specifically includes: S31: Based on the color contrast index obtained in S2, the color contrast values ​​of different pixels in the fused enhanced image are subjected to threshold segmentation processing, and the high contrast area exceeding the threshold is determined as the pigment deposition area of ​​the insect target; S32: Based on the topological connectivity index calculated in S2, the connected region analysis is performed on the pigment deposition area of ​​the insect body target, and a comprehensive evaluation is performed based on the connectivity and area size within the area to distinguish the core area of ​​insect swarm aggregation from the edge area of ​​insect swarm diffusion; S33: Extract the boundaries of the insect swarm core area and the diffusion edge area respectively, establish the geometric outline of the insect pest area based on the spatial coordinates, and convert it into a vector representation to obtain a vector map of the insect pest distribution.

[0009] Optionally, the S32 specifically includes: S321: Perform 8-neighborhood connected region labeling processing on the pigment deposition area image extracted in S31, identify all connected insect body target areas, and record the area, perimeter, center of gravity coordinates and boundary contact length with adjacent areas of each connected area; S322: Calculate the connectivity index of each connected area. The formula is: ,in, Indicates the The connectivity of the target area of ​​the insect body, represents the pixel area of ​​the region, Indicates the boundary perimeter of the area; S323: Setting the connectivity threshold , will satisfy The connected area is defined as the core area of ​​insect swarm gathering; the rest of the area that meets The area is designated as the swarm spread edge zone.

[0010] Optionally, the S4 specifically includes: S41: Perform scale normalization on the pest distribution vector map generated in S3, unifying the vector contour data of the insect swarm core area and the diffusion edge area to the same spatial scale as the hydropower station pest and disease feature database; S42: Extracting geometric characteristic parameters of pest vectors from the pest distribution vector map, including the average area, boundary perimeter, regional shape factor, and spatial layout pattern index of the target insect area, as input features for matching analysis; S43: calling a pre-stored standard vector feature template in a pre-built hydropower station pest and disease feature library, and performing feature matching between the geometric feature parameters of the pest vector to be identified and the template in the feature library using a cosine similarity algorithm to obtain a matching similarity score; S44: Select the standard feature template with the highest similarity score and a score greater than 0.85, determine the corresponding target pest species, and obtain the biological natural enemy species that match the pest species through the pest record information in the hydropower station pest feature database.

[0011] Optionally, the S43 specifically includes: S431: Combining the geometric feature parameters of the pest vector extracted in S42 into a feature vector , represent the average area, boundary perimeter, regional shape factor and spatial layout pattern index respectively; S432: Extract the corresponding feature vectors of the standard feature template from the hydropower station pest feature database, and record them as ,in Indicates the Standard feature templates; S433: For each standard feature template, use the cosine similarity algorithm to calculate the feature vector to be identified With the standard feature template vector The cosine similarity between them is: ,in, Represents the feature vector to be identified and the The similarity score between the standard feature template vectors ranges from 0 to 1; The first vector diagram of the distribution of pests to be identified is extracted. geometric characteristic parameter components; Indicates the first The first standard template Geometric characteristic parameters; It is the serial index of the geometric feature parameter, with a value of 1 to 4.

[0012] Optionally, the S5 specifically includes: S51: Based on the pest distribution vector map obtained in S3, calculate the area of ​​the insect swarm core area and the diffusion edge area, respectively, and record them as the area of ​​the core area. and the diffusion edge area , and use this to determine the spatial weight relationship and calculate the spatial weight value of the cluster core area and the spatial weight value of the diffuse edge area , the expressions are: and ; S52: Based on the spatial weight value obtained in S51, determine the priority level of natural enemy release; S53: Recommended number of natural enemies corresponding to target pests and diseases identified in S4 , combined with the spatial weight value and regional area calculated by S51, calculate the actual number of deployments in each region , the formula is: ,in, Indicates the The number of natural enemies released in each area, represents the spatial weight value of the area, Indicates the area of ​​the region; S54: Based on the geometric center coordinates of the insect swarm gathering core area and the diffusion edge area as the release coordinates, combined with the release quantity calculated in S53, a target prevention and control plan including regional release priority, release coordinates and corresponding release quantity is formed.

[0013] Optionally, the priority levels of natural enemy release include: When the spatial weight of the core area is gathered When the concentration core area is set as the priority level 1 area, the diffusion edge area is set as the priority level 2 area; when When the concentration core area and the diffusion edge area are both set as priority level 2 areas; when When the diffusion edge area is set as the priority level 1 area, the gathering core area is set as the priority level 2 area.

[0014] A biocontrol device for a hydropower station based on pest and disease identification, used to implement the aforementioned biocontrol method for a hydropower station based on pest and disease identification, comprises the following modules: Image acquisition module: used to collect image data of the surface of hydropower station equipment and its surrounding environment through a multispectral camera device, the multispectral image includes visible light band and near infrared band, and output the initial image set; Image processing module: connected to the image acquisition module, used to perform multi-band fusion enhancement processing on the initial image set, wherein the color contrast index is calculated based on the visible light band, the topological connectivity index is calculated based on the near-infrared band, and a fusion enhanced image is generated; Pest extraction module: This module is connected to the image processing module and is used to extract the pigment deposition area of ​​the insect target based on the color contrast index. It also divides the insect swarm into the core area and the diffusion edge area based on the topological connectivity index to form a vector map of the insect pest distribution. Pest and disease identification module: This module is connected to the pest extraction module and is used to match the geometric characteristic parameters of the pest distribution vector with the standard templates in the pre-built hydropower station pest and disease feature library. It uses cosine similarity calculation to output the matching results and determine the target pest and disease species and their corresponding biological natural enemy types. Control decision-making module: connected to the pest identification module, it is used to calculate the priority level, coordinates and quantity of natural enemy release based on the spatial weight relationship between the core area of ​​insect clusters and the diffusion edge area, the distribution of insect pests and natural enemy information, and generate a target biological control plan; Execution control module: connected to the prevention and control decision module, used to drive the UAV platform to load the directional delivery device and complete the natural enemy delivery operation in the target area according to the prevention and control plan.

[0015] Beneficial effects of the present invention: The present invention, by constructing a pest identification and spatial aggregation analysis mechanism based on multispectral images, can realize the automatic extraction of pest targets in hydropower stations, the effective division of the core area of ​​insect swarm aggregation and the diffusion edge area, and complete the identification of pest species and the intelligent matching of their biological natural enemies by combining with the standard pest feature library, thus solving the problems of low identification efficiency, strong subjectivity in judgment, and difficulty in distinguishing pest species in traditional manual detection methods.

[0016] The present invention forms a quantitative natural enemy deployment strategy by utilizing spatial weight calculation and pest density distribution, and implements it through drone-directed execution, achieving closed-loop linkage control from image acquisition and feature recognition to prevention and control deployment, thereby improving the response speed of biological control measures, regional adaptability and the accuracy of resource deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 Schematic diagram of a biocontrol method for a hydropower station according to an embodiment of the present invention; Figure 2 Schematic diagram of a biological control device for a hydropower station according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0020] It should be noted that references in the specification to "one embodiment," "an embodiment," "exemplary embodiments," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not necessarily every embodiment will include such specific features, structures, or characteristics. Furthermore, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0021] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0022] like Figure 1 As shown, a method for biological control of a hydropower station based on pest identification includes the following steps: S1: An initial set of images of the surface of the hydropower station equipment and its surrounding environment is collected using a multispectral camera device, where the multispectral image includes visible light bands and near-infrared bands; S2: Perform multi-band fusion enhancement processing on the initial image set. The color contrast index is calculated based on the visible light band to highlight the pigment differences of the insect body. The topological connectivity index is calculated based on the near-infrared band to reflect the connectivity structure characteristics of the insect swarm area, and a fusion-enhanced image is generated. S3: Extract the pigment deposition area of ​​the insect target based on the color contrast index, and divide the insect swarm core area and diffusion edge area into two parts based on the topological connectivity index to form a pest distribution vector map. S4: Match the pest distribution vector map with the pre-built hydropower station pest and disease feature database to determine the target pest and disease species and the corresponding biological natural enemy types; S5: Determine the priority level of natural enemy deployment based on the spatial weight relationship between the core area of ​​insect swarm aggregation and the diffusion edge area. Calculate the deployment coordinates and quantity based on the pest density distribution to form a targeted control plan. S6: Implement the prevention plan through the directional delivery device carried by the drone.

[0023] S1 specifically includes: S11: Fixed or mobile acquisition platforms equipped with multispectral imaging modules are deployed inside and outside the hydropower station. The acquisition range is set to cover the main engine surface, transmission facilities, turbine blade casing, and the surrounding environment within 10 meters, ensuring continuous imaging capabilities under unobstructed field of view conditions. S12: Start the multispectral camera device and sequentially collect image data including the visible light band (center wavelength 550 nm ± 10 nm) and the near-infrared band (center wavelength 850 nm ± 20 nm) according to the preset band parameters. The images of different bands are collected in time sequence, and the single exposure time is controlled within 30 ms and the frame rate is maintained at no less than 25 fps. S13: The multispectral image acquisition results are marked according to the shooting timestamp and equipment identification number to form an initial image set with spatial location information, and image integrity verification and abnormal frame removal are performed through the edge computing node to ensure data availability. The above steps ensure the comprehensiveness, stability and traceability of pest monitoring image data in key areas of the hydropower station through the multispectral image acquisition method of band division, high frame rate and spatial annotation, providing a high-quality data foundation for subsequent image enhancement and pest extraction.

[0024] S2 specifically includes: S21: Perform pixel-level brightness normalization on the visible light band images in the initial image set, adjust the image brightness distribution using an enhancement algorithm based on histogram equalization, and enhance the color gradient between the inner and edge areas of the region using a partitioned contrast enhancement algorithm; S22: Calculate the color contrast index for the normalized visible light image. The formula is: ,in, represents the color contrast index, represents the total number of pixels in the image, Respectively The grayscale value of each pixel in the red, green, and blue channels, all in 8-bit integer values ​​( ), which is used to measure the intensity of local color differences in an image; S23: Binarize the near-infrared band image and use the Otsu automatic threshold segmentation algorithm to distinguish the insect body area from the background area. Then, establish the insect body connectivity map and calculate the topological connectivity index. The formula is: ,in, represents the topological connectivity index, represents the pixel area of ​​the largest connected insect body region, Indicates the total area of ​​all insect bodies. This index is used to reflect the degree of insect aggregation; S24: Perform weighted fusion of the color contrast index and the topological connectivity index to generate a fused enhanced image. The fusion method uses the following pixel-level fusion model: ,in, Represents the fused image pixel value, is the pixel value corresponding to the visible light image, is the pixel value corresponding to the near-infrared image, 、 is the fusion weight parameter, satisfying , the default setting is By introducing two quantitative indicators, color contrast and topological connectivity, and using a weighted pixel fusion model to process multispectral images, the image expression ability of insect boundaries and aggregation areas can be effectively enhanced, providing a high-resolution, high-target contrast image basis for subsequent insect pest area extraction and analysis.

[0025] S3 specifically includes: S31: Based on the color contrast index obtained in S2, perform threshold segmentation on the color contrast values ​​of different pixels in the fused enhanced image, and determine the high-contrast areas exceeding the threshold as the pigment deposition areas of the insect target; specifically, set the color contrast segmentation threshold to 120. When the color contrast value of a pixel point is greater than the segmentation threshold, the pixel is marked as an insect target pixel; otherwise, it is marked as a background pixel. In this way, continuously distributed high-contrast areas in the image are extracted as the pigment deposition areas of the insect target; S32: Based on the topological connectivity index calculated in S2, the connected regions of the pigment deposition areas of the insect body targets are analyzed, and a comprehensive evaluation is performed based on the connectivity and area size within the region to distinguish the core area of ​​insect swarm aggregation from the edge area of ​​insect swarm diffusion. Specifically, the area with high connectivity and large area is defined as the core area of ​​insect swarm aggregation, and the remaining areas with low connectivity and small area are defined as the edge area of ​​insect swarm diffusion; S33: Extract the boundaries of the insect swarm core area and the diffusion edge area respectively, establish the geometric outline of the insect pest area based on the spatial coordinates, and convert it into a vector representation to obtain a pest distribution vector map; the above steps determine the target area of ​​the insect body through the color contrast index, and clearly divide the insect swarm core area and the diffusion edge area in combination with the topological connectivity index, which can accurately depict the spatial distribution characteristics of the pest and form a pest distribution vector map that can be used for subsequent precise prevention and control.

[0026] S32 specifically includes: S321: Perform 8-neighborhood connected region labeling processing on the pigment deposition area image extracted in S31, identify all connected insect body target areas, and record the area, perimeter, center of gravity coordinates and boundary contact length with adjacent areas of each connected area; S322: Calculate the connectivity index of each connected area. The formula is: ,in, Indicates the The connectivity of the target area of ​​the insect body, represents the pixel area of ​​the region, Indicates the boundary perimeter of the area; this indicator is used to measure whether the regional structure is compact. A larger value indicates a more concentrated regional shape and a higher degree of aggregation. S323: Setting the connectivity threshold , will satisfy The connected area is defined as the core area of ​​insect swarm gathering; the rest of the area that meets The area is delineated as the insect swarm diffusion edge area; the above steps construct connectivity indicators and combine regional geometric characteristics for graded judgment, which can scientifically distinguish the core aggregation area and edge diffusion area of ​​insect distribution, thereby improving the spatial resolution ability and regional discrimination accuracy of insect pest monitoring, and providing a reliable basis for subsequent release strategies.

[0027] S4 specifically includes: S41: Perform scale normalization on the pest distribution vector map generated in S3, unifying the vector contour data of the insect swarm core area and the diffusion edge area to the same spatial scale as the hydropower station pest and disease feature database to ensure scale consistency in subsequent matching analysis; S42: Extracting geometric characteristic parameters of pest vectors from the pest distribution vector map, including the average area, boundary perimeter, regional shape factor, and spatial layout pattern index of the target insect area, as input features for matching analysis; S43: calling a pre-stored standard vector feature template in a pre-built hydropower station pest and disease feature library, and performing feature matching between the geometric feature parameters of the pest vector to be identified and the template in the feature library using a cosine similarity algorithm to obtain a matching similarity score; S44: Select the standard feature template with the highest similarity score greater than 0.85, determine the corresponding target pest and disease species, and obtain the matching biological natural enemy species through the pest and disease record information in the hydropower station's pest and disease feature library, providing a basis for the subsequent formulation of biological control plans; through the feature matching method of scale normalization and cosine similarity, accurately match the pest distribution vector map to the pre-constructed pest and disease feature library, effectively improving the accuracy and efficiency of target pest and disease species identification, and providing a scientific basis for the subsequent precise deployment of corresponding biological natural enemies.

[0028] S43 specifically includes: S431: Combining the geometric feature parameters of the pest vector extracted in S42 into a feature vector , represent the average area, boundary perimeter, regional shape factor and spatial layout pattern index respectively; S432: Extract the corresponding feature vectors of the standard feature template from the hydropower station pest feature database, and record them as ,in Indicates the Standard feature templates; S433: For each standard feature template, use the cosine similarity algorithm to calculate the feature vector to be identified With the standard feature template vector The cosine similarity between them is: ,in, Represents the feature vector to be identified and the The similarity score between the standard feature template vectors ranges from 0 to 1. The closer the value is to 1, the higher the feature matching degree is. The first vector diagram of the distribution of pests to be identified is extracted. geometric characteristic parameter components; Indicates the first The first standard template Geometric characteristic parameters; is the serial index of the geometric feature parameters, with a value of 1 to 4. The above steps realize the rapid and accurate matching of the vector geometric feature parameters of the pests to be identified with the standard template through clear vectorized feature representation and cosine similarity algorithm, thereby improving the reliability and accuracy of pest and disease identification, and providing a quantitative basis for the accurate identification of pest and disease species and the determination of biological natural enemies.

[0029] The example table of the pest and disease feature database of the hydropower station mentioned above is as follows: In the above table, the average insect body area is the average pixel value of the insect body area obtained from the statistics of the training images, which measures the size characteristics of the insect body; the perimeter represents the pixel count value of the insect body edge outline, reflecting its complexity; the shape factor represents the shape compactness index; the layout index is used to characterize the spatial distribution density and regularity of the insect body, such as the center of gravity aggregation or Voronoi distribution entropy; the image template number represents the identifier corresponding to the standard image of the insect species, which is used for image comparison or model call; the name and number of natural enemies indicate the corresponding species of biological control for the pest and the recommended release rate per unit area; the active temperature and humidity range indicates the climatic conditions where the pest species is most likely to break out, which helps to assist in identification and judgment.

[0030] S5 specifically includes: S51: Based on the pest distribution vector map obtained in S3, calculate the area of ​​the insect swarm core area and the diffusion edge area, respectively, and record them as the area of ​​the core area. and the diffusion edge area , and use this to determine the spatial weight relationship and calculate the spatial weight value of the cluster core area and the spatial weight value of the diffuse edge area , the expressions are: and ; S52: Based on the spatial weight value obtained in S51, determine the priority level of natural enemy release; S53: Recommended number of natural enemies corresponding to target pests and diseases identified in S4 , combined with the spatial weight value and regional area calculated by S51, calculate the actual number of deployments in each region , the formula is: ,in, Indicates the The number of natural enemies released in each area, represents the spatial weight value of the area, Indicates the area of ​​the region; S54: Based on the geometric center coordinates of the insect swarm gathering core area and the diffusion edge area as the release coordinates, combined with the release quantity calculated in S53, a target control plan is formed including regional release priority, release coordinates and corresponding release quantities; the above steps effectively distinguish the control priorities of different insect pest areas through clear spatial weight calculation and release quantity matching methods, improve the spatial accuracy and control effect of biological natural enemy release, and are conducive to optimizing the allocation of control resources to the greatest extent.

[0031] The priorities for natural enemy deployment include: When the spatial weight of the core area is gathered When the concentration core area is set as the priority level 1 area, the diffusion edge area is set as the priority level 2 area; when When the concentration core area and the diffusion edge area are both set as priority level 2 areas; when When the diffusion edge area is set as the priority level 1 area, the gathering core area is set as the priority level 2 area.

[0032] like Figure 2 As shown, a biocontrol device for a hydropower station based on pest and disease identification is used to implement the above-mentioned biocontrol method for a hydropower station based on pest and disease identification, and includes the following modules: Image acquisition module: used to collect image data of the surface of hydropower station equipment and its surrounding environment through a multispectral camera device. The multispectral image includes visible light band and near-infrared band, and output the initial image set; Image processing module: connected to the image acquisition module, used to perform multi-band fusion enhancement processing on the initial image set, wherein the color contrast index is calculated based on the visible light band, the topological connectivity index is calculated based on the near-infrared band, and a fusion enhanced image is generated; Pest extraction module: This module is connected to the image processing module and is used to extract the pigment deposition area of ​​the insect target based on the color contrast index. It also divides the insect swarm into the core area and the diffusion edge area based on the topological connectivity index to form a vector map of the insect pest distribution. Pest and disease identification module: This module is connected to the pest extraction module and is used to match the geometric characteristic parameters of the pest distribution vector with the standard templates in the pre-built hydropower station pest and disease feature library. It uses cosine similarity calculation to output the matching results and determine the target pest and disease species and their corresponding biological natural enemy types. Control decision-making module: connected to the pest identification module, it is used to calculate the priority level, coordinates and quantity of natural enemy release based on the spatial weight relationship between the core area of ​​insect clusters and the diffusion edge area, the distribution of insect pests and natural enemy information, and generate a target biological control plan; Execution control module: connected to the prevention and control decision module, used to drive the UAV platform to load the directional delivery device and complete the natural enemy delivery operation in the target area according to the prevention and control plan.

[0033] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0034] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for biological control of hydropower stations based on pest identification, characterized in that: The following steps are involved: S1: An initial set of images of the surface of the hydropower station equipment and its surrounding environment is collected using a multispectral camera device, where the multispectral image includes visible light bands and near-infrared bands; S2: Perform multi-band fusion enhancement processing on the initial image set, calculate the color contrast index based on the visible light band; calculate the topological connectivity index based on the near-infrared band, and generate a fusion enhanced image; S3: Extract the pigment deposition area of ​​the insect target based on the color contrast index, and divide the insect swarm core area and diffusion edge area into two parts based on the topological connectivity index to form a pest distribution vector map. S4: Match the pest distribution vector map with the pre-built hydropower station pest and disease feature database to determine the target pest and disease species and the corresponding biological natural enemy types; S5: Determine the priority level of natural enemy deployment based on the spatial weight relationship between the core area of ​​insect swarm aggregation and the diffusion edge area. Calculate the deployment coordinates and quantity based on the pest density distribution to form a targeted control plan. S6: Implement the prevention plan through the directional delivery device carried by the drone.

2. A method for biological control of hydropower stations based on pest and disease identification according to claim 1, characterized in that: Said S1 specifically includes: S11: A fixed or mobile data acquisition platform equipped with a multispectral imaging module is deployed inside and outside the hydropower station. The acquisition range is set to cover the main engine surface, transmission facilities, turbine blade casing, and the surrounding environment within 10 meters. S12: Start the multispectral camera device, collect image data including visible light band and near-infrared band in sequence according to preset band parameters, and align the images of different bands in time sequence, control the single exposure time within 30 ms, and maintain the frame rate at no less than 25 fps; S13: Mark the multispectral image acquisition results according to the shooting timestamp and the device identification number to form an initial image set with spatial location information.

3. A method for biological control of hydropower stations based on pest identification according to claim 1, characterized in that: The S2 specifically includes: S21: performing pixel-level brightness normalization processing on the visible light band images in the initial image set; S22: Calculate the color contrast index for the normalized visible light image. The formula is: ,in, represents the color contrast index, represents the total number of pixels in the image, Respectively The grayscale value of each pixel in the red, green and blue channels; S23: Binarize the near-infrared band image and use the Otsu automatic threshold segmentation algorithm to distinguish the insect body area from the background area. Then, establish the insect body connectivity map and calculate the topological connectivity index. The formula is: ,in, represents the topological connectivity index, represents the pixel area of ​​the largest connected insect body region, It represents the total area of ​​all insect body regions; S24: Perform weighted fusion of the color contrast index and the topological connectivity index to generate a fused enhanced image.

4. A method for biological control of hydropower stations based on pest and disease identification according to claim 1, characterized in that: The S3 specifically includes: S31: Based on the color contrast index obtained in S2, the color contrast values ​​of different pixels in the fused enhanced image are subjected to threshold segmentation processing, and the high contrast area exceeding the threshold is determined as the pigment deposition area of ​​the insect target; S32: Based on the topological connectivity index calculated in S2, the connected region analysis is performed on the pigment deposition area of ​​the insect body target, and a comprehensive evaluation is performed based on the connectivity and area size within the area to distinguish the core area of ​​insect swarm aggregation from the edge area of ​​insect swarm diffusion; S33: Extract the boundaries of the insect swarm core area and the diffusion edge area respectively, establish the geometric outline of the insect pest area based on the spatial coordinates, and convert it into a vector representation to obtain a vector map of the insect pest distribution.

5. A method for biological control of hydropower stations based on pest and disease identification according to claim 4, characterized in that: The S32 specifically includes: S321: Perform 8-neighborhood connected region labeling processing on the pigment deposition area image extracted in S31, identify all connected insect body target areas, and record the area, perimeter, center of gravity coordinates and boundary contact length with adjacent areas of each connected area; S322: Calculate the connectivity index of each connected area. The formula is: ,in, Indicates the The connectivity of the target area of ​​the insect body, represents the pixel area of ​​the region, Indicates the boundary perimeter of the area; S323: Setting the connectivity threshold , will satisfy The connected area is defined as the core area of ​​insect swarm gathering; the rest of the area that meets The area is designated as the swarm spread edge zone.

6. A method for biological control of hydropower stations based on pest and disease identification according to claim 1, characterized in that: The S4 specifically includes: S41: Perform scale normalization on the pest distribution vector map generated in S3, unifying the vector contour data of the insect swarm core area and the diffusion edge area to the same spatial scale as the hydropower station pest and disease feature database; S42: Extracting geometric characteristic parameters of pest vectors from the pest distribution vector map, including the average area, boundary perimeter, regional shape factor, and spatial layout pattern index of the target insect area, as input features for matching analysis; S43: calling a pre-stored standard vector feature template in a pre-built hydropower station pest and disease feature library, and performing feature matching between the geometric feature parameters of the pest vector to be identified and the template in the feature library using a cosine similarity algorithm to obtain a matching similarity score; S44: Select the standard feature template with the highest similarity score and a score greater than 0.85, determine the corresponding target pest species, and obtain the biological natural enemy species that match the pest species through the pest record information in the hydropower station pest feature database.

7. A method for biological control of hydropower stations based on pest and disease identification according to claim 6, characterized in that: The S43 specifically includes: S431: Combining the geometric feature parameters of the pest vector extracted in S42 into a feature vector , represent the average area, boundary perimeter, regional shape factor and spatial layout pattern index respectively; S432: Extract the corresponding feature vectors of the standard feature template from the hydropower station pest feature database, and record them as ,in Indicates the Standard feature templates; S433: For each standard feature template, use the cosine similarity algorithm to calculate the feature vector to be identified With the standard feature template vector The cosine similarity between them is: ,in, Represents the feature vector to be identified and the The similarity score between the standard feature template vectors ranges from 0 to 1; The first vector diagram of the distribution of pests to be identified is extracted. geometric characteristic parameter components; Indicates the first The first standard template Geometric characteristic parameters; It is the serial index of the geometric feature parameter, with a value of 1 to 4.

8. The method for biological control of a hydropower station based on pest and disease identification according to claim 1, characterized in that: The S5 specifically includes: S51: Based on the pest distribution vector map obtained in S3, calculate the area of ​​the insect swarm core area and the diffusion edge area, respectively, and record them as the area of ​​the core area. and the diffusion edge area , and use this to determine the spatial weight relationship and calculate the spatial weight value of the cluster core area and the spatial weight value of the diffuse edge area , the expressions are: and ; S52: Based on the spatial weight value obtained in S51, determine the priority level of natural enemy release; S53: Recommended number of natural enemies corresponding to target pests and diseases identified in S4 , combined with the spatial weight value and regional area calculated by S51, calculate the actual number of deployments in each region , the formula is: ,in, Indicates the The number of natural enemies released in each area, represents the spatial weight value of the area, Indicates the area of ​​the region; S54: Based on the geometric center coordinates of the insect swarm gathering core area and the diffusion edge area as the release coordinates, combined with the release quantity calculated in S53, a target prevention and control plan including regional release priority, release coordinates and corresponding release quantity is formed.

9. A method for biological control of hydropower stations based on pest identification according to claim 8, characterized in that: The priority levels for natural enemy deployment include: When the spatial weight of the core area is gathered When the concentration core area is set as the priority level 1 area, the diffusion edge area is set as the priority level 2 area; when When the concentration core area and the diffusion edge area are both set as priority level 2 areas; when When the diffusion edge area is set as the priority level 1 area, the gathering core area is set as the priority level 2 area.

10. A biocontrol device for a hydropower station based on pest identification, used to implement a biocontrol method for a hydropower station based on pest identification as claimed in any one of claims 1 to 9, characterized in that: Includes the following modules: Image acquisition module: used to collect image data of the surface of hydropower station equipment and its surrounding environment through a multispectral camera device, the multispectral image includes visible light band and near infrared band, and output the initial image set; Image processing module: connected to the image acquisition module, used to perform multi-band fusion enhancement processing on the initial image set, wherein the color contrast index is calculated based on the visible light band, the topological connectivity index is calculated based on the near-infrared band, and a fusion enhanced image is generated; Pest extraction module: This module is connected to the image processing module and is used to extract the pigment deposition area of ​​the insect target based on the color contrast index. It also divides the insect swarm into the core area and the diffusion edge area based on the topological connectivity index to form a vector map of the insect pest distribution. Pest and disease identification module: This module is connected to the pest extraction module and is used to match the geometric characteristic parameters of the pest distribution vector with the standard templates in the pre-built hydropower station pest and disease feature library. It uses cosine similarity calculation to output the matching results and determine the target pest and disease species and their corresponding biological natural enemy types. Control decision-making module: connected to the pest identification module, it is used to calculate the priority level, coordinates and quantity of natural enemy release based on the spatial weight relationship between the core area of ​​insect clusters and the diffusion edge area, the distribution of insect pests and natural enemy information, and generate a target biological control plan; Execution control module: connected to the prevention and control decision module, used to drive the UAV platform to load the directional delivery device and complete the natural enemy delivery operation in the target area according to the prevention and control plan.

Images