Forest resource investigation sampling device and method

Abstract

The application provides a forest resource investigation sampling device and method, determines the space-time consistency deviation of the insect sound spectrum at the sampling node in the current period based on the frequency spectrum change difference value of the insect sound spectrum signal at the sampling node in the time dimension and the distribution change characteristics of the insects at the sampling node in the space dimension, performs deviation anomaly analysis on the space-time consistency deviation of the insect sound spectrum at the sampling node in the current period in combination with the influence correlation between different environmental parameters and the outbreak of pests and diseases and the environmental parameter data of the current period, obtains the abnormal confidence value of the insects at the sampling node, clusters the sampling nodes with abnormal confidence values exceeding the abnormal threshold value according to all the abnormal confidence values in combination with a spatial clustering algorithm, and further divides the sampling priority areas of multiple pests and diseases in the target forest area. The scheme of the application can optimize the sampling area of the target forest in combination with multi-dimensional anomaly analysis of monitoring data.

Description

Technical Field

[0001] This application relates to the field of forest resource survey and sampling technology, and in particular to a forest resource survey and sampling device and method. Background Technology

[0002] Forest resource surveys are divided into basic surveys and specialized surveys. Basic surveys target the core elements of forest resources, such as tree species, stock volume, diameter at breast height (DBH), tree height, forest area, and soil physicochemical properties. The purpose is to understand the basic status of forest resources. Specialized surveys, on the other hand, mainly target specific issues, such as forest pests and diseases, forest fire sites, distribution of rare and endangered plants and animals, and carbon storage. The purpose is to solve specific resource management or ecological problems.

[0003] Traditional methods for forest resource surveys and sampling for forest pests and diseases often require a large workforce to penetrate deep into forest areas, working in harsh conditions. Furthermore, large-scale forest surveys are time-consuming and fail to meet the need for rapid information acquisition. The labor and logistical costs are prohibitively high. Current technologies combine IoT sensors and drones to monitor target forests, with data processing centers then processing and analyzing the data to identify the sampling areas. However, environmental factors and sensor limitations can lead to inaccurate data. For example, normally monitored data may become abnormal due to environmental factors, or vice versa, resulting in unrepresentative sampling areas. Therefore, optimizing the sampling area of ​​the target forest by incorporating multi-dimensional anomaly analysis of monitoring data has become a significant challenge for the industry. Summary of the Invention

[0004] Based on this, this application provides a forest resource survey sampling device and method for optimizing the sampling area of ​​a target forest by combining multi-dimensional anomaly analysis of monitoring data.

[0005] Firstly, this application provides a sampling area optimization method for forest resource survey sampling equipment to optimize the sampling area of ​​a target forest area. The method includes the following steps:

[0006] The insect acoustic spectrum signals at various sampling nodes within the target forest area are monitored and collected using intelligent sensors. The insect acoustic spectrum signals are composed of characteristic frequency bands corresponding to the target insects.

[0007] For each sampling node, the difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and its neighboring sampling nodes in the current time period. The distribution variation characteristics of the insects at the sampling node in the spatial dimension are determined based on the insect acoustic spectrum signal at the sampling node in the current time period. The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is determined based on the spectral variation difference value and the distribution variation characteristics.

[0008] Historical environmental parameter data and historical pest and disease labeling data at the sampling nodes are obtained, and the influence correlation between different environmental parameters and pest and disease outbreaks is constructed based on the historical environmental parameter data and the historical pest and disease labeling data.

[0009] By combining the various influencing relationships with the environmental parameter data collected by the smart sensor for the current time period, a deviation anomaly analysis is performed on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node within the current time period to obtain the abnormal confidence value of the insect at the sampling node, and then the abnormal confidence value of the insect at each sampling node is obtained.

[0010] Based on all the outlier confidence values ​​and combined with spatial clustering algorithms, sampling nodes with outlier confidence values ​​exceeding the outlier threshold are clustered, thereby dividing the target forest area into multiple priority sampling areas for pests and diseases.

[0011] In some embodiments, determining the difference in spectral variation of the insect acoustic spectrum signal at a sampling node in the time dimension based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and its neighboring sampling nodes within the current time period specifically includes:

[0012] Acquire the insect acoustic spectrum signals of the sampling node and each adjacent sampling node within the current time period;

[0013] Spectral analysis was performed on the acoustic spectrum signals of each insect to extract the frequency distribution, energy density, and peak frequency of the sampling node and each adjacent sampling node.

[0014] The spectral variation characteristics of the insect acoustic spectrum between the sampling node and each adjacent sampling node in the current time period are obtained by calculating all frequency distributions, energy densities, and peak frequencies.

[0015] Based on all spectral variation characteristics, the difference in spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined.

[0016] In some embodiments, determining the spatial distribution variation characteristics of insects at sampling nodes based on the insect acoustic spectrum signals at the sampling nodes within the current time period specifically includes:

[0017] Acquire insect acoustic spectrum signals at multiple consecutive time points within the current time period at the sampling node;

[0018] Spectral analysis was performed on the insect acoustic spectrum signals at various time points to construct a temporal trajectory reflecting the insect acoustic spectrum activity;

[0019] Based on the time-series trajectory, the spatial distribution variation characteristics of insects at the sampling nodes are extracted.

[0020] In some embodiments, determining the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period based on the spectral variation difference value and the distribution variation characteristics specifically includes:

[0021] Obtain the difference in spectral variation and the corresponding distribution characteristics at the sampling nodes within the current time period;

[0022] The distribution variation characteristics are normalized to obtain the distribution variation characteristic values;

[0023] Determine the weighting coefficients for the spectral variation difference value and the distribution variation characteristic value;

[0024] The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is obtained by weighted fusion based on the difference value of the spectrum change, the characteristic value of the distribution change, and the corresponding weighting coefficient.

[0025] In some embodiments, constructing the correlation between different environmental parameters and pest outbreaks based on the historical environmental parameter data and the historical pest and disease annotation data specifically includes:

[0026] Construct a data alignment sample set between the historical environmental parameter data and the historical pest and disease annotation data;

[0027] Based on historical pest and disease labeling data, each data alignment sample in the data alignment sample set is labeled with a pest and disease outbreak status label;

[0028] Several key environmental factors affecting the outbreak of pests and diseases were extracted from the data-aligned sample set.

[0029] By combining all data aligned with the pest outbreak status labels and all key environmental factors, a correlation model between key environmental factors and pest outbreak status is established, thereby outputting the influence relationship between different environmental parameters and pest outbreaks.

[0030] In some embodiments, by combining various influencing relationships with environmental parameter data collected by the smart sensor for the current time period, a deviation anomaly analysis is performed on the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period to obtain the abnormal confidence value of the insect at the sampling node, specifically including:

[0031] Based on all the influence relationships, construct a set of environmental response sensitivity weights representing the contribution of each key environmental factor;

[0032] Obtain the real-time environmental parameter data corresponding to the sampling node within the current time period, and map it into an environmental parameter vector in a preset order;

[0033] The environmental response factor value for the current time period is calculated based on the environmental response sensitivity weight set and the environmental parameter vector.

[0034] The environmental response factor value is multiplied by the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period to obtain the abnormal confidence value of the insect at the sampling node.

[0035] In some embodiments, sampling nodes whose outlier confidence values ​​exceed an outlier threshold are clustered based on all outlier confidence values ​​combined with a spatial clustering algorithm, thereby delineating multiple priority sampling areas for pests and diseases within the target forest area. Specifically, this includes:

[0036] Obtain the pre-set abnormal threshold;

[0037] From all the sampling nodes, the sampling nodes corresponding to the abnormal confidence values ​​exceeding the abnormal threshold are selected to obtain multiple abnormal sampling nodes;

[0038] Spatial clustering algorithms are applied to cluster all abnormal sampling nodes, and then the sampling priority areas for multiple pests and diseases in the target forest area are planned based on the clustering results.

[0039] Secondly, this application provides a forest resource survey sampling device, which includes a sampling area optimization unit, the sampling area optimization unit comprising:

[0040] The acquisition module is used to monitor and acquire the insect acoustic spectrum signals at various sampling nodes in the target forest area through intelligent sensors. The insect acoustic spectrum signals are composed of characteristic frequency bands corresponding to the target insects.

[0041] The processing module is used to determine the difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and the adjacent sampling nodes in the current time period for each sampling node; determine the distribution variation characteristics of the insects at the sampling node in the spatial dimension based on the insect acoustic spectrum signal at the sampling node in the current time period; and determine the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period based on the spectral variation difference value and the distribution variation characteristics.

[0042] The processing module is also used to acquire historical environmental parameter data and historical pest and disease labeling data at the sampling node, and to construct the influence correlation between different environmental parameters and pest and disease outbreaks based on the historical environmental parameter data and the historical pest and disease labeling data.

[0043] The processing module is also used to perform a deviation anomaly analysis on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node in the current time period by combining the various influencing relationships with the environmental parameter data of the current time period collected by the smart sensor, to obtain the abnormal confidence value of the insect at the sampling node, and then obtain the abnormal confidence value of the insect at each sampling node.

[0044] The sampling area division module is used to cluster sampling nodes whose abnormal confidence values ​​exceed the abnormal threshold based on all abnormal confidence values ​​and spatial clustering algorithms, thereby dividing the target forest area into multiple priority sampling areas for pests and diseases.

[0045] Thirdly, this application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described sampling region optimization method.

[0046] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described sampling region optimization method.

[0047] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0048] The forest resource survey sampling equipment and method provided in this application monitors and collects insect acoustic spectrum signals at various sampling nodes within the target forest area using intelligent sensors. The insect acoustic spectrum signals are composed of characteristic frequency bands corresponding to the target insects. For each sampling node, the spectral variation difference value of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and adjacent sampling nodes within the current time period. The spatial distribution variation characteristics of the insects at the sampling node are determined based on the insect acoustic spectrum signals at the sampling node within the current time period. The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period is determined based on the spectral variation difference value and the distribution variation characteristics. The sampling method acquires the collected data. Historical environmental parameter data and historical pest and disease labeling data at sampling nodes are collected, and the influence correlation between different environmental parameters and pest and disease outbreaks is constructed based on the historical environmental parameter data and historical pest and disease labeling data. Combining the various influence correlations with the environmental parameter data of the current time period collected by the intelligent sensor, the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node in the current time period is analyzed to obtain the abnormal confidence value of the insect at the sampling node, and then the abnormal confidence value of the insect at each sampling node is obtained. Based on all the abnormal confidence values, the sampling nodes with abnormal confidence values ​​exceeding the abnormal threshold are clustered using a spatial clustering algorithm, thereby dividing the sampling priority areas of multiple pests and diseases in the target forest area.

[0049] Therefore, the spectral variation difference value described in this application reflects the uniqueness of the insect acoustic spectrum variation trend at the sampling node within the current time period. The distribution variation feature is used to quantify the distribution density change or spatial aggregation characteristics of insects in the current area. Based on the spectral variation difference value and the distribution variation feature, the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period is determined. This not only overcomes the limitations of single-dimensional analysis but also more accurately characterizes the degree of abnormality in the spatiotemporal coordination of insect activities, providing a core basis for subsequent calculation of anomaly confidence values ​​in conjunction with environmental parameters. The influence correlation is used to quantify the contribution intensity of different environmental parameters to the formation mechanism of pests and diseases. Combining each influence correlation with the current data collected by the intelligent sensor... Environmental parameter data for a given time period are used to perform a deviation anomaly analysis on the spatiotemporal consistency deviation of insect acoustic spectra at sampling nodes within the current time period, resulting in anomaly confidence values ​​for insects at the sampling nodes. This process generates anomaly confidence values ​​by combining spatiotemporal consistency deviation with the correlation between different environmental parameters and pest outbreaks. These anomaly confidence values, to a certain extent, reflect the coupling effect between abnormal behavior of target insects at the corresponding nodes and environmental stress. Finally, based on all anomaly confidence values, a spatial clustering algorithm is used to cluster sampling nodes whose anomaly confidence values ​​exceed anomaly thresholds, thereby delineating multiple priority sampling areas for pests and diseases within the target forest area. In summary, the scheme in this application can optimize the sampling area of ​​the target forest by combining multi-dimensional anomaly analysis of monitoring data. Attached Figure Description

[0050] Figure 1 This is an exemplary flowchart of a sampling region optimization method according to some embodiments of this application;

[0051] Figure 2 This is a schematic diagram illustrating the application scenario of the sampling area optimization unit according to some embodiments of this application;

[0052] Figure 3 This is a schematic flowchart illustrating the process of determining the difference value of spectral variation according to some embodiments of this application;

[0053] Figure 4 This is a schematic diagram of the sampling region optimization unit shown in some embodiments of this application;

[0054] Figure 5 This is a schematic diagram of the structure of a computer device implementing a sampling region optimization method according to some embodiments of this application. Detailed Implementation

[0055] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0056] refer to Figure 1The figure is an exemplary flowchart of a sampling region optimization method according to some embodiments of this application. The sampling region optimization method mainly includes the following steps:

[0057] In step 101, the acoustic spectrum signals of insects at each sampling node in the target forest area are monitored and collected by intelligent sensors. The acoustic spectrum signals of insects are composed of characteristic frequency bands corresponding to the target insects.

[0058] In specific implementation, monitoring and collecting insect acoustic spectrum signals at various sampling nodes within the target forest area using intelligent sensors can be achieved in the following way: Several sampling nodes are pre-deployed within the target forest area, taking into account the living habits of the target insects and the preset spatial density. Each sampling node integrates an environmentally adaptable acoustic sensor module to continuously monitor natural sound sources in its surrounding environment. During actual operation, the sampling nodes synchronously start the acoustic spectrum acquisition task according to the set sampling cycle and store the acquired raw audio signals in digital format. Then, the signals are processed in real time using a built-in recognition module or edge computing unit to extract the insect acoustic spectrum signals within the characteristic frequency band corresponding to the target insects. Other methods can also be used in other embodiments, which are not limited here.

[0059] It should be noted that the smart sensor in this application is an acoustic sensor. The acoustic sensor preferably has high sensitivity and wide frequency response capability, and is combined with a front-end filter device set inside the node to shield interference information such as wind noise and raindrop sound. The characteristic frequency band corresponding to the target insect can be dynamically set according to the frequency range determined by the target insect species in literature or experiments to ensure that the extracted acoustic spectrum data has biological relevance.

[0060] It should be noted that the target insects in this application may include pine sawyer beetles, stink bugs, poplar leafminer moths, pine caterpillars, and longhorn beetles, which are destructive pests of forest resources. The specific target species can be set according to the climate zone, season, and monitoring purpose of the forest area, and this invention does not limit them.

[0061] In some embodiments, reference Figure 2 As shown in the figure, this figure is a schematic diagram of the application scenario of the sampling area optimization unit shown in some embodiments of this application. The figure includes three main components: a collection device, a server, and a data storage device. The collection device is responsible for collecting insect acoustic spectrum signals and environmental parameter data at various sampling nodes in the target forest area, and sending the collected insect acoustic spectrum signals and environmental parameter data to the server through a communication network. The sampling area optimization unit runs in the server, and the server stores the processing results in the data storage device and visualizes them.

[0062] In step 102, for each sampling node, the difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and its neighboring sampling nodes in the current time period. The distribution variation characteristics of the insects at the sampling node in the spatial dimension are determined based on the insect acoustic spectrum signal at the sampling node in the current time period. The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is determined based on the difference in spectral variation and the distribution variation characteristics.

[0063] It should be noted that in this application, the target sampling node is taken as the center, and other sampling nodes within a specified distance from the target sampling node are selected as the adjacent sampling nodes of the target sampling node. The specified distance can be adaptively set according to the actual situation. For example, when the density of the monitored target insects is higher, the specified distance can be set smaller, and vice versa. In other embodiments, the specified distance can also be set according to other actual situations, which is not limited here.

[0064] In some embodiments, reference Figure 3 As shown in the figure, this is a flowchart illustrating the process of determining the spectral variation difference value in some embodiments of this application. In this embodiment, the determination of the spectral variation difference value of the insect acoustic spectrum signal at the sampling node in the time dimension based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and its adjacent sampling nodes within the current time period can be achieved by the following steps:

[0065] In step 1031, the insect acoustic spectrum signals of the sampling nodes and each adjacent sampling node within the current time period are obtained;

[0066] In step 1032, spectral analysis is performed on the acoustic spectrum signals of each insect to extract the frequency distribution, energy density and peak frequency of the sampling node and each adjacent sampling node;

[0067] In step 1033, the spectral variation characteristics of the insect acoustic spectrum between the sampling node and each adjacent sampling node in the current time period are calculated by using all frequency distributions, energy densities and peak frequencies.

[0068] In step 1034, the difference in spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on all spectral variation characteristics.

[0069] In practice, the frequency distribution, energy density, and peak frequency of each insect acoustic spectrum signal are extracted by performing spectral analysis. This can be achieved as follows: First, the collected insect acoustic spectrum signals are preprocessed, including noise removal and filtering, to ensure signal quality. Next, a Fast Fourier Transform is performed on each preprocessed insect acoustic spectrum signal to convert the time-domain signal into a frequency-domain signal. The converted frequency-domain signal represents the intensity distribution (i.e., frequency distribution) of different frequencies. Then, the energy density of different frequency intervals is obtained by integrating the squared amplitude of each frequency interval. The energy density represents the energy proportion of each frequency interval in the signal. Finally, the peak frequency is extracted from the spectrum. This can be done by finding the frequency point with the maximum energy in the spectrum and determining its corresponding frequency value, which is the peak frequency, effectively representing the characteristic frequency of the insect acoustic spectrum signal.

[0070] It should be noted that, for example, short-time Fourier transform can also be used to perform spectral analysis on the collected insect acoustic spectrum signal. That is, firstly, the original time-domain signal is segmented according to a certain time window, and a window function is applied to each small segment of the signal to reduce boundary effects. Then, frequency domain transformation is performed on each segment of the signal to obtain the spectral changes of the signal in the time series. Then, the spectrum within each time window is integrated and statistically analyzed to obtain the frequency distribution, and the energy distribution value of each frequency band is extracted as the energy density. Furthermore, the frequency with the largest frequency amplitude in each time window is identified as the peak frequency of that period. Alternatively, the Mel frequency cepstral coefficient extraction technique, which is widely used in speech signal analysis, can be used to perform spectral analysis on the insect acoustic spectrum signal of the sampling node. Other methods can also be used in other embodiments, which will not be elaborated here.

[0071] In specific implementation, the spectral variation characteristics of the insect acoustic spectrum between the sampling node and each adjacent sampling node in the current time period can be obtained by calculating all frequency distributions, energy densities, and peak frequencies. This can be achieved in the following way: First, the frequency distribution, energy density, and peak frequency of each sampling node in the current time period are compared in pairs with its corresponding adjacent sampling nodes. For frequency distribution, the differences in the overall distribution pattern of frequency amplitude can be compared by analyzing the differences. For energy density, the differences in its energy coverage range in the main frequency band can be assessed. For peak frequency, the degree of numerical deviation can be directly compared. Next, the differences of the above three types of characteristics are uniformly quantified to construct the difference degree describing the spectral variation. Preferably, according to the ecological response sensitivity of the insect acoustic spectrum signal in different frequency bands, different analysis priorities can be set for frequency distribution, energy density, and peak frequency. By weighted fusion of each difference degree, the spectral variation characteristics of the insect acoustic spectrum between the sampling node and each adjacent sampling node in the current time period are finally obtained. The spectral variation characteristics represent the degree of difference between the insect acoustic spectrum of the sampling node and each adjacent sampling node in the current time period.

[0072] In specific implementation, the difference in spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension based on all spectral variation characteristics can be achieved in the following way: corresponding weights can be set by the spatial distance between the sampling node and its adjacent sampling nodes. For example, the greater the distance, the smaller the weight, and the closer the distance, the larger the weight. Then, the spectral variation characteristics between the sampling node and its corresponding adjacent sampling nodes are weighted and fused based on each weight to obtain the difference in spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension. The difference in spectral variation reflects the uniqueness of the insect acoustic spectrum variation trend at the sampling node in the current time period, providing dynamic temporal support for subsequent spatial feature fusion and abnormal confidence value calculation.

[0073] In some embodiments, determining the spatial distribution characteristics of insects at sampling nodes based on the insect acoustic spectrum signals at the sampling nodes within the current time period can be achieved using the following steps:

[0074] Acquire insect acoustic spectrum signals at multiple consecutive time points within the current time period at the sampling node;

[0075] Spectral analysis was performed on the insect acoustic spectrum signals at various time points to construct a temporal trajectory reflecting the insect acoustic spectrum activity;

[0076] Based on the time-series trajectory, the spatial distribution variation characteristics of insects at the sampling nodes are extracted.

[0077] In practice, the spectral analysis of insect acoustic signals at various time points, and the subsequent construction of a time-series trajectory reflecting insect acoustic activity, can be achieved as follows: First, spectral analysis is performed on the insect acoustic signals at various time points to extract the acoustic attributes of the target insect within the corresponding frequency band. These acoustic attributes include, but are not limited to, energy peaks, dominant frequency positions, and spectral envelope morphology. Next, these acoustic attributes are arranged chronologically to construct a time-series trajectory reflecting insect acoustic activity. This time-series trajectory reflects the insect activity performance of the sampling node within the current time period. Specifically, based on this time-series trajectory, characterization parameters for the sampling node are extracted. The spatial distribution variation characteristics of insects at a location can be achieved in the following way: the distribution variation characteristics of insects at a sampling node can be constructed based on the peak change rate of energy peak in the time gradient, the number of jumps and the average jump amplitude of the dominant frequency position in the time gradient, and the change rate of the envelope bandwidth of the spectral envelope shape in the time gradient. The distribution variation characteristics can be used to quantify the distribution density change or spatial aggregation and dispersion characteristics of insects in the current area, providing accurate spatial dynamic basis for subsequent spatiotemporal consistency modeling. Other methods can also be used in other embodiments, which are not limited here.

[0078] It should be noted that, in addition to determining the spatial distribution characteristics of insects at the sampling nodes based on the insect acoustic spectrum signals at the sampling nodes during the current time period, in another embodiment, images of the sampling nodes can be acquired by an optical camera, and the spatial distribution characteristics of insects at the sampling nodes can be extracted from the acquired images using image processing techniques. Other methods can also be used in other embodiments, which will not be elaborated here.

[0079] In addition, in some embodiments, determining the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period based on the spectral variation difference value and the distribution variation characteristics can be achieved by the following steps:

[0080] Obtain the difference in spectral variation and the corresponding distribution characteristics at the sampling nodes within the current time period;

[0081] The distribution variation characteristics are normalized to obtain the distribution variation characteristic values;

[0082] Determine the weighting coefficients for the spectral variation difference value and the distribution variation characteristic value;

[0083] The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is obtained by weighted fusion based on the difference value of the spectrum change, the characteristic value of the distribution change, and the corresponding weighting coefficient.

[0084] In specific implementation, the distribution change characteristics are normalized to obtain the distribution change characteristic value. This can be achieved by fusing the peak change rate, the number of jumps, the average jump amplitude, and the envelope bandwidth change rate in the distribution change characteristics according to preset weighting coefficients. The distribution change characteristic value is a comprehensive representation index of the distribution change characteristics. In specific implementation, the weighting coefficients of the spectral change difference value and the distribution change characteristic value can be determined by setting the weighting coefficients of the spectral change difference value and the distribution change characteristic value based on the contribution ratio of spectral change difference and distribution change to abnormal insect activity in historical sample training data. For example, if the distribution change characteristic value has a more stable discrimination ability in dynamic change scenarios, it can be assigned a higher weighting coefficient. Other methods can also be used to set the weighting coefficients in other embodiments, which are not limited here.

[0085] It should be noted that the spatiotemporal consistency deviation in this application characterizes the degree of coordination between the temporal evolution trend and the spatial diffusion trend of insect acoustic spectrum activity at the node, providing a unified measurement basis for subsequent anomaly detection and sampling area division. In other embodiments, a fusion method based on fuzzy logic or interval credibility reasoning can also be introduced to calculate the spatiotemporal consistency deviation, which is not limited in this application.

[0086] It should be noted that the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is determined based on the difference value of the spectrum change and the distribution change characteristics. This not only overcomes the limitations of single-dimensional analysis, but also more accurately characterizes the degree of abnormality in the spatiotemporal coordination of insect activities, providing a core basis for subsequent calculation of abnormal confidence values ​​in combination with environmental parameters.

[0087] In step 103, historical environmental parameter data and historical pest and disease labeling data at the sampling node are obtained, and the influence correlation between different environmental parameters and pest and disease outbreaks is constructed based on the historical environmental parameter data and the historical pest and disease labeling data.

[0088] In specific implementation, the historical environmental parameter data and historical pest and disease labeling data at the sampling nodes can be obtained in the following way: the historical forestry database can be accessed to retrieve the historical environmental parameter data and historical pest and disease labeling data corresponding to each sampling node. The historical environmental parameter data may include temperature, humidity, soil moisture content and carbon dioxide concentration, etc., and the pest and disease labeling data may include the time of occurrence of target insect pests and diseases, insect species, occurrence level and spatial distribution information, etc.

[0089] It should be noted that in this application, when pre-deploying several sampling nodes, not only are each sampling node equipped with an environmentally adaptable acoustic sensor module, but each sampling node also deploys multiple environmental parameter monitoring sensors. The environmental parameter data obtained by the monitoring sensors are organized and stored in the forestry database according to different time granularities such as daily, weekly, or monthly. The monitoring sensors include: temperature monitoring sensors, humidity monitoring sensors, soil moisture content monitoring sensors, and carbon dioxide concentration monitoring sensors, etc. At the same time, combined with manual inspection records, remote sensing image analysis results, or historical reports from the pest and disease monitoring platform, historical pest and disease labeling data at the sampling nodes is extracted. The historical pest and disease labeling data may include the time of occurrence of target insect pests and diseases, insect species, occurrence level, and spatial distribution information, and can be synchronously mapped according to the time alignment method with environmental parameters.

[0090] In some embodiments, constructing the correlation between different environmental parameters and pest outbreaks based on the historical environmental parameter data and the historical pest and disease annotation data can be achieved by the following steps:

[0091] Construct a data alignment sample set between the historical environmental parameter data and the historical pest and disease annotation data;

[0092] Based on historical pest and disease labeling data, each data alignment sample in the data alignment sample set is labeled with a pest and disease outbreak status label;

[0093] Several key environmental factors affecting the outbreak of pests and diseases were extracted from the data-aligned sample set.

[0094] By combining all data aligned with the pest outbreak status labels and all key environmental factors, a correlation model between key environmental factors and pest outbreak status is established, thereby outputting the influence relationship between different environmental parameters and pest outbreaks.

[0095] In specific implementation, the data alignment sample set for the historical environmental parameter data and the historical pest and disease labeling data can be constructed in the following way: The acquired historical environmental parameter data and historical pest and disease labeling data can be preprocessed, including timestamp standardization, missing value handling, and spatial node alignment. Subsequently, based on unified sampling node numbers or spatial coordinates, the environmental parameter data and pest and disease labeling data are paired according to the same time window (e.g., day, week, or month) to form a data alignment sample set with input (i.e., environmental parameters) and output (i.e., pest and disease status). Currently, labeling each data-aligned sample in the data-aligned sample set with a pest outbreak status label based on historical pest labeling data can be achieved in the following ways: binary labeling (i.e., whether an outbreak has occurred) or multi-level labeling (i.e., mild, moderate, severe) can be used to label each data-aligned sample in the data-aligned sample set with a pest outbreak status label. The pest outbreak status label can be generated based on historical pest severity, pest density, or expert judgment criteria; no specific limitation is made here. In practice, multiple key environmental factors influencing pest outbreaks are extracted from the data-aligned sample set. Factors can be implemented in the following way: In the constructed data-aligned sample set, the environmental parameters within each sampling time window are statistically analyzed to extract multiple key environmental factors affecting the outbreak of pests and diseases, such as the average value and fluctuation range of temperature and humidity, changes in soil moisture content, and the number of consecutive dry or wet days. In specific implementation, a correlation model between key environmental factors and the outbreak status is established by combining all data-aligned samples labeled with the outbreak status label and all key environmental factors. The influence correlation between different environmental parameters and the outbreak of pests and diseases can be output in the following way: Based on all the labeled data-aligned samples mentioned above, a mapping relationship model between environmental parameters and the outbreak status can be constructed using data mining methods (such as decision trees, random forests, support vector machines, etc.). This model can evaluate the degree of influence of different key environmental factors on the outbreak of pests and diseases through training and cross-validation, and quantify the correlation or causal strength between each key environmental factor and the outbreak event. Finally, the modeling results are transformed into interpretable influence correlations. For example, a set of environmental response sensitive weight vectors is generated for subsequent weighted fusion processing so that they can be used in conjunction with the environmental parameters of the current time period to achieve dynamic adjustment of abnormal confidence values.

[0096] It should be noted that the influence correlation described in this application is used to quantify the contribution intensity of different environmental parameters to the formation mechanism of pests and diseases. More specifically, by combining all data aligned samples labeled with the pest and disease outbreak status with all key environmental factors to establish a correlation model between key environmental factors and the pest and disease outbreak status, and then outputting the influence correlation between different environmental parameters and pest and disease outbreaks, the following method can be used: First, preprocess the data aligned sample set of all labeled pest and disease outbreak statuses. The preprocessing includes missing value imputation, outlier removal, and normalization of the dimensions of key environmental factors to ensure that the data of different dimensions have a uniform analytical scale. Then, construct an environmental factor vector based on the extracted key environmental factors, and use the pest and disease outbreak status label of each sample as a supervision variable to form an input sample set for modeling. Next, select appropriate environmental features-labels. Machine learning algorithms for relationship modeling are used for training and modeling. Preferably, a logistic regression model can be used to analyze the linear influence trend of various key environmental factors on the outbreak of pests and diseases. Alternatively, a random forest model can be used to discover the nonlinear interaction relationship between multidimensional environmental factors. In some embodiments, a cross-validation mechanism can also be introduced to improve the generalization performance of the model. After training, the importance ranking of each key environmental factor and its mapping relationship with the pest and disease state label are evaluated. The output is used to reflect the influence correlation of different environmental parameters on the risk of pest and disease outbreaks under different states. The influence correlation can be used to quantify the contribution intensity of different environmental parameters to the formation mechanism of pests and diseases and serve as the weight basis in the subsequent calculation of anomaly confidence values. In other embodiments, time series modeling, causal inference analysis and other techniques can be combined to further expand the model accuracy and application breadth, which is not limited here.

[0097] In step 104, by combining the various influencing relationships with the environmental parameter data collected by the smart sensor for the current time period, a deviation anomaly analysis is performed on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node within the current time period to obtain the abnormal confidence value of the insect at the sampling node, and then the abnormal confidence value of the insect at each sampling node is obtained.

[0098] In some embodiments, the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period is analyzed by combining various influencing relationships with the environmental parameter data collected by the smart sensor for the current time period, and the abnormal confidence value of the insect at the sampling node is obtained by the following steps:

[0099] Based on all the influence relationships, construct a set of environmental response sensitivity weights representing the contribution of each key environmental factor;

[0100] Obtain the real-time environmental parameter data corresponding to the sampling node within the current time period, and map it into an environmental parameter vector in a preset order;

[0101] The environmental response factor value for the current time period is calculated based on the environmental response sensitivity weight set and the environmental parameter vector.

[0102] The environmental response factor value is multiplied by the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period to obtain the abnormal confidence value of the insect at the sampling node.

[0103] In practical implementation, constructing an environmental response sensitivity weight set representing the contribution of each key environmental factor based on all influence relationships can be achieved in the following way: Historical data can be trained using machine learning algorithms based on all influence relationships to evaluate the influence weight of each key environmental factor on pest outbreaks. Then, through correlation analysis and importance ranking of key environmental factors, combined with error feedback during training, the sensitivity weight of each key environmental factor in spatiotemporal consistency analysis can be optimized. For example, if a key environmental factor has a strong correlation with pest outbreaks in most cases, it can be assigned a higher weight; conversely, if an environmental factor has a smaller impact on pest outbreaks, its corresponding weight can be lower. Finally, based on these training and analysis results, an environmental response sensitivity weight set is constructed. This weight set reflects the triggering ability of different environmental parameters on abnormal pest activity in the current region. This set can be weighted and fused with real-time environmental parameter data in subsequent anomaly confidence value calculations to dynamically adjust spatiotemporal consistency deviations. Currently, acquiring real-time environmental parameter data corresponding to sampling nodes within the current time period and mapping it into an environmental parameter vector according to a preset order can be achieved in the following way: Real-time environmental parameter data is obtained by real-time collection of environmental parameter data corresponding to sampling nodes within the current time period using various deployed environmental parameter monitoring sensors. The collected real-time environmental parameter data is then preprocessed, including denoising and calibration to reduce noise interference in the real-time environmental parameter data. The preprocessed real-time environmental parameter data is then standardized to avoid the influence of different parameter dimensions. Standardization can employ Z-score standardization or Min-Max standardization. Finally, based on the order of the monitoring sensors corresponding to the sampling nodes or the set parameter priority, the standardized data from each monitoring sensor are arranged sequentially into a vector. For example, if the order is temperature, humidity, and soil moisture content, then the environmental parameter vector can be represented as: [temperature value, humidity value, soil moisture content value]. Each item in this vector represents the standardized data of a certain environmental factor within the current time period.

[0104] In addition, in specific implementation, the environmental response factor value for the current period can be calculated based on the set of environmental response sensitive weights and the environmental parameter vector in the following way: each item in the environmental parameter vector (i.e., each key environmental factor) is multiplied one by one with its corresponding environmental response sensitive weight, and then all the product results are summed to obtain the final environmental response factor value. The environmental response factor value reflects the degree of influence of the environmental conditions of the sampling node on the outbreak of target insect pests and diseases in the current period.

[0105] It should be noted that the abnormal confidence value in this application represents the degree to which the behavior of the target insect at the corresponding node deviates from the normal pattern in the current spatiotemporal context, and can be used to reflect the coupling effect between the abnormal behavior of the target insect at the corresponding node and environmental stress.

[0106] In step 105, based on all the abnormal confidence values ​​and combined with the spatial clustering algorithm, the sampling nodes whose abnormal confidence values ​​exceed the abnormal threshold are clustered, thereby dividing the sampling priority areas for multiple diseases and pests in the target forest area.

[0107] In some embodiments, clustering sampling nodes whose outlier confidence values ​​exceed an outlier threshold based on all outlier confidence values ​​combined with a spatial clustering algorithm, thereby delineating multiple priority sampling areas for pests and diseases within the target forest area, can be achieved through the following steps:

[0108] Obtain the pre-set abnormal threshold;

[0109] From all the sampling nodes, the sampling nodes corresponding to the abnormal confidence values ​​exceeding the abnormal threshold are selected to obtain multiple abnormal sampling nodes;

[0110] Spatial clustering algorithms are applied to cluster all abnormal sampling nodes, and then the sampling priority areas for multiple pests and diseases in the target forest area are planned based on the clustering results.

[0111] It should be noted that the anomaly threshold in this application can be set to an appropriate anomaly confidence value threshold based on historical data or through methods such as cross-validation. This anomaly threshold is used to filter out sampling nodes that exceed this value. Specifically, if the anomaly confidence value of a certain sampling node exceeds this anomaly threshold, it indicates that the abnormal activity of the sampling node is significant and belongs to a high-risk area.

[0112] In specific implementation, the spatial clustering algorithm can be used to cluster all abnormal sampling nodes in the following way: set an appropriate distance threshold and minimum number of sample points, and then cluster all abnormal sampling nodes based on the distance threshold and minimum number of sample points to obtain multiple clusters of abnormal sampling nodes. Other methods can also be used in other embodiments, which will not be elaborated here.

[0113] In addition, in specific implementation, the sampling priority areas for multiple pests and diseases in the target forest area can be planned based on the clustering results in the following way: the sampling area is planned according to the clusters of abnormal sampling nodes. For example, for each cluster, the sampling range is extended outward by 10 meters from the center of each abnormal sampling node in the cluster. Finally, the sampling area of ​​the corresponding cluster is planned according to the maximum sampling range of each abnormal sampling node in the cluster, and the sampling area is used as the sampling priority area of ​​the corresponding cluster. In other embodiments, the sampling range can be set according to the actual situation or other methods can be used to plan the sampling priority area. This is not limited here.

[0114] Furthermore, in another aspect of this application, in some embodiments, this application provides a forest resource survey sampling device, which includes a sampling area optimization unit, with reference to... Figure 4 The figure is a schematic diagram of the structure of a sampling region optimization unit according to some embodiments of this application. The sampling region optimization unit 400 includes: a data acquisition module 401, a processing module 402, and a sampling region division module 403, which are described below:

[0115] The acquisition module 401 in this application is mainly used to monitor and acquire the insect acoustic spectrum signals at each sampling node in the target forest area through intelligent sensors. The insect acoustic spectrum signals are composed of the characteristic frequency bands corresponding to the target insects.

[0116] Processing module 402, in this application, is mainly used to determine the difference value of the spectral change of the insect acoustic spectrum signal at the sampling node in the time dimension based on the spectral change characteristics of the insect acoustic spectrum between the sampling node and the adjacent sampling nodes in the current time period for each sampling node; determine the distribution change characteristics of the insect at the sampling node in the spatial dimension based on the insect acoustic spectrum signal at the sampling node in the current time period; and determine the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period based on the spectral change difference value and the distribution change characteristics.

[0117] The processing module 402 described in this application is also used to acquire historical environmental parameter data and historical pest and disease labeling data at the sampling node, and to construct the influence correlation between different environmental parameters and pest and disease outbreaks based on the historical environmental parameter data and the historical pest and disease labeling data.

[0118] The processing module 402 described in this application is also used to perform a deviation anomaly analysis on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node in the current time period by combining the various influence correlations with the environmental parameter data of the current time period collected by the smart sensor, to obtain the abnormal confidence value of the insect at the sampling node, and then to obtain the abnormal confidence value of the insect at each sampling node.

[0119] The sampling area division module 403 in this application is mainly used to cluster sampling nodes whose abnormal confidence values ​​exceed the abnormal threshold based on all abnormal confidence values ​​and spatial clustering algorithms, thereby dividing multiple priority sampling areas for pests and diseases in the target forest area.

[0120] Each module in the aforementioned sampling region optimization unit can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0121] In another embodiment, this application provides a computer device, which may be a server, and its internal structure diagram may be as follows. Figure 5 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores sampling region optimization data. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a sampling region optimization method.

[0122] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0123] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described sampling region optimization method embodiment.

[0124] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described sampling region optimization method embodiment.

[0125] In one embodiment, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps described in the above-described sampling region optimization method embodiment.

[0126] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0127] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0128] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A sampling area optimization method, used by forest resource survey sampling equipment to optimize the sampling area of ​​a target forest area, characterized in that, The method includes the following steps: The insect acoustic spectrum signals at various sampling nodes within the target forest area are monitored and collected using intelligent sensors. The insect acoustic spectrum signals are composed of characteristic frequency bands corresponding to the target insects. For each sampling node, the difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and its neighboring sampling nodes in the current time period. The distribution variation characteristics of the insects at the sampling node in the spatial dimension are determined based on the insect acoustic spectrum signal at the sampling node in the current time period. The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is determined based on the spectral variation difference value and the distribution variation characteristics. Historical environmental parameter data and historical pest and disease labeling data at the sampling nodes are obtained, and the influence correlation between different environmental parameters and pest and disease outbreaks is constructed based on the historical environmental parameter data and the historical pest and disease labeling data. By combining the various influencing relationships with the environmental parameter data collected by the smart sensor for the current time period, a deviation anomaly analysis is performed on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node within the current time period to obtain the abnormal confidence value of the insect at the sampling node, and then the abnormal confidence value of the insect at each sampling node is obtained. Based on all the outlier confidence values ​​and combined with spatial clustering algorithms, sampling nodes with outlier confidence values ​​exceeding the outlier threshold are clustered, thereby dividing the target forest area into multiple priority sampling areas for pests and diseases.

2. The method as described in claim 1, characterized in that, The difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined based on the spectral variation characteristics between the sampling node and its neighboring sampling nodes within the current time period. Specifically, this includes: Acquire the insect acoustic spectrum signals of the sampling node and each adjacent sampling node within the current time period; Spectral analysis was performed on the acoustic spectrum signals of each insect to extract the frequency distribution, energy density, and peak frequency of the sampling node and each adjacent sampling node. The spectral variation characteristics of the insect acoustic spectrum between the sampling node and each adjacent sampling node in the current time period are obtained by calculating all frequency distributions, energy densities, and peak frequencies. Based on all spectral variation characteristics, the difference in spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension is determined.

3. The method as described in claim 1, characterized in that, Determining the spatial distribution characteristics of insects at sampling nodes based on the insect acoustic spectrum signals at the current time period specifically includes: Acquire insect acoustic spectrum signals at multiple consecutive time points within the current time period at the sampling node; Spectral analysis was performed on the insect acoustic spectrum signals at various time points to construct a temporal trajectory reflecting the insect acoustic spectrum activity; Based on the time-series trajectory, the spatial distribution variation characteristics of insects at the sampling nodes are extracted.

4. The method as described in claim 1, characterized in that, The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period is determined based on the spectral variation difference value and the distribution variation characteristics, specifically including: Obtain the difference in spectral variation and the corresponding distribution characteristics at the sampling nodes within the current time period; The distribution variation characteristics are normalized to obtain the distribution variation characteristic values; Determine the weighting coefficients for the spectral variation difference value and the distribution variation characteristic value; The spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period is obtained by weighted fusion based on the difference value of the spectrum change, the characteristic value of the distribution change, and the corresponding weighting coefficient.

5. The method as described in claim 1, characterized in that, The specific steps for constructing the correlation between different environmental parameters and pest outbreaks based on the historical environmental parameter data and the historical pest and disease annotation data include: Construct a data alignment sample set between the historical environmental parameter data and the historical pest and disease annotation data; Based on historical pest and disease labeling data, each data alignment sample in the data alignment sample set is labeled with a pest and disease outbreak status label; Several key environmental factors affecting the outbreak of pests and diseases were extracted from the data-aligned sample set. By combining all data aligned with the pest outbreak status labels and all key environmental factors, a correlation model between key environmental factors and pest outbreak status is established, thereby outputting the influence relationship between different environmental parameters and pest outbreaks.

6. The method as described in claim 1, characterized in that, By combining various influencing relationships with the environmental parameter data collected by the intelligent sensor for the current time period, a deviation anomaly analysis is performed on the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node within the current time period. The specific abnormal confidence values ​​of the insects at the sampling node include: Based on all the influence relationships, construct a set of environmental response sensitivity weights representing the contribution of each key environmental factor; Obtain the real-time environmental parameter data corresponding to the sampling node within the current time period, and map it into an environmental parameter vector in a preset order; The environmental response factor value for the current time period is calculated based on the environmental response sensitivity weight set and the environmental parameter vector. The environmental response factor value is multiplied by the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period to obtain the abnormal confidence value of the insect at the sampling node.

7. The method as described in claim 1, characterized in that, Based on all the outlier confidence values ​​and spatial clustering algorithms, sampling nodes with outlier confidence values ​​exceeding the outlier threshold are clustered to further delineate multiple priority sampling areas for pests and diseases within the target forest area, specifically including: Obtain the pre-set abnormal threshold; From all the sampling nodes, the sampling nodes corresponding to the abnormal confidence values ​​exceeding the abnormal threshold are selected to obtain multiple abnormal sampling nodes; Spatial clustering algorithms are applied to cluster all abnormal sampling nodes, and then the sampling priority areas for multiple pests and diseases in the target forest area are planned based on the clustering results.

8. A forest resource survey sampling device, comprising a sampling area optimization unit, characterized in that, The sampling area optimization unit includes: The acquisition module is used to monitor and acquire the insect acoustic spectrum signals at various sampling nodes in the target forest area through intelligent sensors. The insect acoustic spectrum signals are composed of characteristic frequency bands corresponding to the target insects. The processing module is used to determine the difference in the spectral variation of the insect acoustic spectrum signal at the sampling node in the time dimension based on the spectral variation characteristics of the insect acoustic spectrum between the sampling node and the adjacent sampling nodes in the current time period for each sampling node; determine the distribution variation characteristics of the insects at the sampling node in the spatial dimension based on the insect acoustic spectrum signal at the sampling node in the current time period; and determine the spatiotemporal consistency deviation of the insect acoustic spectrum at the sampling node in the current time period based on the spectral variation difference value and the distribution variation characteristics. The processing module is also used to acquire historical environmental parameter data and historical pest and disease labeling data at the sampling node, and to construct the influence correlation between different environmental parameters and pest and disease outbreaks based on the historical environmental parameter data and the historical pest and disease labeling data. The processing module is also used to perform a deviation anomaly analysis on the spatiotemporal consistency deviation of the insect sound spectrum at the sampling node in the current time period by combining the various influencing relationships with the environmental parameter data of the current time period collected by the smart sensor, to obtain the abnormal confidence value of the insect at the sampling node, and then obtain the abnormal confidence value of the insect at each sampling node. The sampling area division module is used to cluster sampling nodes whose abnormal confidence values ​​exceed the abnormal threshold based on all abnormal confidence values ​​and spatial clustering algorithms, thereby dividing the target forest area into multiple priority sampling areas for pests and diseases.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the sampling region optimization method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the sampling region optimization method as described in any one of claims 1 to 7.

Images