Pest Identification System and Method Based on Data Classification

Abstract

The present invention discloses a pest identification system and method based on data classification, which relates to the technical field of pest identification. A trained pest classification model is used for identification and classification, and a pest identification data set is constructed. If the damage degree of the identified pest exceeds the expectation, pest outbreak data is collected, and a trained pest outbreak factor identification model is used to analyze the causes of pest outbreaks in the pine forest area. A pest value is generated from the pest outbreak data. When the pest value exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, a planting treatment plan is output by the planting treatment plan library; when the adjustment of the pine forest planting state in the pine forest area fails to achieve the expected effect, a corresponding pest treatment plan is output by the pest emergency treatment knowledge graph. By constructing a multi-level alarm mechanism and a multi-level emergency treatment mechanism, pest treatment is realized at different levels when there is already a pest outbreak, reducing the impact of pests.

Description

Technical Field

[0001] The present invention relates to the technical field of pest identification, in particular to a pest identification system and method based on data classification. Background Art

[0002] As a member of the Scolytidae family in Coleoptera, Tomicus piniperda is a natural enemy of pine trees, especially preferring precious tree species such as Pinus yunnanensis and Pinus armandii. Their large-scale outbreaks can quickly damage entire pine forests, causing huge losses to forest resources. In addition, since pine forests play an important role in the ecosystem, such as soil and water conservation and climate regulation, the rampant spread of Tomicus piniperda may also have a profound impact on the ecological environment. To effectively control the damage caused by Tomicus piniperda, comprehensive prevention and control measures need to be taken, including forest management, biological control, chemical control, and strengthening quarantine.

[0003] The impact of Tomicus piniperda on Pinus yunnanensis is extremely significant. They can directly feed on pine shoots and overwinter in the tree trunks in winter. The damage to shoots and trunks causes the growth of Pinus yunnanensis to be hindered, the leaves to turn red and fall off, and ultimately the entire tree to die. In addition, the pest infestation may also spread viruses and pathogens, exacerbating the occurrence of the pest infestation and threatening the forest ecological security and forestry economic production in Yunnan.

[0004] In the Chinese invention patent with the application publication number CN113743513A, a pest classification method and device based on multi-feature data are disclosed; the method includes: acquiring an image to be recognized; extracting various feature data of the image; determining the optimal information combination method; combining the feature data with the original image information according to the optimal information combination method to obtain fusion information; inputting the fusion information into a pest classification model for recognition to obtain a pest classification result; the pest classification model is a trained deep convolutional network model. It can effectively identify the pest categories, has strong feature expression and classification capabilities, provides information for subsequent pest control such as intelligent variable spraying, and also has good classification accuracy for normal captured images and enlarged images that are not targeted at the pest body.

[0005] Before pest control, it is necessary to identify and classify various types of pests. The existing identification methods usually collect image data of insects through an image acquisition device, and then determine information such as the type and quantity of insects through image recognition. However, in fact, there may be certain similarities in the morphology between different types of insects. Especially when the current planting state of the pine forest meets the conditions for pest outbreaks, it may lead to a large increase in the number of insects with similar species and morphologies at the same time, resulting in a relatively high risk of misidentifying and classifying pests. Further, in the case where multiple types of pest outbreaks may occur simultaneously, if timely targeted treatment cannot be carried out, the economic pine forest may also suffer large-scale grazing, resulting in significant economic losses.

[0006] To this end, the present invention provides a pest identification system and method based on data classification. Summary of the Invention

[0007] (1) Technical problems to be solved

[0008] In view of the deficiencies of the prior art, the present invention provides a pest identification system and method based on data classification. By using a trained pest classification model for identification and classification, a pest identification data set is constructed. If the identified pest damage degree exceeds the expectation, pest outbreak data is collected, and a trained pest outbreak factor identification model is used to analyze the causes of pest outbreaks in the pine forest area to obtain outbreak cause data. A pest value is generated from the pest outbreak data. When the pest value exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, a planting treatment plan is output by the planting treatment plan library; when the adjustment of the pine forest planting state in the pine forest area fails to achieve the expected effect, a corresponding pest treatment plan is output by the pest emergency treatment knowledge graph. By constructing a multi-level alarm mechanism and a multi-level emergency treatment mechanism, when a pest outbreak already exists, pest treatment is implemented at different levels to reduce the impact of pests, thus solving the problems in the background technology.

[0009] (2) Technical solutions

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0011] A pest identification method based on data classification includes, when the deviation between the collected pine forest growth state data and the target value exceeds the expectation, collecting sound and image data of pests in the pine forest area;

[0012] Using the image and sound characteristics as inputs, using a trained pest classification model for identification and classification, and constructing a pest identification data set;

[0013] If the identified pest damage degree exceeds the expectation, collecting pest outbreak data, and using a trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area to obtain outbreak cause data;

[0014] Generating a pest value from the pest outbreak data. When the pest value exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, a planting treatment plan is output by the planting treatment plan library;

[0015] When the adjustment of the pine forest planting state in the pine forest area fails to achieve the expected effect, according to the pest outbreak characteristics, a corresponding pest treatment plan is output by the pest emergency treatment knowledge graph.

[0016] Furthermore, collect and record the stand density, stand aspect, and connectivity of landscape patches in the pine forest area, and summarize them to generate a pine forest growth state data set;

[0017] Generate a state coefficient from the pine forest growth state data set. If the state coefficient exceeds the fluctuation threshold, set a number of data collection points evenly within the pine forest area, collect pest images at the data collection points, record the sounds during the pests' activities, and generate a pest infestation audio-visual data set after summarization.

[0018] Furthermore, label the sound and image data in the pest infestation audio-visual data set, including pest species and sound characteristics. After preprocessing the pest infestation image and sound data, extract features from the image and sound data respectively.

[0019] Concatenate or fuse the feature vectors of the image and sound to form a comprehensive feature vector, train a multi-modal neural network with the comprehensive feature vector, and obtain a trained pest infestation classification model after testing and verification.

[0020] Furthermore, preprocess the collected pest infestation images and sound data, and obtain corresponding image features and sound features through feature extraction; use the trained pest infestation classification model to identify and classify with the features of the image and sound as inputs, obtain corresponding identification and classification results, and generate a pest infestation identification data set after summarization.

[0021] Furthermore, after obtaining the pest infestation types and quantities in the pine forest area, use the trained pest infestation hazard assessment model to perform a hazard score with them as inputs, and obtain the corresponding hazard score.

[0022] Summarize the hazard scores of several different pest infestations obtained continuously. If the sum of the hazard scores exceeds the preset hazard threshold, send an inducement analysis instruction to the outside.

[0023] Furthermore, the sensor network in the pine forest area collects environmental data in real time, correlates the planting state data and environmental condition data with the pest infestation identification data, and generates a pest infestation outbreak data set after summarization.

[0024] After receiving the inducement analysis instruction, use the trained pest infestation outbreak factor identification model to analyze the inducement of the pest infestation outbreak in the pine forest area with the data in the pest infestation outbreak data set as inputs, identify and obtain the outbreak inducements that may cause the pest infestation outbreak, and generate an inducement data set after summarization.

[0025] Furthermore, after identifying the pest infestation in the pine forest area, according to the distribution state of the pest infestation, the trained classifier divides the pest infestation into several settlements, and determines the hazard score within each settlement and the location information between each settlement.

[0026] Construct a pest infestation value based on the hazard score and location information of the pest infestation settlements. If the pest infestation value exceeds the pest infestation threshold, send a first-level alarm instruction to the outside.

[0027] Further, by pre-collecting or formulating several planting treatment plans and aggregating them, a planting treatment plan library is generated;

[0028] After receiving a first-level alarm instruction, after extracting the characteristics of the outbreak incentive data in the pine forest area, the corresponding incentive characteristics are obtained; according to the correspondence between the incentive characteristics and the planting treatment plan, the corresponding planting treatment plan is output from the planting treatment plan library.

[0029] Further, execute the planting treatment plan to adjust the planting status in the pine forest area. After the end of the pre-set observation period, continuously collect and identify pest data through the sensor network in the pine forest area;

[0030] If the number of times of receiving the first-level alarm instruction continuously exceeds the expectation, the time node when receiving an alarm instruction once is used as the alarm node, and the alarm node distribution and the proportion of each pest value exceeding the pest threshold are used as the warning proportion.

[0031] Further, a warning value Jtp is constructed based on the distribution of the alarm nodes and the warning proportion. If the obtained warning value Jtp exceeds the warning threshold, a second-level alarm instruction is sent to the outside. The method for constructing the warning value Jtp is as follows:

[0032]

[0033] In the formula: |t1 - t0| is the receiving interval of the first-level alarm instruction, m is the number of alarm nodes, x i is the warning proportion of the environmental conditions at the i-th alarm node; e can take the value of 2.718; weight coefficients: 0 ≤ α ≤ 1, 0 ≤ β ≤ 1, and α + β = 1.

[0034] Further, taking pest outbreak emergency treatment as the target word, a pest emergency treatment knowledge graph is pre-constructed; after receiving the second-level alarm instruction, the characteristics of each data in the pest outbreak data set are extracted to obtain the corresponding pest outbreak characteristics; the corresponding pest treatment plan is output from the pest emergency treatment knowledge graph.

[0035] A pest identification system based on data classification includes a data collection unit that collects pest sound and image data in the pine forest area when the deviation between the collected pine forest growth status data and the target value exceeds the expectation;

[0036] A pest classification and identification unit takes image and sound characteristics as inputs, uses a trained pest classification model for identification and classification, and constructs a pest identification data set;

[0037] The hazard analysis unit, if the identified pest hazard level exceeds the expectation, collects pest outbreak data, uses the trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area, and obtains outbreak cause data;

[0038] The pest warning unit generates a pest value from the pest outbreak data. When the pest value exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, the planting treatment plan is output from the planting treatment plan library;

[0039] The emergency treatment unit, when the adjustment of the pine forest planting status in the pine forest area fails to achieve the expected effect, according to the pest outbreak characteristics, outputs the corresponding pest treatment plan from the pest emergency treatment knowledge graph.

[0040] (III) Beneficial effects

[0041] The present invention provides a pest identification system and method based on data classification, having the following beneficial effects:

[0042] 1. By combining sound features and image features, when identifying and classifying pests, the accuracy is higher and the confidence level is higher compared to single sound recognition or image recognition and classification.

[0043] 2. Identify the hazards in the pine forest area, determine the hazard level of the current type of pests, and can judge whether pest treatment is needed. Timely pest treatment can prevent further pest outbreaks; through data analysis and identification, determine the causes leading to pest outbreaks. In the scenario where pest outbreaks already exist, conduct targeted treatment of the current pests according to the pest outbreak causes, and play a role in delaying pest outbreaks and damage.

[0044] 3. Construct a pest value Cop based on the current pest outbreak status. According to the pest value Cop, the current pest outbreak level can be quantified and evaluated. If the severity exceeds the expectation, starting from this point, timely treatment can be carried out to achieve pest treatment.

[0045] 4. Output the planting treatment plan from the planting treatment plan library, adjust the current planting status of the pine forest according to the planting treatment plan, conduct preliminary treatment of pests, and reduce the risk of further pest outbreaks on the basis of identifying and classifying pests in the pine forest area.

[0046] 5. Monitor and observe the previous treatment measures to construct a warning value, construct a secondary alarm based on the warning value, and establish a multi-level alarm mechanism. If the previous measures fail to achieve the expected effect, further treatment can be carried out in a timely manner to avoid the impact of multi-type pest outbreaks on the growth of the pine forest group.

[0047] 6. Output the corresponding pest control plan from the pest emergency treatment knowledge graph and directly deal with the pests, such as spraying pesticides, introducing natural enemies of pests, or dealing with the environmental conditions in the pine forest area, etc.; further deal with the pests at different levels to avoid the further deterioration of the pest outbreak situation; construct a multi-level alarm mechanism and a multi-level emergency treatment mechanism by judging the severity of the pest outbreak at multiple levels, so as to achieve pest control at different levels when there is already a pest outbreak and reduce the impact of pests. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Schematic flow chart of the pest identification method based on data classification according to the present invention;

[0049] Figure 2 Schematic structural diagram of the pest identification system based on data classification according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] Please refer to Figure 1 , the present invention provides a pest identification method based on data classification, including,

[0052] Step 1. When the deviation between the collected pine forest growth state data and the target value exceeds the expectation, collect the sound and image data of pests in the pine forest area;

[0053] The content of the above step 1 is as follows:

[0054] Step 101. After determining the pine forest area of the pine forest, collect and record the stand canopy density Yp, stand slope aspect Yb, connectivity Yo of landscape patches and other pine forest growth state data in the pine forest area through a collection device carried by a drone, such as an image collection device and a surveying and mapping device, etc., and generate a pine forest growth state data set after summarization;

[0055] Under dimensionless conditions, generate the state coefficient Btp from the pine forest growth state data set in the following manner:

[0056]

[0057] Weight coefficient: 0 ≤ F1 ≤ 1, 0 ≤ F2 ≤ 1 and F2 + F1 = 1; the weight coefficient is obtained by referring to the analytic hierarchy process;

[0058] Set a fluctuation threshold in advance based on historical data and expectations for pine forest planting management;

[0059] If the state coefficient Btp exceeds the fluctuation threshold, it indicates that there may be a certain deviation between the current growth state of the pine forest and the expected value, and the gap from the ideal conditions exceeds the expectation. In this case, a pest outbreak may be triggered or pests already exist, affecting the further growth of the pine forest. At this time, a data collection instruction is sent to the outside;

[0060] When in use, based on the constructed state coefficient Btp, the current growth state of the pine forest area can be judged. When there are abnormalities in the current growth or development state of the pine forest, there may be a certain risk of pest outbreak in the pine forest area, or pests have already occurred. Starting from this, pests can be monitored in advance to ensure the growth state of the pine forest;

[0061] Step 102: After receiving the data collection instruction, evenly set several data collection points in the pine forest area, and use mobile devices, drones or fixed cameras at the data collection points to collect pest images, such as images of adult and larval pests in different environments; use high-sensitivity microphones or acoustic sensors to record the sounds of pest activities in different environments, such as the chirping of locusts and the crawling sounds of aphids, and generate a pest audio-visual data set after summarization;

[0062] When in use, combine the contents in Steps 101 and 102:

[0063] By collecting pest image and sound data in the pine forest area, when using them for pest identification and classification, the classification results are more accurate and reliable.

[0064] Before controlling pests, it is necessary to identify and classify various types of pests. The existing identification methods usually collect image data of insects through image acquisition devices, and then determine information such as the types and quantities of insects through image recognition. However, in fact, there may also be certain similarities in the morphologies of different types of insects. Especially when the current planting state of the pine forest meets the conditions for pest outbreaks, the number of insects with similar species and morphologies may increase significantly at the same time, resulting in a relatively high risk of misidentifying and classifying pests. Furthermore, in the case where multiple types of pest outbreaks may occur simultaneously, if targeted treatment cannot be carried out in a timely manner, the economic pine forest may also suffer large-scale damage due to being eaten, resulting in relatively large economic losses.

[0065] Step Two: Use the trained pest classification model to identify and classify with image and sound features as inputs, and construct a pest identification data set;

[0066] The above Step Two includes the following contents:

[0067] Step 201: Label the sound and image data in the pest audio-visual data set, including pest species and sound characteristics, such as appearance characteristics, etc.; after preprocessing the image and sound data of the pest damage, extract features from the image and sound data respectively, such as Mel Frequency Cepstral Coefficients and spectrograms, etc.;

[0068] Concatenate or fuse the feature vectors of the image and sound to form a comprehensive feature vector, train a multi-modal neural network with the comprehensive feature vector, and obtain a trained pest classification model after testing and verification;

[0069] Step 202: Preprocess the collected pest image and sound data, and obtain the corresponding image features and sound features through feature extraction;

[0070] Using the features of the image and sound as inputs, use the trained pest classification model for recognition and classification, obtain the corresponding recognition and classification results, including information such as pest species and confidence, etc., and generate a pest recognition data set after summarization;

[0071] When in use, combine the contents in Steps 201 and 202:

[0072] Based on the labeling of the sound data and image data, obtain a trained pest classification model. By combining the sound features and image features, the accuracy rate is higher and the confidence is higher when identifying and classifying pests compared to single sound recognition or image recognition and classification.

[0073] Step 3: If the identified pest damage level exceeds the expectation, collect pest outbreak data, and use the trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area to obtain outbreak cause data;

[0074] The said Step 3 includes the following contents:

[0075] Step 301: Train a neural convolutional network with the labeled sample data to obtain a trained pest damage evaluation model;

[0076] After obtaining the pest types and quantities in the pine forest area, use the trained pest damage evaluation model to score the damage with them as inputs, and obtain the corresponding damage score; summarize the damage scores of several different pests continuously obtained. If the sum of the damage scores exceeds the preset damage threshold, send an instruction for cause analysis to the outside;

[0077] When in use, after identifying the pest category information in the pine forest area, identify the damage score in the pine forest area based on historical data and labeled data, determine the damage level of the current type of pest, and can judge whether it is necessary to deal with the pest. Deal with the pest in time to prevent further pest outbreaks;

[0078] Step 302: Arrange a sensor network within the pine forest area. The sensor network collects environmental data in real time, such as environmental condition data like temperature and humidity. Associate the planting status data, environmental condition data, and pest identification data, and after summarization, establish a pest outbreak data set.

[0079] Step 303: Train a machine learning algorithm with the labeled sample data to obtain a trained pest outbreak factor identification model. After receiving an inducement analysis instruction, use the data within the pest outbreak data set as input, and use the trained pest outbreak factor identification model to conduct an inducement analysis of the pest outbreak in the pine forest area, identify and obtain the outbreak inducements that may lead to a pest outbreak, and generate an inducement data set after summarization.

[0080] When in use, combine the content in Steps 301 to 303:

[0081] When there is already a trend of pest outbreak in the pine forest area, collect various data within the pine forest area. Through data analysis and identification, determine the inducements leading to the pest outbreak. In the scenario where there is already a pest outbreak, conduct targeted treatment of the current pest according to the pest outbreak inducements, which can play a role in delaying the pest outbreak and damage.

[0082] Step Four: Generate a pest value Cop from the pest outbreak data. When the pest value Cop exceeds the pest threshold, output a planting treatment plan from the planting treatment plan library according to the correspondence between the inducement characteristics and the planting treatment plan.

[0083] The above Step Four includes the following content:

[0084] Step 401: After identifying the pests in the pine forest area, according to the distribution state of the pests, use the trained classifier to divide the pest area into several settlements, and determine the harm score within each settlement and the position information between each settlement.

[0085] Under dimensionless conditions, construct a pest value Cop based on the harm score and position information of the pest settlements in the following way:

[0086]

[0087] In the formula: N is the number of pest settlements, X i is the position of the i-th pest settlement, X0 is the central position, k is a parameter controlling distance attenuation, with a value within [0, 1], W(X i ) is the weight coefficient of the i-th pest settlement, with a value within [0, 1]; S(X i ) is the harm score of the i-th pest settlement; weight coefficient: 0 ≤ K1 ≤ 1, 0 ≤ K2 ≤ 1, and K1 + K2 = 1;

[0088] According to historical data and management expectations for pests, a pest threshold is set in advance. If the pest value Cop exceeds the pest threshold, it indicates that the pest situation in the pine forest area may be relatively serious and may be in an outbreak state. At this time, a first-level alarm instruction is sent to the outside.

[0089] When in use, when the pest has broken out, a pest value Cop is constructed based on the current pest outbreak state. Based on the pest value Cop, the current pest outbreak degree can be quantified and evaluated. If the severity exceeds the expectation, starting from this point, by dealing with it in a timely manner, the pests can be treated.

[0090] Step 402: Collect or formulate several planting treatment plans in advance, such as irrigation, fertilization, or pest control, etc. After summarization, a planting treatment plan library is generated. After receiving the first-level alarm instruction, after extracting the characteristics of the outbreak inducement data in the pine forest area, the corresponding inducement characteristics are obtained. According to the correspondence between the inducement characteristics and the planting treatment plans, the corresponding planting treatment plan is output from the planting treatment plan library.

[0091] When in use, combine the content in Steps 401 to 402:

[0092] After identifying and obtaining the cause of the current pest outbreak, the planting treatment plan is output from the planting treatment plan library. According to the planting treatment plan, the current planting state of the pine forest is adjusted to initially treat the pests, and based on the identification and classification of the pests in the pine forest area, the risk of further pest outbreaks is reduced.

[0093] Step Five: When the adjustment of the planting state of the pine forest in the pine forest area fails to achieve the expected effect, according to the pest outbreak characteristics, the corresponding pest treatment plan is output from the pest emergency treatment knowledge graph;

[0094] The said Step Five includes the following content:

[0095] Step 501: Execute the planting treatment plan to adjust the planting state in the pine forest area. After the end of the pre-set observation period, continuously collect and identify pest data through the sensor network in the pine forest area. If the number of continuously received first-level alarm instructions exceeds the expectation, take the time node when a single alarm instruction is received as the alarm node, and take the alarm node distribution and the proportion of each pest value exceeding the pest threshold as the warning proportion;

[0096] Step 502: Under dimensionless conditions, construct a warning value Jtp according to the distribution of the alarm nodes and the warning proportion, and the method is as follows:

[0097]

[0098] Where: |t1 - t0| is the receiving interval of the first-level alarm instruction, m is the number of alarm nodes, and x i is the warning ratio of the environmental conditions at the i-th alarm node; e can take the value of 2.718; weight coefficients: 0 ≤ α ≤ 1, 0 ≤ β ≤ 1, and α + β = 1;

[0099] According to historical data and the management expectation of pests, the warning threshold is set in advance; if the obtained warning value Jtp exceeds the warning threshold, it indicates that the pest outbreak problem in the pine forest area has not been solved. At this time, a second-level alarm instruction is sent to the outside;

[0100] When in use, on the basis of having received the first-level alarm instruction and having made corresponding treatments, by monitoring and observing the previous treatment measures to construct a warning value, and constructing a second-level alarm based on the warning value, so as to build a multi-level alarm mechanism. If the previous measures fail to achieve the expected effect, further treatments can be made in a timely manner to avoid the impact of multi-type pest outbreaks on the growth of the pine forest group;

[0101] Step 503: Using the emergency treatment of pest outbreaks as the target word, after in-depth retrieval and entity relationship construction, a knowledge graph of pest emergency treatment is constructed in advance;

[0102] After receiving the second-level alarm instruction, extract the features of each data in the pest outbreak data set to obtain the corresponding pest outbreak features; according to the correspondence between the pest outbreak features and the pest treatment plan, the corresponding pest treatment plan is output by the knowledge graph of pest emergency treatment;

[0103] When in use, combine the content in Steps 501 to 503:

[0104] The corresponding pest treatment plan is output by the knowledge graph of pest emergency treatment to directly treat the pests, such as spraying pesticides, introducing natural enemies of pests, or treating the environmental conditions in the pine forest area, etc.; by further treating the pests at different levels, the further deterioration of the pest outbreak situation is avoided;

[0105] On the basis of accurately identifying the pest situation in the pine forest area, through multi-level judgment of the severity of the pest outbreak, a multi-level alarm mechanism and a multi-level emergency treatment mechanism are constructed, so as to realize pest treatment at different levels when there is already a pest outbreak, and reduce the impact of pests.

[0106] Please refer to Figure 2 , the present invention provides a pest identification system based on data classification, including,

[0107] A data collection unit, when the deviation between the collected pine forest growth state data and the target value exceeds the expectation, collects the sound and image data of pests in the pine forest area;

[0108] The pest classification and identification unit takes image and sound features as inputs, uses a trained pest classification model for identification and classification, and constructs a pest identification dataset;

[0109] The hazard analysis unit, if the identified pest hazard level exceeds the expectation, collects pest outbreak data, uses a trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area, and obtains outbreak cause data;

[0110] The pest warning unit generates a pest value Cop from the pest outbreak data. When the pest value Cop exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, the planting treatment plan is output from the planting treatment plan library;

[0111] The emergency treatment unit, when the adjustment of the pine forest planting status in the pine forest area fails to achieve the expected effect, outputs the corresponding pest treatment plan according to the pest outbreak characteristics from the pest emergency treatment knowledge graph.

[0112] The Analytic Hierarchy Process (AHP) is a decision-making method that decomposes the elements related to the decision into levels such as goals, criteria, and solutions, and conducts qualitative and quantitative analysis on this basis. It is particularly suitable for dealing with target systems with hierarchical and interleaved evaluation indicators, and when the target value is difficult to quantitatively describe, the Analytic Hierarchy Process is an effective decision-making tool.

[0113] The core of the Analytic Hierarchy Process lies in decomposing the decision problem into multiple levels to form a hierarchical structure, which usually includes a goal layer, a criterion layer, a sub-criterion layer, and a solution layer. By solving the eigenvector of the judgment matrix, the priority weights of each element at each level with respect to an element at the previous level are obtained, and finally, the weighted sum method is used to hierarchically merge the final weights of each alternative solution with respect to the overall goal, so as to find the optimal solution.

[0114] The method for constructing the pest emergency treatment knowledge graph can refer to the following content:

[0115] Collect and organize data. Data sources: Data related to pest emergency treatment can be collected from multiple channels, including agricultural expert knowledge, literature data, field research data, etc. Data organization: Organize the collected data, including removing duplicate data, correcting incorrect data, and unifying the data format, etc.

[0116] Knowledge modeling, defining the domain knowledge system: According to the characteristics and requirements of pest emergency treatment, define the domain knowledge system, including the types of pests, symptoms, transmission routes, prevention and control methods, etc. Determine the relationship types: Clearly define the relationship types between different knowledge points, such as "symptom - pest", "prevention and control method - pest", etc.

[0117] Knowledge extraction, semi-structured data extraction: For data that already has a certain structure (such as tables, databases, etc.), extraction can be carried out through rules. Unstructured data extraction: For unstructured data such as text and images, natural language processing (NLP) and image recognition technologies can be used for extraction. For example, use NLP technology to automatically obtain knowledge about pests and their control methods from literature and expert data.

[0118] Knowledge fusion, entity recognition and linking: Identify the same entities in different data sources and link them to eliminate redundancy and conflicts. Relationship integration: Integrate the relationships between the same entities in different data sources to form a complete knowledge network.

[0119] Knowledge storage, select a graph database: Select a graph database suitable for storing knowledge graphs, such as Neo4j, etc. Data import and storage: Import the extracted and fused knowledge into the graph database for storage.

[0120] Knowledge reasoning and update, knowledge reasoning: Based on existing knowledge, infer unknown knowledge, such as inferring possible pest types according to the symptoms of pests. Knowledge update: As new data is continuously added and technology develops, it is necessary to regularly update and maintain the knowledge graph.

[0121] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0122] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here.

[0123] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0124] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0125] As described above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A pest identification method based on data classification, characterized in that: Including, When the deviation between the collected pine forest growth status data and the target value exceeds the expectation, collect the sound and image data of pests in the pine forest area; use the trained pest classification model to identify and classify with the image and sound features as the input, and construct a pest identification data set; If the identified pest damage degree exceeds the expectation, collect pest outbreak data, and use the trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area to obtain outbreak cause data, including: after obtaining the pest types and quantities in the pine forest area, use them as the input, and use the trained pest damage evaluation model to perform a damage score to obtain the corresponding damage score; sum up the damage scores of several different pests continuously obtained. If the sum of the damage scores exceeds the preset damage threshold, send an instruction for cause analysis to the outside; the sensor network in the pine forest area collects environmental data in real time, associates the planting status data and environmental condition data with the pest identification data, and summarizes them to establish a pest outbreak data set; after receiving the instruction for cause analysis, use the data in the pest outbreak data set as the input, and use the trained pest outbreak factor identification model to analyze the causes of pest outbreaks in the pine forest area, identify and obtain the outbreak causes that may lead to pest outbreaks, and summarize them to generate a cause data set; Generate a pest value from the pest outbreak data. When the pest value exceeds the pest threshold, according to the correspondence between the cause characteristics and the planting treatment plan, output the planting treatment plan from the planting treatment plan library, including: after identifying the pests in the pine forest area, according to the distribution state of the pests, use the trained classifier to divide the pest area into several settlements, and determine the damage score within each settlement and the position information between each settlement; construct a pest value based on the damage score and position information of the pest settlement. If the pest value exceeds the pest threshold, send a first-level alarm instruction to the outside; If the number of continuously received first-level alarm instructions exceeds the expectation, use the time node when a single alarm instruction is received as the alarm node, and use the alarm node distribution and the proportion of each pest value exceeding the pest threshold as the warning proportion; Construct the warning value based on the distribution of alarm nodes and the warning ratio , if the obtained warning value exceeds the warning threshold, send a secondary alarm instruction to the outside, and the method of constructing the warning value is as follows: ; where: is the receiving interval of the first-level alarm instruction, is the number of alarm nodes, is the warning ratio on the i-th alarm node; takes the value of 2.718; weight coefficient: ， , and ; When the adjustment of the pine forest planting status in the pine forest area fails to achieve the expected effect, according to the pest outbreak characteristics, output the corresponding pest treatment plan from the pest emergency treatment knowledge graph, including: pre-construct the pest emergency treatment knowledge graph with pest outbreak emergency treatment as the target word; after receiving the second-level alarm instruction, extract the characteristics of each data in the pest outbreak data set to obtain the corresponding pest outbreak characteristics; output the corresponding pest treatment plan from the pest emergency treatment knowledge graph.

2. The pest identification method based on data classification according to claim 1, characterized in that: Collect and record the stand density, stand slope aspect and connectivity of landscape patches in the pine forest area, and summarize them to generate a pine forest growth status data set; Generate a status coefficient from the pine forest growth status data set. If the status coefficient exceeds the fluctuation threshold, evenly set several data collection points in the pine forest area, collect pest images at the data collection points, and record the sounds when the pests are active and summarize them to generate a pest audio-visual data set.

3. The pest identification method based on data classification according to claim 2, characterized in that: Label the sound and image data in the pest infestation audio-visual data set, including pest species and sound characteristics. After preprocessing the pest infestation image and sound data, extract features from the image and sound data respectively; Concatenate or fuse the feature vectors of the image and sound to form a comprehensive feature vector, train a multi-modal neural network with the comprehensive feature vector, and obtain a trained pest infestation classification model after passing the test and verification.

4. The pest identification method based on data classification according to claim 3, characterized in that: Preprocess the collected pest infestation image and sound data, and obtain the corresponding image features and sound features through feature extraction; Use the trained pest infestation classification model to identify and classify with the features of the image and sound as inputs, obtain the corresponding identification and classification results, and generate a pest infestation identification data set after summarization.

5. The pest identification method based on data classification according to claim 4, characterized in that: Collect or formulate several planting treatment plans in advance, and generate a planting treatment plan library after summarization; After receiving a first-level alarm instruction, extract features from the outbreak cause data in the pine forest area to obtain the corresponding cause features; according to the correspondence between the cause features and the planting treatment plans, output the corresponding planting treatment plan from the planting treatment plan library.

6. The pest identification method based on data classification according to claim 5, characterized in that: Execute the planting treatment plan to adjust the planting status in the pine forest area. After the preset observation period ends, continuously collect and identify pest infestation data through the sensor network in the pine forest area.

7. A pest identification system based on data classification, used to implement the pest identification method based on data classification according to claim 1, characterized in that: A data collection unit, when the deviation between the collected pine forest growth status data and the target value exceeds the expectation, collect the sound and image data of pest infestation in the pine forest area; A pest infestation classification and identification unit, using the trained pest infestation classification model to identify and classify with the image and sound characteristics as inputs, and construct a pest infestation identification data set; A hazard analysis unit, if the identified pest infestation hazard level exceeds the expectation, collect pest infestation outbreak data, and use the trained pest infestation outbreak factor identification model to analyze the cause of the pest infestation outbreak in the pine forest area to obtain the outbreak cause data; A pest infestation warning unit, generate a pest infestation value from the pest infestation outbreak data, and when the pest infestation value exceeds the pest infestation threshold, output the planting treatment plan from the planting treatment plan library according to the correspondence between the cause features and the planting treatment plans; An emergency treatment unit, when the adjustment of the pine forest planting status in the pine forest area fails to achieve the expected effect, output the corresponding pest infestation treatment plan from the pest infestation emergency treatment knowledge graph according to the pest infestation outbreak characteristics.

Images