Pest and disease monitoring method and system based on image processing

Abstract

The invention discloses a disease and pest monitoring method and system based on image processing, and the method comprises the following steps: firstly determining the types of regional diseases and pests according to a regional disease and pest biological sample database; determining growth and development cycles of various diseases and pests according to the disease and pest species in combination with a disease and pest biological sample database; and constructing a regional pest and disease damage growth simulation model, and performing fitting normalization processing on the food intake change curve of various types of pests and diseases in the region to form a regional plant silkworm feeding invasion degree value change curve. According to the method, a regional biological sample database is used as a sample set, screening and analysis are carried out on the sample set, a regional plant silkworm food change curve is constructed, regional plant images are obtained through aerial photography of an unmanned aerial vehicle to determine the silkworm food invasion degree, and the variety number and the development situation of regional diseases and pests can be preliminarily determined; and reliable data support can be provided for subsequent pest and disease control.

Description

Technical Field

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

[0002] In the vast natural world, pests and diseases are like an invisible storm, which often cause serious damage to crops and forests. With the sustained and stable development of agriculture, various problems encountered in agricultural planting and development have received more and more widespread attention. Pest and disease disaster prevention and control has always been the most important existence in agricultural planting; and the damage of forest pests and diseases to the ecological environment is even more difficult to estimate. The forests protected now have almost no mechanism to resist pests and diseases. Once pests and diseases occur in large numbers, they will be irreparable. To achieve the purpose of disaster prevention, it is crucial to conscientiously monitor and forecast forest pests and diseases. Only by timely and accurately grasping the development trends of forest pests and diseases can we make correct prevention and control decisions in a timely manner and take effective prevention and control measures to achieve the goal. To achieve the purpose of early treatment and small treatment; currently there are many modes for disease and insect pest monitoring, among which front-end monitoring is the most common. For example, the disease and insect pest monitoring and early warning system launched by a Zhejiang Co., Ltd. is equipped with microclimate collection equipment, habitat monitoring equipment, insect information collection equipment, pathogen spore capture and cultivation system and early warning forecast system, expert system, information management platform, etc., to monitor and manage the disease and insect conditions, crop growth, disaster conditions, air temperature, humidity, dew point, soil temperature, light intensity and other important parameters in the process of crop growth at each monitoring point in real time. Although the system can realize real-time remote monitoring and diagnosis, it needs to deploy a large number of front-end monitoring and collection equipment, which affects the entire system for the operating status of outdoor equipment, and the system stability is poor. Summary of the invention

[0003] The purpose of the present invention is to provide a pest monitoring method and system based on image processing in order to solve the above problems.

[0004] The present invention achieves the above-mentioned purpose through the following technical solutions:

[0005] A pest monitoring method based on image processing comprises the following steps:

[0006] S1. First determine the types of regional pests and diseases based on the regional pest and disease biological sample database;

[0007] S2. Determine the growth and development cycle of various types of pests and diseases based on the pest and disease types and the pest and disease biological sample database;

[0008] S3. Analyze the food intake during the development cycle of pests and diseases in combination with the pests and diseases biological sample database to determine the change curve of food intake during the development cycle of pests and diseases;

[0009] S4, constructing a regional pest growth simulation model according to steps S1-S3, fitting and normalizing the change curves of the feeding amount of various types of pests in the region according to the regional pest growth simulation model to form a change curve of the regional plant cannibalism damage degree value, and determining the degree of pest development in the region according to the plant cannibalism damage degree value;

[0010] S5. Obtain high-definition regional plant images through drones, extract damaged edge features in the plant images, mark the damaged edge features, compare the marked damaged edge features with the corresponding plant standard features to analyze the plant damage ratio, and determine the plant cannibalization damage degree value;

[0011] S6. Input the plant erosion damage degree value obtained by the drone into the regional pest growth simulation model to determine the development degree of regional pests and diseases.

[0012] Preferably, in step S4, the hostile relationship between the pests and diseases is determined according to the regional pest and disease biological sample database in step S1, and the hostile relationship between the pests and diseases is combined with the growth and development cycle of the pests and diseases to determine the coexistence time period of the hostile pests and diseases. According to the coexistence time period of the hostile pests and diseases, the plant cannibalism damage degree value change curve is corrected, and the original plant cannibalism damage degree value is F 0 The corrected plant damage value is F 1 , F 1 =F 0 (1-&), where & is the correction factor, and the value of & is between 0.1 and 0.25.

[0013] Preferably, the plant cannibalization damage degree variation curve has the time-related corresponding pest and disease type as the horizontal axis and the plant damage degree value as the vertical axis, wherein the plant damage degree value is based on the estimated damaged area as the benchmark value.

[0014] Preferably, the regional pest biological sample database is a pest statistics database over the years and includes plant species, pest species, pest appearance time, pest disappearance time, hostile relations between pests, and food intake of pests in different time periods.

[0015] Preferably, according to step S5, the drone also synchronously acquires infrared area plant images, extracts biological dynamic features in the infrared area plant images, determines biological species based on the biological dynamic features compared with biological standard features, verifies the degree of development of pests and diseases in step S7, and updates the regional pest and disease sample database.

[0016] A pest monitoring system based on image processing includes a drone, an image acquisition module is mounted on the drone, and an image processing module is built in the drone. The image acquisition module transmits the acquired regional plant image to the image processing module, and the image processing module processes the received image and sends the plant erosion damage degree value to the ground central control terminal. The ground central control terminal includes a regional pest growth simulation module, which inputs the plant erosion damage degree value into a regional pest growth simulation model and outputs the regional pest development degree.

[0017] Preferably, the image processing module includes a feature extraction unit, a feature marking unit, a feature analysis unit and an encroachment degree analysis unit. The feature extraction unit is used to extract damaged edge features of the image, the feature marking unit is used to mark edge points of the damaged edge features, the feature analysis unit is used to compare each marked damaged edge feature with the corresponding plant standard feature to analyze the plant damage ratio, and the encroachment degree analysis unit is used to determine the encroachment degree value according to the plant damage ratio.

[0018] Preferably, the regional pest and disease growth simulation module includes a regional biological type analysis unit, a development cycle analysis unit and an encroachment degree correction unit. The regional biological type analysis unit screens the pests and diseases according to the regional pest and disease biological sample database. The development cycle analysis unit analyzes the pest and disease feeding amount during the development cycle according to the screened pests and diseases in combination with the corresponding growth curve. The encroachment degree correction unit determines the hostile relationship between pest and disease types according to the regional pest and disease biological sample database, determines the coexistence time period of hostile pests and diseases according to the hostile relationship between pest and disease types and the pest and disease growth and development cycle, and corrects the plant encroachment degree value change curve according to the coexistence time period of hostile pests and diseases.

[0019] The beneficial effects of the present invention are:

[0020] The present invention uses a regional pest and disease biological database as a sample set, analyzes the pest and disease types, pest and disease development cycles, and changes in feeding amount during the pest and disease development cycle in the sample set, and establishes a pest and disease growth simulation model. According to the pest and disease growth simulation model, the feeding amount in different time periods can be fitted into a nibbling damage degree change curve, that is, the feeding of various types of pests and diseases in different time periods is positively correlated with the degree of plant damage. Regular patrols by unmanned aerial vehicles are used to obtain the percentage of regional plant damage, and then the development trend of pests and diseases in the region and the number of pest and disease types can be indirectly determined. The overall monitoring is more efficient, and there is no need to arrange a large number of collection equipment at the front end. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0022] Figure 1 It is a work flow chart of the present invention.

[0023] Figure 2 It is the overall system architecture diagram of the present invention. DETAILED DESCRIPTION

[0024] The following is combined with Figure 1-2 , the technical solution of the present invention is further explained:

[0025] Embodiment 1

[0026] like Figure 1 As shown, a pest monitoring method based on image processing includes the following steps:

[0027] S1. According to the regional pest and disease biological sample database, first determine the types of regional pests and diseases; that is, first retrieve the pest and disease data in the regional pest and disease biological sample database as a sample set, and use the decision tree algorithm to screen and determine the types of pests and diseases in the region, such as aphids, spider mites, white grubs, mole crickets, longhorn beetles, etc., wherein the screening determines the damage of different types of pests and diseases to different parts of the plant. For example, aphids, spider mites, etc., they harm plants by sucking plant sap, mainly causing damage to plant leaves; while white grubs, mole crickets, etc., mainly harm the roots of plants; longhorn beetle pests, mainly harm the branches and trunks of plants; usually, the appearance time of pests and diseases that cause damage to different parts of plants will overlap, in order to reduce the overall data processing complexity, so this embodiment selects the leaf damage degree for integrated analysis.

[0028] It should be noted that the regional pest and disease biological sample database is a statistical database of pests and diseases over the years and includes the types of plants planted over the years, the types of pests and diseases, the time when pests and diseases appeared, the time when pests and diseases disappeared, the hostile relationship between pests and diseases, and the amount of food eaten by pests and diseases in different time periods. In other words, based on the pest and disease related elements in the biological sample database, it can help analyze the ability of pests and diseases to harm regional plants over time.

[0029] S2. Determine the growth and development cycle of various pests and diseases based on the pest and disease types and the pest and disease biological sample database. That is, first determine the pest and disease types that infringe on the leaves, and combine the pest and disease appearance and disappearance time nodes in the biological sample database to determine the entire survival cycle of the pest and disease in the natural environment.

[0030] S3. According to the growth and development cycle of pests and diseases combined with the pest and disease biological sample database, the feeding amount during the development cycle is analyzed to determine the feeding amount change curve of the pest and disease development cycle. In other words, according to the life cycle of the corresponding pest and disease combined with the biological sample database, the feeding amount of the pest and disease in different time periods of the life cycle can be determined, that is, the amount of leaf erosion, and the feeding amount change curve of the corresponding pest and disease can be established using MATLAB, which can help the subsequent regional pest and disease analysis.

[0031] S4, construct a regional pest growth simulation model according to steps S1-S3, according to the regional pest growth simulation model, fit and normalize the change curves of the feeding amount of various types of pests and diseases in the region to form a change curve of the degree of plant cannibalization damage, and determine the degree of development of pests and diseases in the region according to the plant cannibalization damage value. In other words, MATLAB software is used to establish a growth analysis model in combination with the types of pests and diseases and the feeding amount at different time periods, and corresponding growth analysis models are constructed for different pests and diseases that infringe on leaves, and corresponding cannibalization damage value change curves are generated, and then each change curve is fitted to generate a regional plant cannibalization damage degree change curve, and then the actual degree of damage in the region is determined by different time periods, which can indirectly determine the development trend of pests and diseases, and at the same time, the development trend and the corresponding types of pests and diseases can be determined, which can help to provide reliable model support for the subsequent formulation of prevention and control plans.

[0032] It should be noted that the plant cannibalization damage degree variation curve is a curve with the time-related corresponding pest and disease type as the horizontal axis and the plant damage degree value as the vertical axis, wherein the plant damage degree value is based on the estimated damaged area as the benchmark value.

[0033] In some embodiments, the hostile relationship between the pests and diseases is determined according to the regional pest and disease biological sample database in step S1, and the hostile relationship between the pests and diseases is combined with the growth and development cycle of the pests and diseases to determine the coexistence time period of the hostile pests and diseases. According to the coexistence time period of the hostile pests and diseases, the plant cannibalism damage degree value change curve is corrected, and the original plant cannibalism damage degree value is F 0 The corrected plant damage value is F 1 , F 1 =F 0 (1-&), where & is the correction coefficient, and & is between 0.1 and 0.25. That is to say, according to the hostile relationship between pests and diseases in the natural environment, the samples of pests and diseases in the region with hostile relationships are determined, and then combined with the overlapping time period of the survival and development cycle of pests and diseases in the biological sample database, that is, the coexistence time period, and then the corresponding plant cannibalization damage degree change curve is adjusted according to the coexistence time period, that is, when the degree of damage to plants by the pests and diseases with hostile relationships is weakened during the coexistence time period, the cannibalization degree value is corrected downward.

[0034] S5. Obtain high-definition regional plant images through drones, extract damaged edge features in plant images, mark edge points of damaged edge features, compare each marked damaged edge feature with the corresponding plant standard feature to analyze the percentage of plant damage, and determine the degree of plant cannibalization damage. In other words, low-altitude aerial photography of the area is performed by mounting an RGB high-definition camera on a DJ-mini drone to obtain high-definition images of regional plants. At the same time, the high-definition images are processed by the drone's built-in image processing module, and the CNN algorithm is used to extract the damaged leaf features of the plants in the image. Then, special edge points are marked on the damaged edge features, and the damaged special edge points are compared with the standard features to determine the percentage of damaged edge area, and then the degree of plant damage, i.e., the degree of cannibalization damage, can be determined.

[0035] In some embodiments, an infrared camera is also mounted on the drone, and the drone also synchronously acquires infrared region plant images, extracts biological dynamic features in the infrared region plant images, determines biological species based on the biological dynamic features compared with biological standard features, verifies the degree of development of pests and diseases in step S7, and updates the regional pest and disease sample database. That is, in this embodiment, the infrared image acquired by the infrared camera can further correct the plant damage degree value, which is more accurate, and at the same time, the biological species can be determined through the acquired infrared image, and the biological species in the biological sample database can be updated.

[0036] S6. Input the plant erosion damage value obtained by the drone into the regional pest growth simulation model to determine the development degree of regional pests. In other words, the plant erosion damage value fed back by the drone to the ground control and the time point are input into the software, which can directly determine the type of pests and diseases and the development trend of pests and diseases corresponding to the erosion damage value. There is no need to deploy a large number of collection equipment at the front end, and the monitoring is more efficient and convenient.

[0037] Embodiment 2

[0038] like Figure 2As shown, a pest monitoring system based on image processing includes a drone, an image acquisition module is mounted on the drone, and an image processing module is built in the drone. The image acquisition module transmits the acquired regional plant image to the image processing module, and the image processing module processes the received image and sends the plant cannibalization damage degree value to the ground central control terminal. The ground central control terminal includes a regional pest growth simulation module, which inputs the plant cannibalization damage degree value into the regional pest growth simulation model and outputs the regional pest development degree. In other words, through the regional pest growth analysis model built by the ground central control terminal, and the drone obtains the regional plant cannibalization damage value in real time, the drone feeds back the regional plant cannibalization damage value to the ground central control terminal, and the ground central control terminal can determine the current pest development trend and the approximate types and groups of pests and diseases based on the received regional plant cannibalization damage value and time node.

[0039] Specifically, the image processing module includes a feature extraction unit, a feature marking unit, a feature analysis unit and an encroachment degree analysis unit. The feature extraction unit is used to extract the damaged edge features of the image, the feature marking unit is used to mark the edge points of the damaged edge features, the feature analysis unit is used to compare the damaged edge features after each mark with the corresponding plant standard features to analyze the plant damage ratio, and the encroachment degree analysis unit is used to determine the encroachment degree value according to the plant damage ratio. That is to say, the high-definition regional plant image is obtained by the drone, the damaged edge features in the plant image are extracted, and the damaged edge features are marked with edge points. The damaged edge features are compared with the corresponding plant standard features to analyze the plant damage ratio, and the plant encroachment degree value is determined. That is to say, the DJ-mini drone is mounted with an RGB high-definition camera to take low-altitude aerial photos of the area to obtain high-definition images of regional plants. At the same time, the high-definition images are processed by the built-in image processing module of the drone, and the damaged leaf features of the plants in the image are extracted using the CNN algorithm. Then, the damaged edge features are marked with special edge points, and the damaged edge area ratio is determined by comparing the damaged special edge points with the standard features, so as to determine the plant damage degree, that is, the encroachment degree value.

[0040] Specifically, the regional pest and disease growth simulation module includes a regional biological type analysis unit, a development cycle analysis unit and an encroachment degree correction unit. The regional biological type analysis unit screens the pests and diseases according to the regional pest and disease biological sample database. The development cycle analysis unit analyzes the feeding amount of the pests and diseases during the development cycle according to the screened pests and diseases in combination with the corresponding growth curve; the encroachment degree correction unit determines the hostile relationship between the pest and disease types according to the regional pest and disease biological sample database, determines the coexistence time period of the hostile pests and diseases according to the hostile relationship between the pest and disease types and the pest and disease growth and development cycle, and corrects the plant encroachment degree value change curve according to the coexistence time period of the hostile pests and diseases. In other words, MATLAB software is used to establish a growth analysis model in combination with the types of pests and diseases and the feeding amount in different time periods. Corresponding growth analysis models are constructed for different pests and diseases that infringe on leaves, and corresponding curves of changes in the degree of erosion and damage are generated. Subsequently, each change curve is fitted to generate a curve of changes in the degree of erosion and damage of regional plants. The actual degree of damage to the region can be determined through different time periods, which can indirectly determine the development trend of pests and diseases. At the same time, the development trend and the corresponding types of pests and diseases can be determined, which can provide reliable model support for the subsequent formulation of prevention and control plans.

[0041] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A pest monitoring method based on image processing, characterized in that: The following steps are involved: S1. First determine the types of regional pests and diseases based on the regional pest and disease biological sample database; S2. Determine the growth and development cycle of various types of pests and diseases based on the pest and disease types and the pest and disease biological sample database; S3. Analyze the food intake during the development cycle of pests and diseases in combination with the pests and diseases biological sample database to determine the change curve of food intake during the development cycle of pests and diseases; S4, constructing a regional pest growth simulation model according to steps S1-S3, fitting and normalizing the change curves of the feeding amount of various types of pests in the region according to the regional pest growth simulation model to form a change curve of the regional plant cannibalism damage degree value, and determining the degree of pest development in the region according to the plant cannibalism damage degree value; S5. Obtain high-definition regional plant images through drones, extract damaged edge features in the plant images, mark the damaged edge features, compare the marked damaged edge features with the corresponding plant standard features to analyze the plant damage ratio, and determine the plant cannibalization damage degree value; S6. Input the plant erosion damage degree value obtained by the drone into the regional pest growth simulation model to determine the development degree of regional pests and diseases.

2. The pest monitoring method based on image processing as claimed in claim 1, characterized in that: In the step S4, the hostile relationship between the types of pests and diseases is determined according to the regional pest and disease biological sample database of step S1, and the coexistence time period of the hostile pests and diseases is determined according to the hostile relationship between the types of pests and diseases and the growth and development cycle of the pests and diseases. According to the coexistence time period of the hostile pests and diseases, the plant cannibalism damage degree value change curve is corrected, the original plant cannibalism damage degree value is F0, and the corrected plant damage degree value is F1, F1=F0(1-&), wherein & is the correction coefficient, and & takes a value between 0.1-0.

25.

3. The pest monitoring method based on image processing according to claim 1 or 2, characterized in that: The plant cannibalism damage degree variation curve has the time-related corresponding pest and disease type as the horizontal axis and the plant damage degree value as the vertical axis, wherein the plant damage degree value is based on the estimated damaged area as the benchmark value.

4. The pest monitoring method based on image processing as claimed in claim 1, characterized in that: The regional pest and disease biological sample database is a pest and disease statistical database over the years and includes plant species, pest and disease types, pest and disease appearance time, pest and disease disappearance time, hostile relationships between pests and diseases, and the amount of food consumed by pests and diseases in different time periods.

5. The pest monitoring method based on image processing as claimed in claim 1, characterized in that: According to step S5, the drone also synchronously acquires infrared area plant images, extracts biological dynamic features in the infrared area plant images, determines biological species based on the biological dynamic features compared with biological standard features, verifies the degree of development of pests and diseases in step S7, and updates the regional pest and disease sample database.

6. A pest monitoring system based on image processing, comprising a drone, characterized in that: The drone is mounted with an image acquisition module, and the drone is built with an image processing module. The image acquisition module transmits the acquired regional plant images to the image processing module, and the image processing module processes the received images and sends the plant encroachment degree value to the ground central control terminal. The ground central control terminal includes a regional pest and disease growth simulation module, which inputs the plant encroachment degree value into the regional pest and disease growth simulation model and outputs the regional pest and disease development degree.

7. The pest and disease monitoring system based on image processing as described in claim 6, wherein the image processing module includes a feature extraction unit, a feature marking unit, a feature analysis unit and an encroachment degree analysis unit, the feature extraction unit is used to extract damaged edge features of the image, the feature marking unit is used to mark edge points of the damaged edge features, the feature analysis unit is used to compare each marked damaged edge feature with the corresponding plant standard feature to analyze the plant damage ratio, and the encroachment degree analysis unit is used to determine the encroachment degree value according to the plant damage ratio.

8. The pest monitoring system based on image processing as claimed in claim 6, characterized in that: The regional pest and disease growth simulation module includes a regional biological type analysis unit, a development cycle analysis unit and an encroachment degree correction unit. The regional biological type analysis unit screens the pests and diseases according to the regional pest and disease biological sample database. The development cycle analysis unit analyzes the feeding amount of the pests and diseases during the development cycle according to the screened pests and diseases in combination with the corresponding growth curve; the encroachment degree correction unit determines the hostile relationship between the pest and disease types according to the regional pest and disease biological sample database, determines the coexistence time period of the hostile pests and diseases according to the hostile relationship between the pest and disease types and the pest and disease growth and development cycle, and corrects the plant encroachment degree value change curve according to the coexistence time period of the hostile pests and diseases.

Images