Intelligent integrated control method for agricultural equipment

Abstract

The invention discloses an intelligent integrated control method for agricultural machinery equipment, relates to the technical field of agricultural machinery equipment scheduling, and improves the efficiency and accuracy of agricultural machinery equipment scheduling. According to the invention, the farmland visualization graph and the farmland planting chain are established according to the real-time area image data of each operation area, the feature image segment of each plant is acquired according to the pre-stored crop planting scheme, and the farmland visualization graph is updated in real time according to the real-time area image data of each operation area. An operation demand pointer is generated on the farmland planting chain, an agricultural machinery driving road is set on the farmland visualization graph according to the region boundary line, a plurality of agricultural machinery equipment nodes are matched according to the operation demand pointer, and a prepared agricultural machinery driving route is set for each agricultural machinery equipment node according to a matching result; and mapping each prepared agricultural machine driving route to an agricultural machine driving road for driving simulation, and then selecting and executing the agricultural machine equipment nodes and the agricultural machine driving route.

Description

Technical Field

[0001] The invention relates to the technical field of agricultural machinery and equipment scheduling, and in particular to an intelligent integrated control method for agricultural machinery and equipment. Background Art

[0002] With the rapid development of modern agricultural technology, the demand for precision agriculture and intelligent agricultural machinery management is increasing. Traditional farmland management methods often rely on manual inspections and experience-based judgments, which is not only inefficient but also difficult to grasp the dynamic changes of farmland in real time.

[0003] Modern agricultural technology has introduced drones and intelligent agricultural machinery and equipment. Through the real-time regional image data taken by drones, detailed information of farmland can be quickly obtained, including crop growth status, soil moisture, pest and disease conditions, etc. However, due to the large variety of agricultural machinery and equipment, large real-time scenes, and complex plant species, it is difficult for existing technologies to accurately schedule and accurately detect agricultural machinery and equipment and farmland operation needs. For this purpose, an intelligent integrated control method for agricultural machinery and equipment is provided. Summary of the invention

[0004] In order to solve the above technical problems, an object of the present invention is to provide an intelligent integrated control method for agricultural machinery.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent integrated control method for agricultural machinery equipment comprises the following steps:

[0007] Step S1, dividing the target farmland into several operation areas, and then setting up several drones to generate real-time regional image data of each operation area;

[0008] Step S2, according to the pixel difference between each pixel, a number of clusters are set in the real-time regional image data, pixels at the edge position of each cluster are selected to establish a boundary guide vector, and then according to the corresponding relationship between the boundary guide vectors of adjacent clusters, a regional boundary line is divided in the real-time regional image data;

[0009] Step S3: establishing a farmland visualization map and a farmland planting chain according to the real-time regional image data of each operation area, setting a number of agricultural machinery and equipment, obtaining the real-time position information of each agricultural machinery and equipment according to the Beidou positioning device, and setting a number of agricultural machinery and equipment nodes on the farmland visualization map according to the real-time position information of each agricultural machinery and equipment;

[0010] Step S4, update the farmland visualization map in real time according to the real-time regional image data of each operation area, generate an operation demand pointer on the farmland planting chain, set the agricultural machinery driving road on the farmland visualization map according to the regional boundary line, and then match multiple agricultural machinery equipment nodes according to the pre-stored operation demand pointer, set a prepared agricultural machinery driving route for each agricultural machinery equipment node according to the matching result, map each prepared agricultural machinery driving route to the agricultural machinery driving road for driving simulation, and then select the execution agricultural machinery equipment node and agricultural machinery driving route and execute them.

[0011] Furthermore, the generation process of the real-time regional image data includes:

[0012] According to the distribution of crop planting in the target farmland, the target farmland is divided into n operation areas of different areas, where n is a natural number greater than 0;

[0013] The drone is equipped with a camera and a Beidou positioning device, and a fixed shooting range is set for each camera. Four positions are selected at the edge of the target farmland as starting positions, and the diagonally opposite positions of each starting position are mutually end positions, and then four drone flight routes are established with each starting position and end position. During the flight of the drone, real-time video data of its shooting range is continuously obtained through the camera;

[0014] An acquisition detection cycle is set, and each time an acquisition detection cycle starts or ends, a real-time video data segment generated by a previous acquisition detection cycle is intercepted, and the real-time video data segment is divided into several pieces of real-time regional image data by frame.

[0015] Further, the process of setting a plurality of clusters in the real-time regional image data includes:

[0016] Obtain the pixel value of each pixel in each real-time regional image data, establish a two-dimensional coordinate system, and map the real-time regional image data generated by the same drone in each collection and detection cycle onto the two-dimensional coordinate system in chronological order;

[0017] A cluster pixel difference threshold is set to determine whether the pixel difference between adjacent pixels in the real-time regional image data is less than or equal to the cluster pixel difference threshold. If the pixel difference is less than or equal to the cluster pixel difference threshold, the corresponding two pixels are recorded as the same cluster, otherwise no operation is performed.

[0018] Furthermore, the process of establishing the boundary guidance vector includes:

[0019] Select the pixel at the edge of each cluster as the starting point, and select the pixel with the largest pixel difference from the starting point as the secondary end point, and then establish several boundary guide vectors at the edge of each cluster;

[0020] Set the angle threshold and plane distance threshold. Since there are common boundaries between adjacent clusters, there must be some boundary guide vectors between adjacent clusters whose direction angles are greater than the angle threshold. Then obtain the angle degree and Euclidean distance between each boundary guide vector in the adjacent cluster. If the two angle degrees are greater than the angle degree and the Euclidean distance is less than or equal to the plane distance threshold, the corresponding two boundary guide vectors are matched with each other, otherwise no operation is performed.

[0021] Furthermore, the process of dividing the region boundary line in the real-time region image data includes:

[0022] Extend each mutually matching boundary guide vector until they intersect, record the intersection point as the appearance boundary point, connect adjacent appearance boundary points in sequence, and then generate corresponding appearance boundary lines between each cluster, set a number of feature pointers, and the feature pointers are used to extract feature structures and feature pixel value intervals in real-time regional image data;

[0023] Each cluster is traversed through each feature pointer, and each feature structure and feature pixel value interval are obtained according to the traversal result, wherein the feature pixel value interval represents the color distribution or color change characteristics of the corresponding image area;

[0024] The same label is set for clusters with the same characteristic structure and characteristic pixel value interval, and the appearance boundary lines of the edge positions of the regions formed by adjacent clusters with the same label are merged to obtain the region boundary lines.

[0025] Furthermore, the process of establishing the farmland visualization map includes:

[0026] In the process of each UAV arriving at the terminal position along the UAV flight route, each time a collection and detection cycle ends, a number of operation area images and individual plant images are segmented in each real-time area image data according to the area boundary line and the appearance boundary line;

[0027] A farmland visualization map is established based on images of each operation area and images of individual plants, wherein the farmland visualization map is composed of n operation visualization areas, and each operation visualization area contains a number of real-time image models of plants;

[0028] The crop planting plan records the plant species, initial planting time and management operation steps of each operation area, wherein the management operation steps include operation contents and interval time between each operation content, wherein the operation content includes the name of the operation content and the image data of the corresponding plant before executing the relevant operation content;

[0029] According to the crop planting plan records and the location distribution of each operation area in the farmland visualization map, a farmland planting chain is established on the farmland visualization map, where the farmland planting chain consists of n operation nodes;

[0030] Based on the image data of each plant before performing the relevant operation content, the characteristic image fragments of the corresponding plant during each operation content are obtained, and according to the initial planting time of each operation area and the interval time between each operation content in the crop planting plan record, an independent initial detection time and several final operation detection times are set for each operation node.

[0031] Furthermore, several agricultural machinery equipment are set up, and the agricultural machinery equipment is provided with a Beidou positioning device, and then the real-time position information of each agricultural machinery equipment is obtained according to the Beidou positioning device, and several agricultural machinery equipment nodes are set on the farmland visualization map according to the real-time position information of each agricultural machinery equipment, and each agricultural machinery equipment node is marked with a corresponding agricultural machinery equipment number and executable operation content.

[0032] Furthermore, the process of generating the job requirement pointer includes:

[0033] For any operation node, the timing starts from the initial detection time it carries. When a drone starts to collect real-time video data of its corresponding operation area, it is determined whether the current time has reached the next final operation detection time. If not, then at the end of each collection and detection cycle, the corresponding real-time image model of the plant in the farmland visualization map is updated according to the real-time area image data generated in the previous collection and detection cycle, and then the real-time image model of the plant is determined through the feature image fragments to determine whether it meets the requirements for the next operation, and an operation requirement pointer is generated according to the judgment result;

[0034] If the current time has not reached the next final job detection time, but the number of plant real-time image models marked with "executable next job content" is less than the executable threshold, no operation will be performed until the current time reaches the next final job detection time, and a job demand pointer will be directly generated for the corresponding job node.

[0035] Furthermore, the process of preparing the agricultural machinery driving route for driving simulation includes:

[0036] Set the unit moving speed for each agricultural machinery equipment node. Whenever an operation node generates an operation demand pointer, match the agricultural machinery equipment node with the corresponding operation content name and in an idle state on the farmland visualization map according to the operation content name in the operation demand pointer;

[0037] The agricultural machinery driving roads are set based on the distribution of the farmland visualization map on the regional boundary line, and several prepared agricultural machinery driving routes are generated with the operation node with the operation demand pointer as the end position and the current position of each matching agricultural machinery equipment node as the starting position, and the agricultural machinery driving route currently being executed is mapped on the agricultural machinery driving road;

[0038] Map all the prepared agricultural machinery driving routes to the agricultural machinery driving roads for driving simulation. If any of the prepared agricultural machinery driving routes collides with the agricultural machinery driving route that is already being executed during the simulated driving process, the corresponding prepared agricultural machinery driving route will be eliminated, otherwise the corresponding prepared agricultural machinery driving route will be retained;

[0039] Count the estimated execution time of the reserved reserve agricultural machinery driving routes, and then select the reserve agricultural machinery driving route with the shortest estimated execution time and the corresponding agricultural machinery equipment node as the agricultural machinery driving route and the executing agricultural machinery equipment node;

[0040] The agricultural machinery driving route is sent to the agricultural machinery equipment corresponding to the agricultural machinery equipment node, and then the agricultural machinery equipment drives according to the received agricultural machinery driving route, and the corresponding agricultural machinery equipment node is marked as working.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. The present invention sets several clusters in the real-time regional image data according to the pixel difference between each pixel, selects the pixels at the edge position of each cluster to establish the boundary guidance vector, and then divides the regional boundary line in the real-time regional image data according to the corresponding relationship between the boundary guidance vectors of adjacent clusters, thereby realizing the rapid identification of the boundaries of each working area and providing an accurate data basis for the driving route planning of agricultural machinery equipment.

[0043] 2. The present invention obtains the location information of agricultural machinery and equipment in real time through the Beidou positioning device, and accurately sets the nodes of agricultural machinery and equipment on the farmland visualization map, so that agricultural machinery and equipment can be intelligently dispatched according to real-time data, reducing manual intervention and resource waste. At the same time, by matching the operation demand pointer with the pre-stored operation demand pointer, the operation route of agricultural machinery and equipment is optimized, and the operation efficiency and accuracy are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention.

[0045] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0046] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0047] like Figure 1 As shown, an intelligent integrated control method for agricultural machinery equipment includes the following steps:

[0048] Step S1, dividing the target farmland into several operation areas, and then setting up several drones to generate real-time regional image data of each operation area;

[0049] Step S2, according to the pixel difference between each pixel, a number of clusters are set in the real-time regional image data, pixels at the edge position of each cluster are selected to establish a boundary guide vector, and then according to the corresponding relationship between the boundary guide vectors of adjacent clusters, a regional boundary line is divided in the real-time regional image data;

[0050] Step S3: establishing a farmland visualization map and a farmland planting chain according to the real-time regional image data of each operation area, setting a number of agricultural machinery and equipment, obtaining the real-time position information of each agricultural machinery and equipment according to the Beidou positioning device, and setting a number of agricultural machinery and equipment nodes on the farmland visualization map according to the real-time position information of each agricultural machinery and equipment;

[0051] Step S4, update the farmland visualization map in real time according to the real-time regional image data of each operation area, generate an operation demand pointer on the farmland planting chain, set the agricultural machinery driving road on the farmland visualization map according to the regional boundary line, and then match multiple agricultural machinery equipment nodes according to the pre-stored operation demand pointer, set a prepared agricultural machinery driving route for each agricultural machinery equipment node according to the matching result, map each prepared agricultural machinery driving route to the agricultural machinery driving road for driving simulation, and then select the execution agricultural machinery equipment node and agricultural machinery driving route and execute them.

[0052] Furthermore, step S1 is implemented by the following process:

[0053] According to the distribution of crop planting in the target farmland, the target farmland is divided into n operation areas of different areas. It should be noted that there are any differences in plant varieties or planting times between the various operation areas;

[0054] Numbers a1, a2, a3, ..., a are set for each work area. n , where n is a natural number greater than 0;

[0055] Set up four drones, each of which is equipped with a camera and a Beidou positioning device, set a fixed shooting range for each camera, select four positions at the edge of the target farmland as the starting positions, and make the diagonally opposite positions of each starting position the end position, and then establish four drone flight routes with each starting position and the end position, so that the drone can traverse the entire target farmland through the camera along each drone flight route. It should be noted that the width of the drone flight route is equal to the width of the camera's shooting range;

[0056] Four drones simultaneously start from different starting positions and fly over the target farmland along different unmanned flight routes. During the flight of the drones, the cameras continuously acquire real-time video data of the shooting range;

[0057] An acquisition detection cycle is set, and the duration of the acquisition detection cycle is generally 50 to 100 ms. Whenever an acquisition detection cycle starts or ends, a real-time video data segment generated by the previous acquisition detection cycle is intercepted, and the real-time video data segment is divided into several pieces of real-time regional image data by frame.

[0058] Further, step S2 is implemented by the following process:

[0059] Obtain the pixel value of each pixel in each real-time regional image data. Since the plant species or planting time of each operation area in the target area are different, the plants in the real-time regional image data generated by each operation area at the same time must have differences in color or morphology;

[0060] Establish a two-dimensional coordinate system, and map the real-time regional image data generated by the same UAV in each collection and detection cycle onto the two-dimensional coordinate system in chronological order;

[0061] Set the same cluster pixel difference threshold, and then judge whether the pixel difference between adjacent pixels in the real-time regional image data is less than or equal to the same cluster pixel difference threshold. If the pixel difference is less than or equal to the same cluster pixel difference threshold, the corresponding two pixels are recorded as the same cluster, otherwise no operation is performed;

[0062] At the same time, since the species and planting time of the plants in the same operation area are the same, the plants in the same operation area have the same appearance characteristics. At the same time, since the same operation area is a separate individual, each plant has a certain spatial hierarchy in the real-time regional image data, that is, there is a pixel difference boundary between the corresponding parts;

[0063] Select the pixel at the edge of each cluster as the starting point, and select the pixel with the largest pixel difference from the starting point as the secondary end point, and then establish several boundary guide vectors at the edge of each cluster, set the angle threshold and the plane distance threshold. Since there is a common boundary between adjacent clusters, there must be some boundary guide vectors between adjacent clusters whose direction angle is greater than the angle threshold, and then obtain the angle degree and Euclidean distance between each boundary guide vector in the adjacent cluster. If the two angle degrees are greater than the angle degree, and the Euclidean distance is less than or equal to the plane distance threshold, the corresponding two boundary guide vectors are matched with each other, otherwise no operation is performed;

[0064] Extend each mutually matching boundary guide vector until they intersect, record the intersection point as the appearance boundary point, connect the adjacent appearance boundary points in sequence, and then generate the corresponding appearance boundary lines between each cluster;

[0065] A plurality of feature pointers are set, wherein the feature pointers are used to extract feature structures and feature pixel value intervals in real-time regional image data, and each cluster is traversed through each feature pointer, and each feature structure and feature pixel value interval is obtained according to the traversal result, wherein the feature pixel value interval represents the color distribution or color change characteristics of the corresponding image area;

[0066] The same label is set for clusters with the same characteristic structure and characteristic pixel value interval, and the appearance boundary lines of the edge positions of the regions formed by adjacent clusters with the same label are merged to obtain the region boundary lines.

[0067] Further, step S3 is implemented by the following process:

[0068] In the process of each UAV arriving at the terminal position along the UAV flight route, each time a collection and detection cycle ends, a number of operation area images and individual plant images are segmented in each real-time area image data according to the area boundary line and the appearance boundary line;

[0069] A farmland visualization map is established based on images of each operation area and images of individual plants, wherein the farmland visualization map is composed of n operation visualization areas, and each operation visualization area contains a number of real-time image models of plants;

[0070] The crop planting plan records the plant species, initial planting time and management operation steps of each operation area;

[0071] The management operation steps include operation contents and intervals between each operation content, wherein the operation contents include fertilization, sowing, irrigation, topdressing, etc., and image data of corresponding plants before executing the relevant operation contents;

[0072] According to the crop planting plan records and the position distribution of each operation area in the farmland visualization map, a farmland planting chain is established on the farmland visualization map, where the farmland planting chain consists of n operation nodes, and each operation node corresponds to an operation visualization area;

[0073] Based on the image data of each plant before performing the relevant operation content, the characteristic image fragments of the corresponding plant during each operation content are obtained, and according to the initial planting time of each operation area and the interval time between each operation content in the crop planting plan record, an independent initial detection time and several final operation detection times are set for each operation node.

[0074] A number of agricultural machinery equipment are set up, and the agricultural machinery equipment is provided with a Beidou positioning device, and then the real-time position information of each agricultural machinery equipment is obtained according to the Beidou positioning device, and a number of agricultural machinery equipment nodes are set on the farmland visualization map according to the real-time position information of each agricultural machinery equipment, and each agricultural machinery equipment node is marked with a corresponding agricultural machinery equipment number and executable operation content.

[0075] Further, step S4 is implemented by the following process:

[0076] For any operation node, the timing starts from the initial detection time it carries. When a drone starts to collect real-time video data of its corresponding operation area, it is determined whether the current time has reached the next final operation detection time. If not, at the end of each collection and detection cycle, the corresponding real-time image model of the plant in the farmland visualization map is updated according to the real-time regional image data generated in the previous collection and detection cycle.

[0077] Retrieve the corresponding feature image fragments according to the plant species and the current operation content, and then judge whether the plant real-time image model meets the next operation content through the feature image fragments;

[0078] If the judgment is satisfied, the "next operation content can be executed" mark is set for the plant real-time image model, otherwise no operation is performed;

[0079] An executable threshold is set according to the number of plants planted in each operation area. If the current time has not reached the next final operation detection time, but the number of plant real-time image models marked with "next operation content can be executed" is greater than or equal to the executable threshold, a job requirement pointer is generated for the corresponding operation node, and the job requirement pointer includes the name of the operation content, the type of plant planted, and the number;

[0080] If the current time has not reached the next final job detection time, but the number of plant real-time image models marked with "executable next job content" is less than the executable threshold, no operation will be performed until the current time reaches the next final job detection time, and a job demand pointer will be directly generated for the corresponding job node.

[0081] Furthermore, a unit moving speed is set for each agricultural machinery equipment node. Whenever an operation node generates an operation demand pointer, the agricultural machinery equipment node with the corresponding operation content name and in an idle state on the farmland visualization map is matched according to the operation content name in the operation demand pointer;

[0082] The agricultural machinery driving roads are set based on the distribution of the farmland visualization map on the regional boundary line, and several prepared agricultural machinery driving routes are generated with the operation node with the operation demand pointer as the end position and the current position of each matching agricultural machinery equipment node as the starting position, and the agricultural machinery driving route currently being executed is mapped on the agricultural machinery driving road;

[0083] Map all the prepared agricultural machinery driving routes to the agricultural machinery driving roads for driving simulation. If any prepared agricultural machinery driving route collides with the agricultural machinery driving route that is already being executed during the simulated driving process, the corresponding prepared agricultural machinery driving route will be eliminated. Otherwise, the corresponding prepared agricultural machinery driving route will be retained, and the estimated execution time of the retained prepared agricultural machinery driving routes will be calculated, and then the prepared agricultural machinery driving route with the shortest estimated execution time and the corresponding agricultural machinery equipment node will be selected as the agricultural machinery driving route and the executing agricultural machinery equipment node;

[0084] The agricultural machinery driving route is sent to the agricultural machinery equipment corresponding to the agricultural machinery equipment node, and then the agricultural machinery equipment drives according to the received agricultural machinery driving route, and the corresponding agricultural machinery equipment node is marked as working. The above agricultural machinery equipment scheduling process is repeated until the operation content of the crops in all operation areas is completed.

[0085] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent integrated control method for agricultural machinery, characterized in that: The following steps are involved: Step S1, dividing the target farmland into several operation areas, and then setting up several drones to generate real-time regional image data of each operation area; Step S2, according to the pixel difference between each pixel, a number of clusters are set in the real-time regional image data, pixels at the edge position of each cluster are selected to establish a boundary guide vector, and then according to the corresponding relationship between the boundary guide vectors of adjacent clusters, a regional boundary line is divided in the real-time regional image data; Step S3: establishing a farmland visualization map and a farmland planting chain according to the real-time regional image data of each operation area, setting a number of agricultural machinery and equipment, obtaining the real-time position information of each agricultural machinery and equipment according to the Beidou positioning device, and setting a number of agricultural machinery and equipment nodes on the farmland visualization map according to the real-time position information of each agricultural machinery and equipment; Step S4, update the farmland visualization map in real time according to the real-time regional image data of each operation area, generate an operation demand pointer on the farmland planting chain, set the agricultural machinery driving road on the farmland visualization map according to the regional boundary line, and then match multiple agricultural machinery equipment nodes according to the pre-stored operation demand pointer, set a prepared agricultural machinery driving route for each agricultural machinery equipment node according to the matching result, map each prepared agricultural machinery driving route to the agricultural machinery driving road for driving simulation, and then select the execution agricultural machinery equipment node and agricultural machinery driving route and execute them.

2. The intelligent integrated control method for agricultural machinery according to claim 1, characterized in that: The generation process of the real-time regional image data includes: According to the distribution of crop planting in the target farmland, the target farmland is divided into n operation areas of different areas, where n is a natural number greater than 0. The drone is equipped with a camera and a Beidou positioning device. During the flight of the drone, the camera continuously obtains real-time video data of its shooting range; An acquisition detection cycle is set, and each time an acquisition detection cycle starts or ends, a real-time video data segment generated by a previous acquisition detection cycle is intercepted, and the real-time video data segment is divided into several pieces of real-time regional image data by frame.

3. The intelligent integrated control method for agricultural machinery according to claim 2, characterized in that: The process of setting a number of clusters within the real-time area image data includes: Obtain the pixel value of each pixel in each real-time regional image data, establish a two-dimensional coordinate system, and map the real-time regional image data generated by the same drone in each collection and detection cycle onto the two-dimensional coordinate system in chronological order; A cluster pixel difference threshold is set to determine whether the pixel difference between adjacent pixels in the real-time regional image data is less than or equal to the cluster pixel difference threshold. If the pixel difference is less than or equal to the cluster pixel difference threshold, the corresponding two pixels are recorded as the same cluster, otherwise no operation is performed.

4. The intelligent integrated control method for agricultural machinery according to claim 3, characterized in that: The process of establishing the boundary guidance vector includes: Select the pixel at the edge of each cluster as the starting point, and select the pixel with the largest pixel difference from the starting point as the secondary end point, and then establish several boundary guide vectors at the edge of each cluster; Set the angle threshold and plane distance threshold. Since there are common boundaries between adjacent clusters, there must be some boundary guide vectors between adjacent clusters whose direction angles are greater than the angle threshold. Then obtain the angle degree and Euclidean distance between each boundary guide vector in the adjacent cluster. If the two angle degrees are greater than the angle degree and the Euclidean distance is less than or equal to the plane distance threshold, the corresponding two boundary guide vectors are matched with each other, otherwise no operation is performed.

5. The intelligent integrated control method for agricultural machinery according to claim 4, characterized in that: The process of dividing the area boundary line in the real-time area image data includes: Extend each mutually matching boundary guide vector until they intersect, record the intersection point as the appearance boundary point, connect adjacent appearance boundary points in sequence, and then generate corresponding appearance boundary lines between each cluster, and set several feature pointers; Each cluster is traversed through each feature pointer, and each feature structure and feature pixel value interval is obtained according to the traversal results. The same label is set for clusters with the same feature structure and feature pixel value interval, and the appearance boundary lines of the edge positions of the regions formed by adjacent clusters with the same label are merged to obtain the region boundary lines.

6. The intelligent integrated control method for agricultural machinery according to claim 5, characterized in that: The process of establishing the farmland visualization map includes: A plurality of operation area images and plant individual images are segmented from each real-time area image data, and a farmland visualization map is established based on each operation area image and plant individual image, wherein the farmland visualization map is composed of n operation visualization areas, and each operation visualization area contains a plurality of plant real-time image models; The crop planting plan records the plant species, initial planting time and management operation steps of each operation area, wherein the management operation steps include operation contents and interval time between each operation content, wherein the operation content includes the name of the operation content and the image data of the corresponding plant before executing the relevant operation content; According to the crop planting plan records and the position distribution of each operation area in the farmland visualization map, a farmland planting chain is established on the farmland visualization map, where the farmland planting chain consists of n operation nodes. According to the image data of each plant before performing the relevant operation content, the characteristic image fragments of the corresponding plant during each operation content are obtained, and according to the initial planting time of each operation area in the crop planting plan records and the interval time between each operation content, an independent initial detection time and several final operation detection times are set for each operation node.

7. The intelligent integrated control method for agricultural machinery according to claim 6, characterized in that: A number of agricultural machinery equipment are set up, and the agricultural machinery equipment is provided with a Beidou positioning device, and then the real-time position information of each agricultural machinery equipment is obtained according to the Beidou positioning device, and a number of agricultural machinery equipment nodes are set on the farmland visualization map according to the real-time position information of each agricultural machinery equipment, and each agricultural machinery equipment node is marked with a corresponding agricultural machinery equipment number and executable operation content.

8. The intelligent integrated control method for agricultural machinery according to claim 7, characterized in that: The generation process of the job demand pointer includes: For any operation node, the timing starts from the initial detection time it carries. When a drone starts to collect real-time video data of its corresponding operation area, it is determined whether the current time has reached the next final operation detection time. If not, then at the end of each collection and detection cycle, the corresponding real-time image model of the plant in the farmland visualization map is updated according to the real-time area image data generated in the previous collection and detection cycle, and then the real-time image model of the plant is determined through the feature image fragments to determine whether it meets the requirements for the next operation, and an operation requirement pointer is generated according to the judgment result; If the current time has not reached the next final job detection time, but the number of plant real-time image models marked with "next job content can be executed" is less than the executable threshold, no operation will be performed until the current time reaches the next final job detection time, and a job demand pointer will be directly generated for the corresponding job node.

9. The intelligent integrated control method for agricultural machinery according to claim 8, characterized in that: The process of preparing the agricultural machinery driving route for driving simulation includes: Set the unit moving speed for each agricultural machinery equipment node. Whenever an operation node generates an operation demand pointer, match the agricultural machinery equipment node with the corresponding operation content name and in an idle state on the farmland visualization map according to the operation content name in the operation demand pointer; The agricultural machinery driving roads are set based on the distribution of the farmland visualization map on the regional boundary line, and several prepared agricultural machinery driving routes are generated with the operation node with the operation demand pointer as the end position and the current position of each matching agricultural machinery equipment node as the starting position, and the agricultural machinery driving route currently being executed is mapped on the agricultural machinery driving road; All prepared agricultural machinery driving routes are mapped to agricultural machinery driving roads for driving simulation, and agricultural machinery driving routes and agricultural machinery equipment nodes are selected according to the driving simulation results.

Images