CN116310489A - Agricultural pest monitoring system and method based on big data - Google Patents

Abstract

The invention provides an agricultural pest monitoring system and method based on big data, wherein the system comprises: the branch and leaf image acquisition module is used for acquiring branch and leaf images of crops in the agricultural garden through the camera trolley; the plant disease and insect pest image acquisition module is used for acquiring plant disease and insect pest images of the crops from the big data platform; the plant disease and insect pest crop determining module is used for determining plant disease and insect pest crops based on the branch and leaf images and the plant disease and insect pest images; and the early warning module is used for early warning the pest and disease crops to management personnel in the agricultural garden. According to the agricultural pest monitoring system and method based on big data, a plurality of personnel with pest judgment experience of crops do not need to be arranged for checking the pests of the crops regularly, and labor cost is reduced.

Description

A system and method for monitoring agricultural diseases and insect pests based on big data

technical field

The invention relates to the technical field of data processing, in particular to a system and method for monitoring agricultural diseases and insect pests based on big data.

Background technique

At present, there are a large number of crops planted in the agricultural park, and it is necessary to arrange a number of personnel with experience in judging crop diseases and insect pests to regularly inspect the crops one by one, and the labor cost is relatively high.

Therefore, a solution is urgently needed.

Contents of the invention

One of the purposes of the present invention is to provide a big data-based agricultural pest monitoring system, without arranging multiple personnel with experience in judging crop pests to regularly check crops one by one, reducing labor costs.

A big data-based agricultural disease and insect pest supervision system provided by the embodiments of the present invention includes:

The branch and leaf image acquisition module is used to obtain the branch and leaf image of the crops in the agricultural garden by the camera dolly;

The image acquisition module of diseases and insect pests is used to acquire the images of diseases and insect pests of the crops from the big data platform;

The crops with diseases and insect pests determination module is used to determine the crops with diseases and insect pests based on the images of branches and leaves and the images of diseases and insect pests;

The early warning module is used to warn the management personnel of the agricultural garden of the crops of the diseases and insect pests.

Preferably, the branch and leaf image acquisition module obtains the branch and leaf images of the crops in the agricultural garden through the camera car, and performs the following operations:

obtaining a bird's-eye view image of the crop;

Based on the bird's-eye view image, plan the patrol route of the camera car;

Controlling the camera trolley to move in the agricultural garden based on the patrol route;

When the camera car arrives at any target plant in the crop, control the camera car to take an external image of the target plant;

Based on the external image, determine the re-shooting position of the target plant that needs to be re-shot;

Controlling the camera dolly to capture the internal image of the supplementary shooting position;

The external image and the internal image are used as branch and leaf images.

Preferably, the branch and leaf image acquisition module plans the patrol route of the camera car based on the bird's-eye view image, and performs the following operations:

Extracting park roads and plant gaps connected to the park roads from the bird's-eye view image;

dividing the plant gap into a plurality of gap segments based on a preset segmentation template;

According to the positional relationship between the gap segment and the road in the park, the gap segment is traversed from near to far;

During each traversal, based on the preset first feature extraction template, perform feature extraction on the traversed gap segments to obtain a gap feature set;

matching the gap feature set with a preset standard gap feature set corresponding to the crop to obtain a first matching degree;

If the first matching degree is less than the preset first matching degree threshold, splicing the previously traversed gap segments as the longest moving gap;

Extracting plant distribution from the bird's-eye view image;

Based on the road in the park, the longest moving gap and the distribution of the plants, the patrolling route of the camera car is planned.

Preferably, based on the external image, the branch and leaf image acquisition module determines the re-shooting position of the target plant that needs to be re-shot, and performs the following operations:

Based on the external image, determine whether image acquisition is necessary for the back of the leaf of the target plant;

If so, use the position corresponding to the back of the leaf of the target plant as the supplementary shooting position;

and / or,

Based on the external image, determine whether image acquisition is necessary on the back of the stem of the target plant;

If so, the position corresponding to the back of the branch stem of the target plant is used as the supplementary shooting position;

and / or,

Based on the external image, a shooting blind area of branches and leaves of the target plant is determined, and the position of the blind shooting area of branches and leaves is used as a supplementary shooting position.

Preferably, the branch and leaf image acquisition module determines whether image acquisition is necessary on the back of the leaf of the target plant based on the external image, and performs the following operations:

extracting leaf outlines from said external image;

Extracting the leaf outline based on a preset second feature extraction template to obtain a first outline feature set;

Matching the first profile feature set with a preset first standard profile feature set corresponding to the crop to obtain a second matching degree;

If the second matching degree is less than or equal to the preset second matching degree threshold, it is determined that image acquisition is necessary for the back of the leaf of the target plant.

Preferably, the branch and leaf image acquisition module determines whether image acquisition is necessary on the back of the branch and stem of the target plant based on the external image, including:

extracting stem outlines from said external image;

Extracting the stem outline based on a preset third feature extraction template to obtain a second outline feature set;

Matching the second profile feature set with the preset second standard profile feature set corresponding to the crops to obtain a third matching degree;

If the third matching degree is less than or equal to the preset third matching degree threshold, it is determined that image acquisition is necessary on the back of the branch of the target plant.

Preferably, the branch and leaf image acquisition module controls the camera trolley to capture the internal image of the supplementary shooting position, and performs the following operations:

extracting foliage voids from said external image;

Determining the re-shooting position next to the gap between the branches and leaves, and using it as the target re-shooting position;

Constructing a direction vector based on the target re-shooting position and the straight line direction from the target re-shooting position to the gap center of the foliage gap;

When the vector angle between two consecutive N direction vectors is within the preset vector angle interval, determine the relative center position from the target supplementary shooting positions corresponding to the N consecutive direction vectors;

Taking the straight line direction from the center of the gap to the relative middle position as the shooting direction of the lens;

Controlling the camera on the camera trolley to go to the center of the gap to shoot the internal image of the supplementary shooting position in the shooting direction of the lens.

Preferably, the image acquisition module of diseases and insect pests obtains the images of diseases and insect pests of the crops from the big data platform, and performs the following operations:

Acquiring a preset image retrieval template of diseases and insect pests corresponding to the crops;

Based on the image retrieval template of diseases and insect pests, the images of diseases and insect pests of the crops are retrieved from the big data platform.

Preferably, the crops with diseases and insect pests determining module determines the crops with diseases and insect pests based on the images of branches and leaves and the images of diseases and insect pests, including:

The pest image is input into the neural network model as a training sample for training to obtain a pest identification model;

The branch and leaf images are input into the pest recognition model to determine crops with pests and diseases.

A method for supervising agricultural diseases and insect pests based on big data provided by the embodiments of the present invention includes:

Step S1: Obtain the branch and leaf images of the crops in the agricultural garden through the camera car;

Step S2: Obtaining images of diseases and insect pests of the crops from the big data platform;

Step S3: Based on the images of branches and leaves and the images of pests and diseases, determine crops with pests and diseases;

Step S4: Alerting the crops of the diseases and insect pests to the management personnel of the agricultural garden.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

The technical solutions of the present invention will be described in further detail below with reference to the accompanying drawings and embodiments.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the description, and are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the attached picture:

Fig. 1 is the schematic diagram of a kind of agricultural disease and insect pest monitoring system based on big data in the embodiment of the present invention;

Fig. 2 is a schematic diagram of a big data-based agricultural pest and disease supervision method in an embodiment of the present invention.

Detailed ways

The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present invention, and are not intended to limit the present invention.

The embodiment of the present invention provides a big data-based agricultural disease and insect pest supervision system, as shown in Figure 1, including:

The branch and leaf image acquisition module 1 is used to obtain the branch and leaf image of the crops in the agricultural garden by the camera dolly;

Pest and disease image acquisition module 2, used to obtain the image of pest and disease of the crops from the big data platform;

The pest-damage crop determining module 3 is configured to determine the pest-damage crop based on the image of the branches and leaves and the image of the pest damage;

The early warning module 4 is used to warn the management personnel of the agricultural garden of the crops of the diseases and insect pests.

The working principle and beneficial effects of the above-mentioned technical scheme are:

The camera car is a mobile car with a camera. The big data platform collects images of crop diseases and insect pests based on big data technology. In a specific application, the camera car moves in the agricultural garden, takes images of branches and leaves of crops, and determines crops with diseases and insect pests based on the images of branches and leaves and the images of diseases and insect pests obtained from the data platform, for example, by comparing the two images. Early warning of plant diseases and insect pests is given to management personnel, for example, notifying management personnel by means of mobile phone information notification. There is no need to arrange for multiple personnel with experience in judging crop diseases and insect pests to regularly inspect crops one by one for diseases and insect pests, which reduces labor costs.

In one embodiment, the branch and leaf image acquisition module 1 obtains the branch and leaf images of the crops in the agricultural garden through a camera car, and performs the following operations:

obtaining a bird's-eye view image of the crop;

Based on the bird's-eye view image, plan the patrol route of the camera car;

Controlling the camera trolley to move in the agricultural garden based on the patrol route;

When the camera car arrives at any target plant in the crop, control the camera car to take an external image of the target plant;

Based on the external image, determine the re-shooting position of the target plant that needs to be re-shot;

Controlling the camera dolly to capture the internal image of the supplementary shooting position;

The external image and the internal image are used as branch and leaf images.

The working principle and beneficial effects of the above-mentioned technical scheme are:

The bird's-eye view image can be obtained by shooting down from the bird's-eye camera set at a high place in the agricultural park. Based on the bird's-eye view image, the route for the camera trolley to move and the distribution of crops in the agricultural park can be determined, so it can be used for planning the patrolling route. Based on the patrol route, the camera car is controlled to move in the agricultural garden. Normally, the branches and leaves of the plant are superimposed and scattered, and the pests and diseases will not only occur on the front of the branches and leaves, but also on the back of the branches and leaves. Therefore, a single external image of the target plant is not enough to determine the pests and diseases. Therefore, based on the external image, determine the target plant. For locations that need to be re-shot internally, such as the back of branches and leaves and internal shooting blind spots, etc., control the camera car to take the internal images of the re-shoot positions, and use the external images and internal images as branch and leaf images to improve the acquisition of branch and leaf images for pest monitoring. comprehensive and reasonable.

In one embodiment, the branch and leaf image acquisition module 1 plans the patrol route of the camera car based on the bird's-eye view image, and performs the following operations:

Extracting park roads and plant gaps connected to the park roads from the bird's-eye view image;

dividing the plant gap into a plurality of gap segments based on a preset segmentation template;

According to the positional relationship between the gap segment and the road in the park, the gap segment is traversed from near to far;

During each traversal, based on the preset first feature extraction template, perform feature extraction on the traversed gap segments to obtain a gap feature set;

matching the gap feature set with a preset standard gap feature set corresponding to the crop to obtain a first matching degree;

If the first matching degree is less than the preset first matching degree threshold, splicing the previously traversed gap segments as the longest moving gap;

Extracting plant distribution from the bird's-eye view image;

Based on the road in the park, the longest moving gap and the distribution of the plants, the patrolling route of the camera car is planned.

The working principle and beneficial effects of the above-mentioned technical scheme are:

There are park roads in the agricultural park, which are mostly set up outside the planting area, and can be used by camera trolleys. Secondly, normally, when the plants are planted, they will be arranged in multiple rows. Therefore, there will be a gap between the two rows of plants, which can be used by the camera trolley. However, as the plants grow, the plants will become lush and the gaps The area will become smaller, and if the camera car is forced to drive in the gap, it may damage the plants. Therefore, the camera car can only shoot images of branches and leaves on the roads of the park and the gaps of plants that can be driven by the roads of the park. Then extract the plant gaps connected with the roads in the park, and divide them into multiple gap ends based on the preset segmentation template; Make a vertical line perpendicular to the mid-perpendicular line every 0.5 meters on the vertical line, cancel the mid-perpendicular line, and each vertical line divides the plant gap into multiple sections. The gap feature set of the gap segment is extracted, and the gap feature set includes: the minimum gap width and the average gap width, etc. The preset standard gap feature set corresponding to the crop is specifically that when the crop is planted in two columns, the gap area between the two columns can accommodate the image features presented outside the gap area when the camera trolley is driving, including: the minimum gap width is 0.35 meters and The average width of the gap is 0.42 meters. Match the gap feature set with the standard gap feature set. If the first matching degree is less than or equal to the preset first matching degree threshold, it means that the gap segment cannot accommodate the camera car, and the previously traversed gap segments are stitched together as the longest moving gap . Then extract the plant distribution from the bird's-eye view image, and plan the patrolling route of the camera car based on the park road, the longest moving gap and the plant distribution. The embodiment of the present invention adaptively determines the longest moving gap that can be used by the shooting car in the plant gap, which improves the applicability of planning the patrol route of the shooting car in the plant planting area, and at the same time, it is more intelligent.

In one embodiment, the branch and leaf image acquisition module 1 determines the position of the target plant for re-shooting that needs to be re-shot based on the external image, and performs the following operations:

Based on the external image, determine whether image acquisition is necessary for the back of the leaf of the target plant;

If so, use the position corresponding to the back of the leaf of the target plant as the supplementary shooting position;

and / or,

Based on the external image, determine whether image acquisition is necessary on the back of the stem of the target plant;

If so, the position corresponding to the back of the branch stem of the target plant is used as the supplementary shooting position;

and / or,

Based on the external image, a shooting blind area of branches and leaves of the target plant is determined, and the position of the blind shooting area of branches and leaves is used as a supplementary shooting position.

The working principle and beneficial effects of the above-mentioned technical scheme are:

The external image of the plant mostly shows the front of the branches and leaves of the plant. However, because the growth angle of the branches and leaves is different from the angle of the external image taken by the camera trolley, a part of the back of some branches and leaves will be exposed, and the pests and diseases mostly appear regionally. Therefore, if If there is no abnormality in the exposed part, there is no need to collect the back side. Therefore, the first and second methods first determine whether image acquisition is necessary on the back of the leaf/stem of the target plant, and if so, use the corresponding position as the non-shooting position. There is no need to re-shoot the back of each leaf and each stem, which reduces shooting resources and improves the shooting efficiency of branch and leaf images. The third type uses the position of the blind spot in the shooting of branches and leaves as the supplementary shooting position.

In one embodiment, the branch and leaf image acquisition module 1 determines whether image acquisition is necessary on the back of the leaf of the target plant based on the external image, and performs the following operations:

extracting leaf outlines from said external image;

Extracting the leaf outline based on a preset second feature extraction template to obtain a first outline feature set;

Matching the first profile feature set with a preset first standard profile feature set corresponding to the crop to obtain a second matching degree;

If the second matching degree is less than or equal to the preset second matching degree threshold, it is determined that image acquisition is necessary for the back of the leaf of the target plant.

The working principle and beneficial effects of the above-mentioned technical scheme are:

Generally, when a part of the back of the leaf is presented on the external image, the outline of the leaf is more inconsistent with the outline of the leaf when it is spread out on the front. Therefore, the first contour feature set of the leaf contour is extracted, and the first contour feature set includes: contour area, contour maximum length and contour minimum length. The preset first standard outline feature set corresponding to the crops is specifically the features of the outline of the leaves of the crop when the front is spread out, for example, including: outline area of 3.2 square centimeters, maximum outline length of 3 cm and minimum outline length of 1.5 cm. Match the first contour feature set with the first standard contour feature set. If the second matching degree is less than or equal to the preset second matching degree threshold, it means that the back of the leaf is less in the external image and should be collected. . The efficiency and accuracy of determining whether image acquisition is necessary for the back of the leaf of the target plant is improved.

In one embodiment, the branch and leaf image acquisition module 1 determines whether it is necessary to perform image acquisition on the back of the branch and stem of the target plant based on the external image, including:

extracting stem outlines from said external image;

Extracting the stem outline based on a preset third feature extraction template to obtain a second outline feature set;

Matching the second profile feature set with the preset second standard profile feature set corresponding to the crops to obtain a third matching degree;

If the third matching degree is less than or equal to the preset third matching degree threshold, it is determined that image acquisition is necessary on the back of the branch of the target plant.

The working principle and beneficial effects of the above-mentioned technical scheme are:

Generally, when part of the back of the stem is presented on the external image, the outline of the stem is more inconsistent with the outline of the stem when it is spread out on the front. Therefore, the second contour feature set of the branch stem contour is extracted, and the second contour feature set includes: contour area, contour maximum length and contour minimum length. The preset second standard contour feature set corresponding to the crops is specifically the characteristics of the contours of the branches and stems of the crops when they are spread out in front, for example, including: a contour area of 5.1 square centimeters, a maximum contour length of 4 cm, and a minimum contour length of 1.1 cm. Match the second profile feature set with the second standard profile feature set. If the third matching degree is less than or equal to the preset third matching degree threshold, it means that the back of the branch is less in the external image, and it should be collection. The determination efficiency and determination accuracy of whether the back of the branch stem of the target plant is necessary for image acquisition are improved.

In one embodiment, the branch and leaf image acquisition module 1 controls the camera trolley to capture the internal image of the supplementary shooting position, and performs the following operations:

extracting foliage voids from said external image;

Determining the re-shooting position next to the gap between the branches and leaves, and using it as the target re-shooting position;

Constructing a direction vector based on the target re-shooting position and the straight line direction from the target re-shooting position to the gap center of the foliage gap;

When the vector angle between two consecutive N direction vectors is within the preset vector angle interval, determine the relative center position from the target supplementary shooting positions corresponding to the N consecutive direction vectors;

Taking the straight line direction from the center of the gap to the relative middle position as the shooting direction of the lens;

Controlling the camera on the camera trolley to go to the center of the gap to shoot the internal image of the supplementary shooting position in the shooting direction of the lens.

The working principle and beneficial effects of the above-mentioned technical scheme are:

There will be gaps between the branches and leaves, and the camera can enter the gaps for supplementary shooting. N is a positive integer. The preset vector included angle ranges from 0° to 120°. When the vector angle between two consecutive N direction vectors is within the preset vector angle interval, it means that the camera can perform a one-time measurement of the target supplementary shooting position corresponding to these consecutive N direction vectors at the target supplementary shooting position. For re-shooting, determine the relative center position from the target re-shoot positions corresponding to these consecutive N direction vectors. The straight line direction from the center of the gap to the relative center position can be used as the shooting direction of the lens. Control the camera on the camera car to go to the center of the gap to shoot with the lens direction to shoot the internal image of the supplementary shooting position. Improve the rationality of the camera car's photo control, reduce re-shooting resources, and improve the efficiency of re-shooting.

In one embodiment, the image acquisition module 2 of diseases and insect pests acquires the images of diseases and insect pests of the crops from the big data platform, and performs the following operations:

Obtaining a preset image retrieval template of diseases and insect pests corresponding to the crops;

Based on the image retrieval template of diseases and insect pests, the images of diseases and insect pests of the crops are retrieved from the big data platform.

The preset image retrieval template of diseases and insect pests corresponding to the crops is specifically: the retrieval condition for retrieving the images of diseases and insect pests of the types of diseases and insect pests that may be produced by the crops. Based on the image retrieval template of diseases and insect pests, images of diseases and insect pests of crops are retrieved from the big data platform.

In one embodiment, the crops with diseases and insect pests determination module 3 determines the crops with diseases and insect pests based on the images of branches and leaves and the images of diseases and insect pests, including:

The pest image is input into the neural network model as a training sample for training to obtain a pest identification model;

The branch and leaf images are input into the pest recognition model to determine crops with pests and diseases.

The images of diseases and insect pests are input into the neural network model as training samples for training, and the identification model of diseases and insect pests is obtained after the training reaches convergence. .

The embodiment of the present invention provides a method for supervising agricultural diseases and insect pests based on big data, as shown in Figure 2, including:

Step S1: Obtain the branch and leaf images of the crops in the agricultural garden through the camera car;

Step S2: Obtaining images of diseases and insect pests of the crops from the big data platform;

Step S3: Based on the images of branches and leaves and the images of pests and diseases, determine crops with pests and diseases;

Step S4: Alerting the crops of the diseases and insect pests to the management personnel of the agricultural garden.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. A big data-based agricultural pest monitoring system, characterized in that it comprises: The branch and leaf image acquisition module is used to obtain the branch and leaf image of the crops in the agricultural garden by the camera dolly; The image acquisition module of diseases and insect pests is used to acquire the images of diseases and insect pests of the crops from the big data platform; The crops with diseases and insect pests determination module is used to determine the crops with diseases and insect pests based on the images of branches and leaves and the images of diseases and insect pests; The early warning module is used to warn the management personnel of the agricultural garden of the crops of the diseases and insect pests. 2. a kind of agricultural disease and insect pest monitoring system based on big data as claimed in claim 1, is characterized in that, described branch and leaf image acquisition module obtains the branch and leaf image of the crop in the agricultural garden by camera dolly, carries out following operation: obtaining a bird's-eye view image of the crop; Based on the bird's-eye view image, plan the patrol route of the camera car; Controlling the camera trolley to move in the agricultural garden based on the patrol route; When the camera car arrives at any target plant in the crop, control the camera car to take an external image of the target plant; Based on the external image, determine the re-shooting position of the target plant that needs to be re-shot; Controlling the camera dolly to capture the internal image of the supplementary shooting position; The external image and the internal image are used as branch and leaf images. 3. A kind of agricultural disease and insect pest monitoring system based on big data as claimed in claim 2, it is characterized in that, described branch and leaf image acquisition module is based on described bird's-eye view image, plans the patrol route of described camera trolley, performs the following operations: Extracting park roads and plant gaps connected to the park roads from the bird's-eye view image; dividing the plant gap into a plurality of gap segments based on a preset segmentation template; According to the positional relationship between the gap segment and the road in the park, the gap segment is traversed from near to far; During each traversal, based on the preset first feature extraction template, perform feature extraction on the traversed gap segments to obtain a gap feature set; matching the gap feature set with a preset standard gap feature set corresponding to the crop to obtain a first matching degree; If the first matching degree is less than the preset first matching degree threshold, splicing the previously traversed gap segments as the longest moving gap; Extracting plant distribution from the bird's-eye view image; Based on the road in the park, the longest moving gap and the distribution of the plants, the patrolling route of the camera car is planned. 4. A kind of big data-based agricultural disease and insect pest supervision system as claimed in claim 2, is characterized in that, described branch and leaf image acquisition module is based on described external image, determines that the inside of described target plant needs to carry out the re-shooting of re-shooting location, do the following: Based on the external image, determine whether image acquisition is necessary for the back of the leaf of the target plant; If so, use the position corresponding to the back of the leaf of the target plant as the supplementary shooting position; and / or, Based on the external image, determine whether image acquisition is necessary on the back of the stem of the target plant; If so, the position corresponding to the back of the branch stem of the target plant is used as the supplementary shooting position; and / or, Based on the external image, a shooting blind area of branches and leaves of the target plant is determined, and the position of the blind shooting area of branches and leaves is used as a supplementary shooting position. 5. A kind of big data-based agricultural disease and insect pest supervision system as claimed in claim 4, is characterized in that, described branch and leaf image acquisition module is based on described external image, determines whether it is necessary to carry out image acquisition on the back side of the leaf of described target plant , do the following: extracting leaf outlines from said external image; Extracting the leaf outline based on a preset second feature extraction template to obtain a first outline feature set; Matching the first profile feature set with a preset first standard profile feature set corresponding to the crop to obtain a second matching degree; If the second matching degree is less than or equal to the preset second matching degree threshold, it is determined that image acquisition is necessary for the back of the leaf of the target plant. 6. A kind of big data-based agricultural disease and insect pest supervision system as claimed in claim 4, it is characterized in that, based on the external image, the branch and leaf image acquisition module determines whether it is necessary to image the back of the branch and stem of the target plant. collection, including: extracting stem outlines from said external image; Extracting the stem outline based on a preset third feature extraction template to obtain a second outline feature set; Matching the second profile feature set with the preset second standard profile feature set corresponding to the crops to obtain a third matching degree; If the third matching degree is less than or equal to the preset third matching degree threshold, it is determined that image acquisition is necessary on the back of the branch of the target plant. 7. A kind of agricultural disease and insect pest monitoring system based on big data as claimed in claim 2, is characterized in that, described branch and leaf image acquisition module controls described camera trolley to take the internal image of described supplementary shooting position, performs the following operations: extracting foliage voids from said external image; Determining the re-shooting position next to the gap between the branches and leaves, and using it as the target re-shooting position; Constructing a direction vector based on the target re-shooting position and the straight line direction from the target re-shooting position to the gap center of the foliage gap; When the vector angle between two consecutive N direction vectors is within the preset vector angle interval, determine the relative center position from the target supplementary shooting positions corresponding to the N consecutive direction vectors; Taking the straight line direction from the center of the gap to the relative middle position as the shooting direction of the lens; Controlling the camera on the camera trolley to go to the center of the gap to shoot the internal image of the supplementary shooting position in the shooting direction of the lens. 8. A kind of agricultural disease and insect pest monitoring system based on big data as claimed in claim 1, is characterized in that, described disease and insect pest image acquisition module obtains the disease and insect pest image of described crop from big data platform, performs the following operations: Acquiring a preset image retrieval template of diseases and insect pests corresponding to the crops; Based on the image retrieval template of diseases and insect pests, the images of diseases and insect pests of the crops are retrieved from the big data platform. 9. A kind of big data-based agricultural disease and insect pest supervision system as claimed in claim 1, is characterized in that, described plant disease and insect pest crop determining module is based on described branch and leaf image and described disease and insect pest image, determines disease and insect pest crop, comprising: The pest image is input into the neural network model as a training sample for training to obtain a pest identification model; The branch and leaf images are input into the pest recognition model to determine crops with pests and diseases. 10. A method for monitoring agricultural pests and diseases based on big data, characterized in that it comprises: Step S1: Obtain the branch and leaf images of the crops in the agricultural garden through the camera car; Step S2: Obtaining images of diseases and insect pests of the crops from the big data platform; Step S3: Based on the images of branches and leaves and the images of pests and diseases, determine crops with pests and diseases; Step S4: Alerting the crops of the diseases and insect pests to the management personnel of the agricultural garden.

Images