Forestry disease and pest monitoring and early warning system

Abstract

The invention provides a forestry disease and pest monitoring and early warning system, which relates to the technical field of forestry informatization monitoring, and is characterized in that a damaged area and an abnormal temperature area are extracted by analyzing thermal imaging and bark falling conditions of the surface of a pine tree trunk, so that a termite damage dilatation index is calculated, a trunk multi-layer detection ring belt is further determined, and an intrusion radius is obtained; and termite danger grades are divided by combining the water content of the tree trunk, so that early warning is carried out. Therefore, the problem of serious forestry resource loss caused by inaccurate termite hazard monitoring, insufficient diffusion trend evaluation and lack of pertinence of prevention and control measures is solved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the field of forestry information monitoring technology, and more specifically, to a forestry pest and disease monitoring and early warning system. Background Art

[0002] Forestry pest and disease monitoring and early warning technologies are crucial components of forest resource protection. With the increasing emphasis on ecological and environmental protection in modern forestry management, precision pest and disease monitoring and early warning technologies have become a research hotspot. Traditional forestry pest and disease monitoring methods rely primarily on manual inspections, trapping devices, and ground surveys. While these methods played an important role in their early stages of development, they are limited by manpower, time, and environmental factors, resulting in low monitoring efficiency and difficulty in providing timely and comprehensive information on the occurrence and spread of pests and diseases. In recent years, advances in sensor technology, image processing, and big data analytics have enabled pest and disease monitoring to evolve towards automation and intelligence. For example, drones equipped with high-resolution cameras or infrared sensors are used to monitor large areas of forests, combined with image processing algorithms to identify pest and disease areas. Alternatively, spectral analysis can be used to detect abnormal reflectance characteristics on plant surfaces and infer pest distribution. However, these technologies are often limited to macroscopic pest and disease distribution monitoring, making it difficult to precisely monitor specific, high-incidence pests (such as termites) within specific areas and to quantitatively assess their spread.

[0003] Existing technologies for forestry pest and disease monitoring and early warning systems have numerous shortcomings. For one thing, early detection of pests and diseases has always been challenging, especially for elusive and rapidly spreading pests like termites. Traditional monitoring methods struggle to capture early signs of their activity, missing the optimal opportunity for control. Furthermore, existing monitoring systems often rely on single data sources, such as infrared thermal imaging or spectral data, which fail to fully reflect the complex characteristics of pests and diseases, leading to biased monitoring results. Furthermore, when analyzing and warning of pest and disease spread trends, existing research is largely limited to static analysis and lacks the ability to dynamically assess the speed and extent of spread. Specifically for termite infestation monitoring, existing technologies lack a quantitative calculation model for the diffusion index, making it impossible to intuitively characterize the extent of termite infestation. More importantly, existing technologies often adopt a one-size-fits-all approach to targeted control measures, making it difficult to integrate the specific distribution characteristics of the pest and disease to provide regionalized and precise application recommendations. This reduces control efficiency and increases the loss of forestry resources. Summary of the Invention

[0004] To address the above technical problems, the present invention provides a forest pest monitoring and early warning system that can, to a certain extent, address the serious loss of forestry resources caused by inaccurate termite damage monitoring, insufficient spread trend assessment, and lack of targeted prevention and control measures.

[0005] According to one aspect of the present invention, a forestry pest and disease monitoring and early warning system is provided, comprising: A data acquisition unit, used to collect infrared thermal imaging data and bark shedding image data of the pine tree trunk surface; a data processing unit, configured to divide the bark shedding image data into damaged areas and healthy areas, and extract abnormal temperature distribution points in the infrared thermal imaging data; a damage degree determination unit, configured to calculate a termite damage expansion index based on the area of ​​the damaged area and the positional relationship of the abnormal temperature distribution point; the termite damage expansion index represents the expansion degree of the termite damage area per unit time; A multi-layer detection ring unit is used to determine a multi-layer detection ring of the tree trunk according to the termite damage expansion index, mark high-incidence spread nodes within the multi-layer detection ring, and calculate the invasion radius of each high-incidence spread node; The hazard warning unit is used to classify the termite hazard warning level when a new damaged area is detected, based on the coverage of the multi-layer detection ring and the invasion radius, combined with the tree trunk moisture content data, and to provide regional pesticide control suggestions.

[0006] Furthermore, the data processing unit extracts the color and texture feature vectors of the superpixel blocks, uses the density peak clustering algorithm to automatically identify the cluster center, divides the bark peeling area and the intact bark area into damaged and healthy categories, and generates regional division results reflecting termite damage through regional merging and morphological optimization.

[0007] Furthermore, the cluster center is identified by calculating the local density value of the superpixel block and the minimum distance value to the point with higher density, identifying the cluster center with both high density and large distance in two-dimensional space, and iteratively assigning cluster labels based on the density reachability principle to finally obtain a complete clustering result.

[0008] Furthermore, the abnormal temperature distribution points are determined by constructing a temperature distribution probability density function through spatial block division, and the temperature distribution difference between the central area and the adjacent areas is detected using Kullback-Leibler divergence. Combined with time series analysis, continuous abnormal points are identified, and finally the abnormal temperature distribution points are determined.

[0009] Furthermore, the operation logic of the temperature distribution probability density function is as follows: For each temperature value to be evaluated, calculate the Gaussian kernel function value between it and all temperature sampling points in the area; The distance weight function is introduced to reflect the influence of spatial position on temperature distribution; The weighted Gaussian kernel function values ​​of all sampling points are accumulated and summed, and then normalized by dividing by the product of the number of sampling points and the bandwidth parameter to obtain the final probability density function value.

[0010] Furthermore, the calculation of the termite damage expansion index includes: Based on the data collected during the observation period, the area growth rate of a single damaged area is calculated. The area growth rate is the difference between the current damaged area and the initial area divided by the number of observation days to obtain the daily area growth rate. Sampling the boundary of the damaged area at equal intervals, dividing the number of abnormal temperature distribution points at all sampling points by the total length of the boundary to obtain the density of the abnormal temperature distribution points; Record the total number of new abnormal temperature distribution points in the damaged area during the observation period, and divide this number by the number of observation days to obtain the average daily number of new abnormal temperature distribution points; Normalizing the daily area growth, the density of the abnormal temperature distribution points, and the average daily number of newly added abnormal temperature distribution points, and performing weighted calculation to obtain the termite damage expansion index; When the damaged area bifurcates or merges, the termite damage expansion index of each sub-area is calculated separately, and the maximum value in each sub-area is selected as the expansion index of the damaged area.

[0011] Furthermore, within the multi-layer detection ring zone, the ring zone area is divided into a plurality of sector-shaped detection units. When the expansion index of a sector-shaped detection unit is greater than the average value of adjacent units, the center point of the unit is marked as a high-incidence spread node.

[0012] Furthermore, the calculation of the invasion radius includes: Taking the marked high-incidence spread node as the center, the area around the node is divided into multiple directions, and sampling is carried out along the multiple directions, and the termite damage expansion index at the sampling point is recorded; When the expansion index of a sampling point in a certain sampling direction is lower than the preset multiple of the expansion index of the central node, the distance from the point to the central node is recorded as the temporary invasion radius in that direction; Counting the temporary intrusion radius in the multiple directions, and calculating the average thereof as the benchmark intrusion radius of the high-incidence intrusion node; Correcting the reference invasion radius to obtain the invasion radius; For adjacent high-incidence spreading nodes, when their invasion radiuses overlap, the invasion radius corresponding to the maximum expansion index in the overlapping area is taken as the effective invasion radius of the area.

[0013] Furthermore, when the invasion radius of multiple adjacent high-incidence spreading nodes overlaps, the spatial distribution gradient and distance attenuation coefficient in the overlapping area are calculated, and the influence of each node is superimposed to obtain the comprehensive expansion index. The area where the comprehensive expansion index exceeds the influence of a single node is marked as a coordinated expansion area. If the area of ​​the overlapping region exceeds half of the invasion radius of a single node, they are merged into one expansion center, and the invasion radius is re-determined based on the comprehensive expansion index.

[0014] Furthermore, the termite damage warning level is divided based on the termite damage expansion index, the invasion radius and the moisture content of the tree trunk.

[0015] Compared to existing technologies, the forestry pest monitoring and early warning system provided by this invention uses thermal imaging and bark loss analysis of pine tree trunk surfaces to identify damaged areas and abnormal temperature zones. This system then calculates a termite damage expansion index, identifies multiple detection zones around the trunk, and determines the invasion radius. This, combined with the trunk moisture content, determines the termite risk level and provides early warnings. This, to a certain extent, addresses the serious loss of forestry resources caused by inaccurate termite damage monitoring, insufficient assessment of spread trends, and lack of targeted prevention and control measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings: Figure 1 4 is a system block diagram of a forestry pest and disease monitoring and early warning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] Below, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0018] Figure 1 FIG. 1 is a system block diagram of a forestry pest and disease monitoring and early warning system according to an embodiment of the present invention. Figure 1 As shown in the figure, the forestry pest and disease monitoring and early warning system includes: A data acquisition unit, used to collect infrared thermal imaging data and bark shedding image data of the pine tree trunk surface; a data processing unit, configured to divide the bark shedding image data into damaged areas and healthy areas, and extract abnormal temperature distribution points in the infrared thermal imaging data; a damage degree determination unit, configured to calculate a termite damage expansion index based on the area of ​​the damaged area and the positional relationship of the abnormal temperature distribution point; the termite damage expansion index represents the expansion degree of the termite damage area per unit time; A multi-layer detection ring unit is used to determine a multi-layer detection ring of the tree trunk according to the termite damage expansion index, mark high-incidence spread nodes within the multi-layer detection ring, and calculate the invasion radius of each high-incidence spread node; The hazard warning unit is used to classify the termite hazard warning level when a new damaged area is detected, based on the coverage of the multi-layer detection ring and the invasion radius, combined with the tree trunk moisture content data, and to provide regional pesticide control suggestions.

[0019] More specifically, the data acquisition unit is implemented as follows: In the woods, infrared thermal imaging data of the surface of pine tree trunks were first collected by infrared thermal imaging sensors installed at different heights on the trunks, with a collection interval of 30 cm. At the same time, bark shedding image data of the trunk surface was collected by high-definition camera equipment, with a field of view of 120 degrees and a collection distance of 1.5 meters. The bark shedding image data was input into a deep segmentation algorithm, and the image was divided into damaged areas and healthy areas using a superpixel-based segmentation method. The damaged areas included features such as bark peeling, depressions, and cracks. Temperature threshold analysis was performed on the infrared thermal imaging data to extract abnormal temperature distribution points with a temperature difference greater than 2°C from the surrounding areas. The abnormal temperature distribution points represented local temperature anomalies caused by termite activity. The spatial distribution information of the damaged areas was spatially aligned with the position information of the abnormal temperature distribution points to establish a trunk surface disease feature map, which was used for the subsequent calculation of the termite damage expansion index.

[0020] Furthermore, based on the data obtained by the data acquisition unit, the data processing unit divides the image into damaged areas and healthy areas using a superpixel-based segmentation method as follows: The image is divided into superpixel blocks of uniform size. The color feature vector in the HSV color space and the texture feature vector based on the gray-level co-occurrence matrix are extracted for each superpixel block. The feature vectors of all superpixel blocks are input into the density peak clustering algorithm, which identifies the cluster centers by calculating the local density and distance in the feature space and automatically obtains the optimal number of clusters. Based on the clustering results, the superpixel blocks are divided into multiple categories, among which the bark peeling areas are naturally clustered into the damaged category due to the discontinuity of texture features and the significant difference in color features, while the intact bark areas are clustered into the healthy category due to the continuity of texture features and the similarity of color features. Adjacent superpixel blocks with similar clustering labels are merged to form continuous damaged areas and healthy areas. Finally, the region boundaries are morphologically optimized to obtain the regional division results reflecting the characteristics of termite damage.

[0021] To identify cluster centers, the Euclidean distance between each superpixel and all other superpixels in feature space is calculated to construct a distance matrix. A cutoff distance is calculated based on the distance matrix, taking the median of all distances. For each superpixel, the number of neighboring points whose distance is less than the cutoff distance is counted to obtain the local density value of that point. The distance from each superpixel to the nearest point with a higher density than itself is also calculated to obtain the minimum distance value for that point. All superpixels are projected into a two-dimensional space composed of local density and minimum distance. Points with both a high local density and a large minimum distance are identified as cluster centers. Based on the identified cluster centers, a cluster label is assigned to each non-central point using the density reachability principle, assigning each point to the category of its nearest neighbor with a higher density than itself. Through iterative propagation, all points are assigned to the corresponding cluster category, resulting in the final clustering result. Because damaged and healthy areas of bark exhibit different density distribution characteristics in feature space, this method can adaptively discover the inherent clustering structure of the data, avoiding the limitation of pre-specified number of clusters.

[0022] The distance between each superpixel block and the nearest point with a higher density than itself is calculated, and the minimum distance value of the point is obtained, which can be expressed by the following formula: ; in, ; in, is the Euclidean distance between the i-th superpixel block and the j-th superpixel block; is the coordinate value of the i-th superpixel block in the k-th dimension feature space; is the feature space dimension; is the minimum distance value of the i-th superpixel block; is the local density value of the i-th superpixel block; The indices of other superpixel blocks that satisfy the local density greater than the i-th superpixel block.

[0023] It should be noted that the criteria for determining the local density value and the minimum distance value are determined in a data-driven manner. Specifically, the local density values ​​of all superpixel blocks are sorted, and their 75% quantile is calculated as the density reference value; the minimum distance values ​​of all superpixel blocks are sorted, and their 75% quantile is calculated as the distance reference value; in the two-dimensional space composed of local density and minimum distance, the two-dimensional space is divided into four quadrants with the density reference value and the distance reference value as the dividing line; points located in the first quadrant (i.e., the local density value is greater than the density reference value and the minimum distance value is greater than the distance reference value) are identified as cluster centers; to avoid cluster centers being too dense, the Mahalanobis distance between each point in the first quadrant is calculated. When the Mahalanobis distance between two points is less than their average value, the point with the larger local density value is retained as the cluster center.

[0024] On the other hand, the damage degree determination unit performs contour detection on the damaged area to obtain the area size and boundary coordinates of each damaged area; performs spatial position mapping on the abnormal temperature distribution points extracted from the infrared thermal imaging data and the damaged area, and establishes a correlation between the temperature abnormality points and the damaged area; selects data for seven consecutive days as an observation period, and calculates the area growth of a single damaged area and the number of newly added abnormal temperature distribution points in each observation period; divides each damaged area into multiple levels according to the area size, and for each level of damaged area, counts the density of abnormal temperature distribution points on its boundary, where the density represents the number of damaged areas. The number of abnormal temperature distribution points along the boundary length; the termite damage expansion index is calculated by weighted summation based on the area growth, the density of abnormal temperature distribution points and the number of newly added abnormal temperature distribution points, where the weight of area growth is 0.4, the weight of the density of abnormal temperature distribution points is 0.35, and the weight of the number of newly added abnormal temperature distribution points is 0.25; when the termite damage expansion index of a damaged area shows an upward trend for three consecutive days, the area is marked as an active expansion area; the termite damage expansion index is used to characterize the activity level and damage expansion trend of termites at different positions on the tree trunk, providing a basis for subsequent early warning classification.

[0025] The abnormal temperature distribution points are determined as follows: The infrared thermal imaging data is spatially partitioned, with each partition being 10×10 cm in size. Within each partition, a kernel density estimation method is used to construct a temperature distribution probability density function, which reflects the distribution characteristics of the temperature values ​​within the partition. Eight adjacent partitions are selected around each partition to form a temperature analysis window. Within the temperature analysis window, the temperature distribution probability density function of the central partition is compared with the probability density functions of the adjacent partitions. When the temperature distribution of the central partition deviates significantly, the Kullback-Leibler divergence is used to calculate the degree of difference in temperature distribution between the central and adjacent partitions. For each temperature point, its contribution to the Kullback-Leibler divergence is calculated. When the contribution value is greater than twice the average contribution value within the temperature analysis window, the temperature point is marked as a candidate anomaly point. A time series analysis is performed on the candidate anomaly points. When a candidate anomaly point at a certain location appears continuously within three consecutive acquisition cycles, it is determined to be an abnormal temperature distribution point. The method for determining abnormal temperature distribution points avoids the limitations of a fixed threshold and can adaptively identify temperature anomalies under different environmental conditions.

[0026] Among them, the probability density function of the temperature distribution in the block area is It can be expressed as follows: ; More specifically, the temperature anomaly measurement value is as follows: ; in, is the temperature value to be evaluated; is the number of temperature sampling points within the block area; is the bandwidth parameter; is the temperature value of the i-th sampling point; is the distance from the i-th sampling point to the center of the block area; is the distance weight function, expressed as . is the temperature anomaly measurement value of the central block area; is the temperature distribution probability density function of the central block area; is the temperature distribution probability density function of the kth adjacent block area, when When it is greater than twice the average contribution value in the temperature analysis window, the corresponding temperature point is marked as a candidate anomaly point.

[0027] The steps to calculate the termite damage expansion index are as follows: Based on the data collected during the observation period, the area growth rate of a single damaged area was first calculated, that is, the difference between the current damaged area and the initial area was divided by the number of observation days to obtain the daily area growth; the damaged area boundary was sampled at equal distances, and the sampling point spacing was set to 5 cm. The number of abnormal temperature distribution points within a radius of 3 cm for each sampling point was counted, and the number of abnormal temperature distribution points of all sampling points was divided by the total length of the boundary to obtain the density of abnormal temperature distribution points; the total number of new abnormal temperature distribution points in the damaged area during the observation period was recorded, and this value was divided by the number of observation days to obtain the average daily number of new abnormal temperature distribution points ; The above three parameters are normalized so that their numerical ranges are all between 0 and 1; the normalized area growth is multiplied by a weight of 0.4, the density of abnormal temperature distribution points is multiplied by a weight of 0.35, and the number of newly added abnormal temperature distribution points is multiplied by a weight of 0.25, and the three are added together to obtain the termite damage expansion index of the damaged area; when the damaged area is bifurcated or merged, the termite damage expansion index of each sub-area is calculated separately, and the maximum value is selected as the expansion index of the area; the calculation of the termite damage expansion index comprehensively considers multiple characteristic parameters of termite damage and can accurately reflect the expansion trend of termite activities.

[0028] For example, in each observation period, the area expansion analysis of the damaged area is performed, the area of ​​the damaged area on the initial sampling day is recorded as the initial area value, the area of ​​the damaged area on the seventh day is recorded as the final area value, the final area value is subtracted from the initial area value and divided by the number of observation days to obtain the daily area growth; the boundary of the damaged area is sampled at equal distances to obtain multiple sampling points, the number of abnormal temperature distribution points within a specific radius of each sampling point is counted, the total number of abnormal temperature distribution points is divided by the total length of the boundary of the damaged area to obtain the density of the abnormal temperature distribution points; the abnormal temperature distribution points that appear in the observation period are cumulatively counted, and the number of newly added abnormal temperature distribution points is recorded. The total number is divided by the number of observation days to obtain the average daily number of new abnormal temperature distribution points; the maximum reference value of each parameter is determined based on historical monitoring data, and the daily area growth, the density of abnormal temperature distribution points and the average daily number of new abnormal temperature distribution points are divided by their corresponding maximum reference values ​​for normalization; the normalized area growth is multiplied by the first weight, the density of abnormal temperature distribution points is multiplied by the second weight, and the number of new abnormal temperature distribution points is multiplied by the third weight, and the three are added together to obtain the termite damage expansion index of the damaged area; when there are multiple branch damaged areas, the termite damage expansion index of each branch area is calculated separately, and the maximum value is selected as the expansion index of the area.

[0029] The multi-layer detection ring unit is based on the coordinate grid system of the trunk surface. Starting from the bottom of the trunk, a detection ring is established every 50 cm, and the width of the detection ring is 20 cm. Within each detection ring, the ring area is divided into 12 sector detection units, and the central angle of each sector detection unit is 30 degrees. The termite damage expansion index in each sector detection unit is calculated. When the expansion index of a sector detection unit is greater than the average value of the adjacent units, the center point of the unit is marked as a high-incidence spread node. For each high-incidence spread node, the termite damage expansion index is calculated according to the size of its termite damage expansion index. Determine the scope of influence, specifically: radiate from the node as the center to the surrounding areas. When the expansion index of the surrounding area shows a decreasing trend and drops to 30% of the node expansion index, this range is the invasion radius; count the spatial distribution patterns of high-incidence spread nodes in each detection ring zone. When the vertical distance between high-incidence spread nodes in adjacent detection ring zones is less than 30 cm, connect these nodes to form a spread channel; the invasion radius of the high-incidence spread node expands with the increase of the termite damage expansion index, which is used to predict the spread range of termite damage and provide a spatial reference for subsequent regional pesticide control.

[0030] When the invasion radius of multiple high-incidence spread nodes overlap, the spatial distribution gradient of the termite damage expansion index of each node in the overlapping area is first calculated, where the gradient value represents the rate of change of the expansion index per unit distance; grid sampling points are established in the overlapping area, and the grid spacing is set to 5 cm; for each sampling point, the distance attenuation coefficient from its to each high-incidence spread node is calculated, which is inversely proportional to the distance and directly proportional to the node's termite damage expansion index; based on the distance attenuation coefficient, the influence of each node is weighted and superimposed to obtain the comprehensive expansion index at the sampling point; when the comprehensive expansion index of the overlapping area is greater than the expansion index generated by a single node, the overlapping area is marked as a coordinated expansion zone, and its damage level is higher than the influence of a single node; if the area of ​​the overlapping area exceeds 50% of the coverage area of ​​the invasion radius of any node, these high-incidence spread nodes are merged into an expansion center, and the invasion radius of the expansion center is recalculated. The size of the invasion radius is determined by the merged comprehensive expansion index; the processing method takes into account the interaction of multiple high-incidence spread nodes and more accurately reflects the actual spread of termite damage.

[0031] For the expansion center formed by the merger, its spatial position is first determined. This position is the weighted centroid of each high-incidence spread node participating in the merger, and the weight is the termite damage expansion index of each node; the comprehensive expansion index after the merger is calculated, specifically: a concentric circle coordinate system is established based on the centroid position, radial sampling is performed every 10 degrees in the range of 0 to 360 degrees, and the expansion index attenuation curve of the original node is statistically calculated on each radial sampling line; the attenuation curves on each sampling line are superimposed and averaged to obtain the radial expansion characteristic curve of the expansion center after the merger; the new invasion radius is determined based on the radial expansion characteristic curve, and when the curve value drops to 25% of the maximum value, the distance from this point to the centroid is determined as the invasion radius; if the radial expansion characteristic curve shows obvious differences in different directions, a variable radius is used to describe the invasion range, and the invasion radius in the main expansion direction is relatively large; when the invasion radius after the merger is less than the maximum invasion radius of the original node, the maximum invasion radius is kept unchanged to ensure effective coverage of the warning range.

[0032] Furthermore, with the marked high-incidence spread node as the center, a polar coordinate system is established around it, and the area around the node is divided into 8 directions, with an angle of 45 degrees in each direction; sampling is carried out in steps of 5 cm along each direction, and the termite damage expansion index at the sampling point is recorded; when the expansion index of a point in the sampling direction is lower than 30% of the expansion index of the central node, the distance from the point to the central node is recorded as the temporary invasion radius in that direction; the temporary invasion radius in the 8 directions is counted, and its mean is calculated as the baseline invasion radius of the high-incidence spread node; the baseline invasion radius is corrected, and the correction factors include: the higher the moisture content of the tree trunk, the larger the correction coefficient; the closer the detection ring is to the ground, the larger the correction coefficient; the greater the density of abnormal temperature distribution points in the surrounding area, the larger the correction coefficient; the baseline invasion radius is multiplied by the comprehensive correction coefficient to obtain the final invasion radius; for adjacent high-incidence spread nodes, when their invasion radii overlap, the invasion radius corresponding to the largest expansion index in the overlapping area is taken as the effective invasion radius of the area. In more detail, the calculation of the invasion radius can be expressed by the following formula: ; in, is the final invasion radius, is the baseline invasion radius, is the moisture content of the tree trunk, To detect the height of the ring from the ground, The reference height is 100 cm. is the density of abnormal temperature distribution points, is the distance from the i-th adjacent high-incidence spreading node to the current node, The reference distance is 50 cm. The influence weight of moisture content is 0.3. The height influence weight is 0.2. The temperature density influence weight is set to 0.25. is the number of adjacent high-incidence spreading nodes.

[0033] When the detection system finds a new damaged area, the hazard warning unit first determines the location of the detection ring where the area is located; analyzes whether the damaged area is within the invasion radius coverage of an existing high-incidence spread node. If it is within the coverage range, the area is marked as an expansion-type damaged area; if it is outside the coverage range, the area is marked as a newly damaged area; collects moisture content data of tree trunks around the damaged area, and performs a layered analysis on the moisture content data, in which the moisture content of the outer bark, xylem and heartwood layers are statistically analyzed separately; based on the termite damage expansion index, invasion radius coverage and trunk moisture content data, the termite hazard warning level is divided into four levels: when a newly damaged area appears and the moisture content of the surrounding tree trunks is 30% higher than the normal value, the area is divided into a first-level warning area; when the expansion index of the expansion-type damaged area continues to grow and is between two levels, the termite hazard warning level is divided into four levels: when a newly damaged area appears and the moisture content of the surrounding tree trunks is 30% higher than the normal value, the area is divided into a first-level warning area; when the expansion index of the expansion-type damaged area continues to grow and is between two levels, the termite hazard warning level is divided into four levels: when a newly damaged area appears and the moisture content of the surrounding tree trunks is 30% higher than the normal value, the area is divided into a first-level warning area; when the expansion index of the expansion-type damaged area continues to grow and is between two levels, the termite hazard warning level is divided into four levels: when the ... When the invasion radius of the above high-incidence spread nodes overlaps, the area is divided into a second-level warning area; when the expansion index of the expansion-type damaged area grows steadily and is only within the invasion radius of a single high-incidence spread node, the area is divided into a third-level warning area; when the expansion index of the damaged area shows a downward trend, the area is divided into a fourth-level warning area; formulate prevention and control recommendations based on the warning level: for the first-level warning area, it is recommended to inject high-concentration pesticides within 2 times the invasion radius around the damaged area and set up a protective isolation belt; for the second-level warning area, it is recommended to spray pesticides in the connecting area between high-incidence spread nodes and increase the monitoring frequency; for the third-level warning area, it is recommended to spray pesticides regularly within the invasion radius; for the fourth-level warning area, it is recommended to regularly observe and record the changing trend of the termite damage expansion index.

[0034] In summary, the forestry pest and disease monitoring and early warning system based on the embodiments of the present invention has been demonstrated. By analyzing thermal imaging and bark loss on the surface of pine tree trunks, it extracts damaged areas and abnormal temperature zones, thereby calculating the termite damage expansion index. Furthermore, it determines multiple detection zones around the trunk to obtain an invasion radius. Combined with the trunk moisture content, this system then classifies termite risk levels and issues early warnings. This, to a certain extent, addresses the serious loss of forestry resources caused by inaccurate termite damage monitoring, insufficient assessment of spread trends, and lack of targeted prevention and control measures.

[0035] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned forestry pest monitoring and early warning method have been described in detail above. Figure 1 The forestry pest and disease monitoring and early warning system has been introduced in detail, and therefore, its repeated description will be omitted.

Claims

1. A forestry pest monitoring and early warning system, characterized in that: include: A data acquisition unit, used to collect infrared thermal imaging data and bark shedding image data of the pine tree trunk surface; a data processing unit, configured to divide the bark shedding image data into damaged areas and healthy areas, and extract abnormal temperature distribution points in the infrared thermal imaging data; a damage degree determination unit, configured to calculate a termite damage expansion index based on the area of ​​the damaged area and the positional relationship of the abnormal temperature distribution point; the termite damage expansion index represents the expansion degree of the termite damage area per unit time; A multi-layer detection ring unit is used to determine a multi-layer detection ring of the tree trunk according to the termite damage expansion index, mark high-incidence spread nodes within the multi-layer detection ring, and calculate the invasion radius of each high-incidence spread node; The hazard warning unit is used to classify the termite hazard warning level when a new damaged area is detected, based on the coverage of the multi-layer detection ring and the invasion radius, combined with the tree trunk moisture content data, and to provide regional pesticide control suggestions.

2. The forestry pest monitoring and early warning system according to claim 1 is characterized in that: The data processing unit extracts the color and texture feature vectors of superpixel blocks, uses the density peak clustering algorithm to automatically identify the cluster center, divides the bark peeling area and the intact bark area into damaged and healthy categories, and generates regional division results reflecting termite damage through regional merging and morphological optimization.

3. The forestry pest monitoring and early warning system according to claim 2 is characterized in that: The cluster centers are identified by calculating the local density value of the superpixel block and the minimum distance value to the point with higher density, identifying the cluster centers with both high density and large distance in two-dimensional space, and iteratively assigning cluster labels based on the density reachability principle to finally obtain a complete clustering result.

4. The forestry pest monitoring and early warning system according to claim 1 is characterized in that: The abnormal temperature distribution points are determined by constructing a temperature distribution probability density function through spatial block division, and using the Kullback-Leibler divergence to detect the temperature distribution difference between the central area and the adjacent areas. Combined with time series analysis, continuous abnormal points are identified, and finally the abnormal temperature distribution points are determined.

5. The forestry pest monitoring and early warning system according to claim 4 is characterized in that: The operation logic of the temperature distribution probability density function is as follows: For each temperature value to be evaluated, calculate the Gaussian kernel function value between it and all temperature sampling points in the area; The distance weight function is introduced to reflect the influence of spatial position on temperature distribution; The weighted Gaussian kernel function values ​​of all sampling points are accumulated and summed, and then normalized by dividing by the product of the number of sampling points and the bandwidth parameter to obtain the final probability density function value.

6. The forestry pest monitoring and early warning system according to claim 1 is characterized in that: The calculation of the termite damage expansion index includes: Based on the data collected during the observation period, the area growth rate of a single damaged area is calculated. The area growth rate is the difference between the current damaged area and the initial area divided by the number of observation days to obtain the daily area growth rate. Sampling the boundary of the damaged area at equal intervals, dividing the number of abnormal temperature distribution points at all sampling points by the total length of the boundary to obtain the density of the abnormal temperature distribution points; Record the total number of new abnormal temperature distribution points in the damaged area during the observation period, and divide this number by the number of observation days to obtain the average daily number of new abnormal temperature distribution points; Normalizing the daily area growth, the density of the abnormal temperature distribution points, and the average daily number of newly added abnormal temperature distribution points, and performing weighted calculation to obtain the termite damage expansion index; When the damaged area bifurcates or merges, the termite damage expansion index of each sub-area is calculated separately, and the maximum value in each sub-area is selected as the expansion index of the damaged area.

7. The forestry pest monitoring and early warning system according to claim 1 is characterized in that: In the multi-layer detection ring zone, the ring zone area is divided into a plurality of sector-shaped detection units. When the expansion index of a sector-shaped detection unit is greater than the average value of adjacent units, the center point of the unit is marked as a high-incidence spread node.

8. The forestry pest monitoring and early warning system according to claim 7 is characterized in that: The calculation of the invasion radius includes: Taking the marked high-incidence spread node as the center, the area around the node is divided into multiple directions, and sampling is carried out along the multiple directions, and the termite damage expansion index at the sampling point is recorded; When the expansion index of a sampling point in a certain sampling direction is lower than the preset multiple of the expansion index of the central node, the distance from the point to the central node is recorded as the temporary invasion radius in that direction; Counting the temporary intrusion radius in the multiple directions, and calculating the average thereof as the benchmark intrusion radius of the high-incidence intrusion node; Correcting the reference invasion radius to obtain the invasion radius; For adjacent high-incidence spreading nodes, when their invasion radiuses overlap, the invasion radius corresponding to the maximum expansion index in the overlapping area is taken as the effective invasion radius of the area.

9. The forestry pest monitoring and early warning system according to claim 8, characterized in that: When the invasion radius of multiple adjacent high-incidence spreading nodes overlaps, the spatial distribution gradient and distance attenuation coefficient in the overlapping area are calculated, and the influence of each node is superimposed to obtain the comprehensive expansion index. The area where the comprehensive expansion index exceeds the influence of a single node is marked as a coordinated expansion area. If the area of ​​the overlapping region exceeds half of the invasion radius of a single node, they are merged into one expansion center, and the invasion radius is re-determined based on the comprehensive expansion index.

10. The forestry pest monitoring and early warning system according to claim 1, characterized in that: The termite damage warning level is divided based on the termite damage expansion index, the invasion radius and the tree trunk moisture content.

Images