A system and method for monitoring crop diseases and insect pests
By combining the technical means of convolutional layer and dynamic weight allocation unit in the crop pest monitoring system, the problems of high difficulty, high cost and poor accuracy in the existing technology are solved, and efficient and accurate pest monitoring are achieved. and prevention.
Patent Information
- Application Number
- CN202210053516.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-18
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-01-18
AI Technical Summary
The prior art is difficult to prevent, costly, and poor accuracy in crop pest monitoring, and is difficult to popularize, and the prevention model is immature.
A crop pest monitoring system was designed, including a central processor, a static monitoring device and a dynamic monitoring drone. The image data is processed using convolutional layers, combined with pest prevention models and dynamic weight allocation units, and generated pest development maps and pesticides. The spraying device performs accurate spraying.
It effectively reduces the cost and difficulty of the pest and disease monitoring system, improves the monitoring accuracy, realizes accurate prevention and real-time early warning, and optimizes the pest and disease prevention model.
Smart Images

Figure CN115291541B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of agricultural planting pest control, and in particular to a crop pest monitoring system and method. Background Art
[0002] Crops such as wheat and corn are the important foundation of food security. Monitoring and controlling pests and diseases during the growth of wheat, corn and other food crops is of great significance for increasing production and efficiency and maintaining food security. Under the current land situation in China, the characteristics of wheat, corn and other crops are large planting areas but scattered planting plots. Even though land contracting has begun to form farms of a certain scale in recent years, it is difficult to achieve centralized management, which brings certain difficulties to the monitoring and prevention of crop growth pests and diseases. Crop pests and diseases should be prevented first. If they are treated after they have occurred, the crops have been damaged by pests and diseases and the losses have already occurred.
[0003] Although the technical solutions proposed by existing technologies for the current status of pest and disease prevention have solved the problem of pest and disease monitoring to a certain extent, the prevention model is immature; secondly, these system methods mainly rely on technical routes such as satellite remote sensing, which are high-cost and have poor monitoring accuracy, and are difficult to popularize. Summary of the invention
[0004] The present invention provides a crop disease and insect pest monitoring system and method, which solve the problems of high difficulty in preventing crop diseases and insect pests, high cost investment in infrastructure construction, high difficulty in popularization, and poor prediction function.
[0005] To solve the above problems, a crop disease and insect pest monitoring system described in the present invention includes a central processing unit, a static monitoring device, a dynamic monitoring drone and a pesticide spraying device. The central processing unit is provided with a communication unit, which communicates with the static monitoring device, the dynamic monitoring drone and the pesticide spraying device respectively. The central processing unit is provided with an intelligent image processing module, and the intelligent image processing module is provided with a convolution layer 1 and a convolution layer 2. The convolution layer 1 processes the image collected from the static monitoring device, and the convolution layer 1 simultaneously transmits the processed image to a database and a matching analysis unit, and the matching analysis unit obtains similar disease and insect pest morphological images from the database for possible disease and insect pest analysis; the convolution layer 2 processes the image collected from the dynamic monitoring drone, and the convolution layer 2 transmits the processed image to a prevention overlap analysis unit. The prevention overlap analysis unit has a built-in disease and insect pest prevention model and a dynamic weight allocation unit, which can be optimized through continuous learning. The overlap prevention analysis unit records the analysis results in the database and transmits them to the external display and warning terminal; the pesticide spraying device sprays pesticides according to the pesticide ratio fed back by the pest area and pest type given by the central processor, and feeds back the actual spraying to the central processor.
[0006] Furthermore, the static monitoring device includes a bracket, an environmental monitoring module, a solar cell panel, a communication module and an image acquisition module. A spiral bracket is provided at the lower part of the bracket; the environmental monitoring module is arranged at the lower part, and the distance from the ground is set according to the crop variety and the key environmental layer that causes pests and diseases; the communication module and the image acquisition module are arranged at the top of the bracket, and the image acquisition module has a built-in intelligent camera module and an image temporary storage module. The intelligent camera module has a built-in automatic photo-taking program, which can automatically change the camera focal length, camera rotation and intermittent photo-taking, take photos of crops from far and near and locally, store the images in the image temporary storage module, and upload the images to the central processor through the communication module.
[0007] Furthermore, the static monitoring devices are staggeredly fixed in the farmland, and a triangle formed by connecting three adjacent static monitoring devices is an equilateral triangle.
[0008] The present invention comprises a method for monitoring crop pests and diseases, the steps comprising:
[0009] Step S01, the static monitoring device collects crop images according to a preset period and transmits them to the convolution layer 1 in the central processor in real time;
[0010] Step S02, the convolution layer 1 receives and analyzes the image transmitted by the convolution layer 1, including the analysis of image color, shape, spots, target interlacing and environmental factors, and stores the results in a database;
[0011] Step S03, the matching analysis unit receives the image parsed by the convolutional layer 1 and searches for an approximate pest and disease trait model from a database;
[0012] Step S04, the matching analysis unit initially gives the pest results. If normal, it is ignored. If at least one pest similar model is matched, a white pest area mark and alarm information are issued. The alarm information includes the possible pest name, location, original image of the relevant image, analysis image, model overlap image and urgency level;
[0013] Step S05, manually confirm the alarm, manually analyze the alarm information, and trigger the dynamic monitoring drone 300 after eliminating misjudgment;
[0014] Step S06, the dynamic monitoring drone performs comprehensive sampling of the area marked as pests and diseases and the adjacent areas within a certain range;
[0015] Step S07, the convolutional layer 2 receives and analyzes the sampled data of the dynamic monitoring drone;
[0016] Step S08, the overlap prevention analysis unit receives the convolutional layer 2 analysis data and obtains the convolutional layer 1 analysis data from the database;
[0017] Step S09, the prevention overlap analysis unit retrieves the pest prevention model and weights, applies the current climate environment weights to the data acquired in step S08, combines the pest prevention model to generate a pest development map with time as the horizontal axis, and generates alarm information, which includes the pest development map and the degree of urgency;
[0018] Step S10, preventing the overlap analysis unit data from self-learning to rebuild the pest prevention model and dynamic weight allocation rules, and storing the relevant data in a database.
[0019] The specific steps of the method for generating the pest and disease development map of the prevention overlap analysis unit in step S09 are as follows:
[0020] Step S901, the overlap prevention analysis unit processes the convolutional layer 1 parsing data with pest and disease alarms and obtains the location coordinates;
[0021] Step S902, the overlap prevention analysis unit obtains images and related data of the convolution layer 1 analysis obtained by the 6 to 8 nearest static monitoring devices of the location coordinates of the pests and diseases;
[0022] Step S903, the overlap prevention analysis unit performs mosaic analysis on the images and confirms the center point of the occurrence of pests and diseases in combination with the pest and disease prevention model;
[0023] Step S904, the overlap prevention analysis unit processes the regional pest and disease data sent by the convolutional layer 2 to lock the affected area;
[0024] Step S905, the overlap prevention analysis unit analyzes the convolution layer 1 and the convolution layer 2 images to find key points in the same area, overlaps and optimizes the points, and generates a current pest and disease spectrum map.
[0025] Step S906, the prevention overlap analysis unit generates a pest and disease development map with time as the horizontal axis based on the current pest and disease spectrum map, the current climate environment weight and the pest and disease prevention model.
[0026] The environmental weights in the pest and disease development map of the prevention and analysis unit in step S09 are overlapped , the spread .
[0027] Furthermore, the environmental weight k ranges from 0.34 to 3.2.
[0028] The present invention effectively solves the problems of high deployment cost, poor monitoring accuracy, high difficulty in popularization and inaccurate prevention of pests and diseases in pest monitoring systems. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 It is a schematic diagram of the overall working framework of the present invention;
[0030] Figure 2 It is a schematic diagram of the structure of the static monitoring device of the present invention;
[0031] Figure 3 This is a schematic diagram of farmland distribution of the static monitoring device of the present invention;
[0032] Figure 4 This is a schematic diagram of the structure of the central processing unit of the present invention;
[0033] Figure 5 It is a schematic diagram of the working process of the present invention;
[0034] Figure 6 A schematic diagram of a workflow for generating a pest and disease development map for preventing overlap analysis units of the present invention;
[0035] Figure 7 It is a schematic diagram of the development of pests and diseases of the present invention; DETAILED DESCRIPTION
[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0037] See also Figure 1 and Figure 4 A crop pest and disease monitoring system includes a central processing unit 100, a static monitoring device 200, a dynamic monitoring drone 300 and a pesticide spraying device 400. The central processing unit 100 is provided with a communication unit for communicating with the static monitoring device 200, the dynamic monitoring drone 300 and the pesticide spraying device 400 respectively. The central processing unit 100 is provided with an intelligent image processing module, and the intelligent image processing module is provided with a convolution layer 1 and a convolution layer 2. The convolution layer 1 processes the image collected from the static monitoring device 200, and the convolution layer 1 transmits the processed image to the database and the matching analysis unit at the same time. At the same time, the matching analysis unit obtains similar pest and disease morphological images from the database for possible pest and disease analysis; the convolution layer 2 processes the image collected from the dynamic monitoring drone 300, and the convolution layer 2 transmits the processed image to the prevention overlap analysis unit. The prevention overlap analysis unit has a built-in pest and disease prevention model and a dynamic weight allocation unit, which can be optimized through continuous learning. The overlap prevention analysis unit records the analysis results in a database and transmits them to an external display and warning terminal for manual review and verification; the pesticide spraying device 400 sprays pesticides based on the pesticide ratio fed back by the pest area and pest type given by the central processor 100, and feeds back the actual spraying to the central processor 100.
[0038] Please refer again Figure 1 and Figure 2 The static monitoring device 200 includes a bracket 201, an environmental monitoring module 202, a solar panel 203, a communication module 204 and an image acquisition module 205. The lower part of the bracket 201 is provided with a spiral bracket, which can be buried in the soil to increase wind resistance and overall stability and reduce land occupation; the environmental monitoring module 202 is arranged at the lower part, and the distance from the ground is set according to the crop variety and the key environmental layer causing pests and diseases; the communication module 204 and the image acquisition module 205 are arranged at the top of the bracket 201, and the image acquisition module 205 has a built-in intelligent camera module and an image temporary storage module. The intelligent camera module has a built-in automatic photo program, which can automatically change the camera focal length, camera rotation and intermittent photo taking, take photos of crops from far and near and locally, and store the images in the image temporary storage module, and upload the images to the central processor 100 through the communication module 204. The image temporary storage module can avoid the loss of images that cannot be uploaded when the signal is not good.
[0039] Please refer again Figure 3 The static monitoring devices 200 are staggered and fixed in the farmland, and the triangle formed by the connection line of three adjacent static monitoring devices 200 is an equilateral triangle, which reduces the number of static monitoring devices 200 and obtains the maximum coverage.
[0040] Please refer again Figure 1 , Figure 4 and Figure 5 The present invention provides a method for monitoring crop pests and diseases, the method steps comprising:
[0041] Step S01, the static monitoring device 200 collects crop images according to a preset period and transmits them to the convolution layer 1 in the central processor 100 in real time;
[0042] Step S02, the convolution layer 1 receives and analyzes the image transmitted by the convolution layer 1, including the analysis of image color, shape, spots, target interlacing and environmental factors, and stores the results in a database;
[0043] Step S03, the matching analysis unit receives the image parsed by the convolutional layer 1 and searches for an approximate pest and disease trait model from a database;
[0044] Step S04, the matching analysis unit initially gives the pest results. If normal, it is ignored. If at least one pest similar model is matched, a white pest area mark and alarm information are issued. The alarm information includes the possible pest name, location, original image of the relevant image, analysis image, model overlap image and urgency level;
[0045] Step S05, manually confirm the alarm, manually analyze the alarm information, and trigger the dynamic monitoring drone 300 after eliminating misjudgment;
[0046] Step S06, the dynamic monitoring drone 300 performs comprehensive sampling of the area marked as pests and diseases and the adjacent areas within a certain range;
[0047] Step S07, the convolutional layer 2 receives and analyzes the sampled data of the dynamic monitoring drone 300;
[0048] Step S08, the overlap prevention analysis unit receives the convolutional layer 2 analysis data and obtains the convolutional layer 1 analysis data from the database;
[0049] Step S09, the prevention overlap analysis unit retrieves the pest prevention model and weights, applies the current climate environment weights to the data acquired in step S08, combines the pest prevention model to generate a pest development map with time as the horizontal axis, and generates alarm information, which includes the pest development map and the degree of urgency;
[0050] Step S10, preventing the overlap analysis unit data from self-learning to reconstruct the pest prevention model and dynamic weight allocation rules, and storing the relevant data in a database.
[0051] Please refer again Figure 1 , Figure 4 , Figure 5 and Figure 6 The following is a detailed description of the generation of the pest and disease development map of the prevention overlap analysis unit in step S09. The specific steps are as follows:
[0052] Step S901, the overlap prevention analysis unit processes the convolutional layer 1 parsing data with pest and disease alarms and obtains the location coordinates;
[0053] Step S902, the overlap prevention analysis unit obtains images and related data of the convolution layer 1 analysis obtained by the 6 to 8 nearest neighboring static monitoring devices 200 of the location coordinates of the pests and diseases;
[0054] Step S903, the overlap prevention analysis unit performs mosaic analysis on the images and confirms the center point of the occurrence of pests and diseases in combination with the pest and disease prevention model;
[0055] Step S904, the overlap prevention analysis unit processes the regional pest and disease data sent by the convolutional layer 2 to lock the affected area;
[0056] Step S905, the overlap prevention analysis unit analyzes the convolution layer 1 and the convolution layer 2 images to find key points in the same area, overlaps and optimizes the points, and generates a current pest and disease spectrum map.
[0057] Step S906, the prevention overlap analysis unit generates a pest and disease development map with time as the horizontal axis based on the current pest and disease spectrum map, the current climate environment weight and the pest and disease prevention model.
[0058] Please refer again Figure 7 , the following is a detailed description of the spread range D of the pest and disease development map of the prevention overlap analysis unit in step S09. The pest and disease development map is a map of the spread range D of pests and diseases with time t as the unit. It should be noted that the "0" point in the figure is the center point of the logarithmic growth period of the current crop pests and diseases. The basic statistical standard of the development map here is the time taken for pests and diseases to develop to the logarithmic growth period of the biological growth "S" curve in a certain range of crops. Pests and diseases in the logarithmic growth period will grow and spread rapidly, and pests and diseases must be controlled within the logarithmic growth period;
[0059] The most important factors for biological growth are environmental factors. The most important factors affecting biological growth are temperature T, humidity H, and eutrophication N. In order to facilitate the subsequent changes in the spread rate of pests and diseases caused by environmental factors, the environmental weight k is needed.
[0060] ,
[0061] The value of k ranges from 0.34 to 3.2.
[0062] The spread range is the continuous accumulation of time factors, and the diffusion speed V also tends to decrease due to the diffusion distance. Eventually, as the organism grows, it enters the death period, and the diffusion speed eventually tends to 0. Therefore, the spread range is the result of the accumulation of time t and diffusion speed changes:
[0063] ,
[0064] ,
[0065] Instantaneous speed
[0066] Instantaneous speed V n It is predicted by the pest prevention model that it is affected by environmental factors and the "S" curve of biological growth, so,
[0067] .
[0068] Although the present invention has been described above with reference to the embodiments, various modifications may be made thereto and parts thereof may be replaced with equivalents without departing from the scope of the present invention. In particular, as long as there is no structural conflict, the various features in the embodiments disclosed by the present invention may be used in combination with each other in any manner, and the fact that these combinations are not exhaustively described in this specification is only for the sake of omitting space and saving resources. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.
Claims
1. A crop pest monitoring system, Features: The invention comprises a central processing unit (100), a static monitoring device (200), a dynamic monitoring drone (300) and a pesticide spraying device (400), wherein the central processing unit (100) is provided with a communication unit for communicating with the static monitoring device (200), the dynamic monitoring drone (300) and the pesticide spraying device (400) respectively, and wherein the central processing unit (100) is provided with an intelligent image processing module, wherein the intelligent image processing module is provided with a convolution layer 1 and a convolution layer 2, wherein the convolution layer 1 processes an image collected from the static monitoring device (200), and the convolution layer 1 simultaneously transmits the processed image to a database and a matching analysis unit, and wherein the matching analysis unit Acquire similar pest morphological images from the database to perform possible pest analysis; the convolution layer 2 processes images collected by the dynamic monitoring drone (300), and the convolution layer 2 transmits the processed images to a prevention overlap analysis unit, the prevention overlap analysis unit has a built-in pest prevention model and a dynamic weight allocation unit, and can be optimized through continuous learning; the prevention overlap analysis unit records the analysis results in the database and transmits them to an external display and warning terminal; the pesticide spraying device (400) sprays pesticides according to the pesticide ratio fed back by the pest area and pest type given by the central processor (100), and feeds back the actual spraying to the central processor (100); The static monitoring device (200) comprises a support (201), an environmental monitoring module (202), a solar cell panel (203), a communication module (204) and an image acquisition module (205); a spiral support is provided at the bottom of the support (201); the environmental monitoring module (202) is arranged at the bottom, and the distance from the ground is set according to the crop variety and the key environmental layer causing pests and diseases; the communication module (204) and the image acquisition module (205) are arranged at the top of the support (201); the image acquisition module (205) has a built-in intelligent camera module and an image temporary storage module; the intelligent camera module has a built-in automatic photography program, which can automatically change the camera focal length, camera rotation and intermittent photography, take photos of crops at a distance and in part, and store the images in the image temporary storage module, and upload the images to the central processor (100) through the communication module (204); The static monitoring devices (200) are staggered and fixed in the farmland, and the triangle formed by connecting three adjacent static monitoring devices (200) is an equilateral triangle; The steps of implementing the crop pest and disease monitoring method based on the above system include: Step S01, the static monitoring device collects crop images according to a preset period and transmits them to the convolution layer 1 in the central processor in real time; Step S02, the convolution layer 1 receives and analyzes the image transmitted by the convolution layer 1, including image color, shape, spots, target interlacing and environmental factors, and stores the results in a database; Step S03, the matching analysis unit receives the image parsed by the convolutional layer 1 and searches for an approximate pest and disease trait model from a database; Step S04, the matching analysis unit initially gives the pest results. If normal, it is ignored. If at least one pest similar model is matched, a white pest area mark and alarm information are issued. The alarm information includes the possible pest name, location, original image of the relevant image, analysis image, model overlap image and urgency level; Step S05, manually confirm the alarm, manually analyze the alarm information, and trigger the dynamic monitoring drone 300 after eliminating misjudgment; Step S06, the dynamic monitoring drone performs comprehensive sampling of the area marked as pests and diseases and the adjacent areas within a certain range; Step S07, the convolutional layer 2 receives and analyzes the sampled data of the dynamic monitoring drone; Step S08, the overlap prevention analysis unit receives the convolutional layer 2 analysis data and obtains the convolutional layer 1 analysis data from the database; Step S09, the prevention overlap analysis unit retrieves the pest prevention model and weights, applies the current climate environment weights to the data acquired in step S08, combines the pest prevention model to generate a pest development map with time as the horizontal axis, and generates alarm information, which includes the pest development map and the degree of urgency; Step S10, preventing the overlap analysis unit data from self-learning to reconstruct the pest prevention model and dynamic weight allocation rules, and storing the relevant data in a database; The specific steps of the method for generating a pest and disease development map for preventing overlapping analysis units in step S09 are as follows: Step S901, the overlap prevention analysis unit processes the convolutional layer 1 parsing data with pest and disease alarms and obtains the location coordinates; Step S902, the overlap prevention analysis unit obtains images and related data of the convolution layer 1 analysis obtained by the 6 to 8 nearest neighboring static monitoring devices of the location coordinates of the pests and diseases; Step S903, the overlap prevention analysis unit performs mosaic analysis on the images and confirms the center point of the occurrence of pests and diseases in combination with the pest and disease prevention model; Step S904, the overlap prevention analysis unit processes the regional pest and disease data sent by the convolutional layer 2 to lock the affected area; Step S905, the overlap prevention analysis unit analyzes the convolution layer 1 and the convolution layer 2 images to find key points in the same area, overlaps and optimizes the points, and generates a current pest and disease spectrum map; Step S906, the prevention overlap analysis unit generates a pest and disease development map with time as the horizontal axis based on the current pest and disease spectrum map, the current climate environment weight and the pest and disease prevention model.
2. A method for monitoring crop diseases and insect pests according to claim 1, Features: The environmental weights in the pest and disease development map of the prevention and analysis unit in step S09 are overlapped , where T is the temperature that affects biological growth, H is the humidity that affects biological growth, and N is the eutrophication that affects biological growth; , , Instantaneous speed , Spread , Among them, D n To monitor the spread of crop pests and diseases, D n-1 is the spread range of crop pests and diseases at the last measurement time, t n is the current measurement time of crop pest and disease monitoring, t n-1 is the last measurement time of crop pest monitoring, V n is the instantaneous speed.
3. A method for monitoring crop diseases and insect pests according to claim 2, Features: The environmental weight k ranges from 0.34 to 3.2.
