An intelligent inter-row weeding system for orchards based on multi-sensor information fusion

Abstract

The present invention relates to the field of intelligent weeding technology, and discloses an orchard intelligent inter-row weeding system with multi-sensor information fusion, including a sensor data module and an intelligent control module. The system obtains image data of all paths and operation data of all time points through the sensor data module, and classifies and forms a data set. The intelligent control module establishes a two-dimensional coordinate system in equal proportion, analyzes the inter-row distance of each fruit tree, and then performs grid processing according to the two-dimensional coordinate system to calculate the trunk diameter of each fruit tree, grasps the growth status of all fruit trees in real time, analyzes the operation trajectory of the weeding robot, and accurately plans the operation trajectory with strong applicability. The intelligent control module analyzes the coverage rate of the weeding operation, judges the growth status of the fruit tree and the growth status of the weeds, and if the expected duration remains unchanged within the monitoring period, when the coverage rate decreases, the average cutting speed decreases, indicating that the weed growth state is vigorous, and the weeding operation frequency needs to be increased, and the weeding effect is accurately evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent weeding, and in particular to an orchard intelligent inter-row weeding system with multi-sensor information fusion. Background Art

[0002] Orchard interplant weeds refer to non-target plants that grow in the open space between trees in the orchard. There is a strong competition for water, nutrients and space between these weeds and fruit trees, resulting in a 10% to 15% reduction in fruit tree yields. In severe cases, they can completely occupy the growth space of fruit trees. Weeds with taller plants will affect the sunlight exposure of fruit trees, reduce the photosynthesis intensity of fruit tree leaves, and thus affect the normal growth and development, flowering and fruiting, and fruit quality of fruit trees. Too many weeds in the orchard will lead to a closed environment, increased air humidity, and a large number of harmful pathogens, affecting the normal growth and development of fruit trees. At the same time, weeds are also an important place for most pathogens to overwinter. Methods for preventing and controlling weeds in the field include the use of chemical agents, weed-proof cloth, organic matter covering, and planting green manure. Specifically, chemical control can be carried out by spraying lethal or stem and leaf herbicides. At the same time, cover with horticultural cloth with good air permeability, which can inhibit the growth of weeds without affecting the normal growth of plants. In addition, using organic matter such as straw and hay to cover the spaces between fruit trees or around tree trays can not only increase the organic matter content in the soil, but also improve the soil structure. Planting leguminous green manure crops such as alfalfa and broad beans between fruit tree rows can not only increase soil organic matter, but also improve the overall economic benefits of the orchard.

[0003] At present, traditional orchard weeding systems lack real-time monitoring capabilities and find it difficult to accurately locate the position of fruit trees and the diameter of trunks. In actual use, the weeder blades can easily accidentally damage the trunks of fruit trees, causing economic losses. They have poor flexibility and applicability. Traditional orchard weeding systems can only deal with short-term weed growth problems and are unable to periodically evaluate weed conditions and provide timely and effective management strategies. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an orchard intelligent inter-row weeding system with multi-sensor information fusion, which has the advantages of precise planning of operation trajectories, strong applicability, precise evaluation of good weeding effect, etc., and solves the problem that traditional orchard weeding systems have poor flexibility and applicability and cannot provide management strategies in a timely manner.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An orchard intelligent inter-row weeding system with multi-sensor information fusion, characterized by comprising a sensor data module and an intelligent control module;

[0007] The sensor data module is composed of an image data unit and a mechanical data unit. The image data unit is connected to a panoramic camera through a network to collect an image data set, and the image data set includes image data of all paths between orchard plants. The mechanical data unit is connected to a weeding robot through a network to collect a mechanical data set, and the mechanical data set includes operation data of the weeding robot at all time points. The sensor data module transmits the image data set and the mechanical data set to the intelligent control module through the network.

[0008] The intelligent control module is composed of a positioning analysis unit, a path analysis unit and a weeding management unit. The positioning analysis unit establishes a two-dimensional coordinate system in proportion to the image data set. , analyze the distance between each fruit tree and generate the corresponding distance between plants , and transmitted to the path analysis unit through the network. The path analysis unit is based on the two-dimensional coordinate system Perform grid processing to generate The grid is combined with the inter-plant data set , calculate the trunk diameter of each fruit tree , analyze the trajectory of the weeding robot and transmitted to the weeding management unit via the network, wherein the weeding management unit analyzes the coverage rate of the weeding operation based on the mechanical data set. The weed control unit is provided with a monitoring cycle of fixed duration. , combined with the distance between plants , trunk diameter and coverage , judge the growth status of fruit trees and weeds, and output the corresponding judgment results.

[0009] Preferably, the expression of the image data set is , to They are the first to the second rows between the orchard plants. The image data of the path, Indicates the time point for obtaining image data of a single path.

[0010] Preferably, the expression of the mechanical data set is , to They are the first time point to the second time point of the weeding robot. The operation data at each time point includes the moving speed and cutting speed. Indicates the time point at which the running data is obtained.

[0011] Preferably, the two-dimensional coordinate system The establishment process is as follows:

[0012] According to the image dataset, the length of all paths from west to east is counted and marked as , to Articles 1 to 2 The length of the paths from west to east is counted, and the length of all the paths from south to north is counted and marked as , to Articles 1 to 2 The length of the path from south to north;

[0013] according to to , filter out the longest east-west path;

[0014] according to to , select the longest north-south path;

[0015] If the longest east-west path intersects with the longest north-south path, a two-dimensional coordinate system is established with the intersection of the two paths as the origin and in equal proportion. ;

[0016] If the longest east-west path does not intersect the longest north-south path, then to as well as to , select any intersecting east-west paths and north-south paths, and establish a two-dimensional coordinate system with the intersection of the intersecting paths as the origin. .

[0017] Preferably, the distance between the plants The calculation process is as follows:

[0018] Extract the image data set The image data of the paths are marked as ;

[0019] Through image processing technology, recognition All fruit trees in the two-dimensional coordinate system Mark the coordinate points of the outer diameter of all fruit trees, where: The coordinate points of the outer diameter of the fruit tree are , The coordinate points of the outer diameter of the fruit tree are ;

[0020]

[0021] In the formula, represents the distance between plants, According to the Euclidean distance calculation formula, we get The outer diameter of the fruit tree The straight-line distance between the outer diameters of fruit trees.

[0022] Preferably, the trunk diameter The calculation formula is as follows:

[0023]

[0024] In the formula, represents the trunk diameter, represents the standard distance of a single grid, represents the standard distance minus The outer diameter of the fruit tree The straight-line distance between the outer diameters of the fruit trees is the difference between The trunk diameter of a fruit tree.

[0025] Preferably, the running track The calculation formula is as follows:

[0026]

[0027] In the formula, Indicates the running track. Indicates the safe distance between the blade and the tree trunk. Indicates based on The trunk diameter and safety distance of the fruit tree, and the calculated safety semi-circumference, It represents the sum of the distance between plants and the safety semi-circumference, which is the distance around the weeding robot. The running track of fruit trees during operation.

[0028] Preferably, the coverage The calculation process is as follows:

[0029] According to the mechanical dataset, the moving speed of the weeding robot at all time points is marked as , to They are the first time point to the second time point of the weeding robot. The moving speed at each time point is marked as , to They are the first time point to the second time point of the weeding robot. Cutting speed at each time point;

[0030]

[0031]

[0032]

[0033]

[0034] In the formula, represents the coverage rate, Indicates that the average moving speed of the weeding robot is obtained according to the average calculation formula , Indicates that the average cutting speed of the weeding robot is obtained according to the average calculation formula , It represents the total length of the running track of the weeding robot when it is operating around all the fruit trees. Indicates the estimated duration obtained by dividing the total length of the running track by the average moving speed , It means the estimated time multiplied by the average cutting speed to get the weeding area. It represents the percentage of weeding area to grid area, which is the coverage rate.

[0035] Preferably, if the monitoring period Internal, interplant distance Gradually shortening and trunk diameter When it grows gradually, it means the fruit tree is in good growth condition.

[0036] Preferably, if the monitoring period Estimated duration If the coverage rate remains unchanged, When it decreases, it means that weeds are growing vigorously and the frequency of weeding operations needs to be increased.

[0037] Compared with the prior art, the present invention provides an orchard intelligent inter-row weeding system with multi-sensor information fusion, which has the following beneficial effects:

[0038] 1. The present invention connects the panoramic camera and the weeding robot through the sensor data module network, obtains the image data of all paths between orchard plants and the operation data of the weeding robot at all time points, and classifies them into image data sets and mechanical data sets. The intelligent control module establishes a two-dimensional coordinate system in proportion to the image data sets. , which helps to accurately locate the location coordinates of each fruit tree, analyze the distance between each fruit tree, and generate the corresponding distance between trees. , the intelligent control module is based on the two-dimensional coordinate system Perform grid processing to generate The grid is combined with the inter-plant data set , calculate the trunk diameter of each fruit tree , real-time grasp of the growth status of all fruit trees and analysis of the running trajectory of the weeding robot , accurate planning of operation trajectory has strong applicability.

[0039] 2. The present invention uses an intelligent control module to analyze the coverage rate of weeding operations based on mechanical data sets. , set a fixed-length monitoring cycle , combined with the distance between plants , trunk diameter and coverage , judge the growth status of fruit trees and weeds, if the monitoring cycle Internal, interplant distance Gradually shortening and trunk diameter When it increases gradually, it means that the fruit tree is growing well. Estimated duration If the coverage rate remains unchanged, When reduced, the average cutting speed A slight decrease indicates that weeds are growing vigorously and it is necessary to increase the frequency of weeding operations and accurately evaluate the weeding effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the system flow of the present invention;

[0041] Figure 2 The figure is a schematic diagram of the running trajectory of the weeding robot of the present invention. DETAILED DESCRIPTION

[0042] 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.

[0043] Because traditional orchard weeding systems lack real-time monitoring capabilities, it is difficult to accurately locate the position of fruit trees and the diameter of tree trunks. In actual use, the weeder blades are prone to accidentally damage the trunks of fruit trees, resulting in economic losses. The flexibility and applicability are poor. Traditional orchard weeding systems can only deal with short-term weed growth problems, and cannot periodically evaluate weed conditions and provide timely and effective management strategies. Therefore, a multi-sensor information fusion orchard intelligent inter-row weeding system is provided. Please refer to Figure 1-Figure 2 ,The system includes a sensor data module and an intelligent control module;

[0044] The sensor data module consists of an image data unit and a mechanical data unit. The image data unit collects image data sets through a network connection with a panoramic camera. The image data sets include image data of all paths between orchard plants. The expression of the image data set is: , to They are the first to the second rows between the orchard plants. The image data of the path, Indicates the time point at which the image data of a single path is obtained. In actual use, orchards usually plant fruit trees at equal distances, and regularly arranged paths are set between each row and column of fruit trees. The panoramic camera collects image data in a single direction along the path, which helps to quickly and accurately locate the position of the fruit trees and the trunk diameter.

[0045] The mechanical data unit collects the mechanical data set through the network connection to the weeding robot. The mechanical data set includes the operation data of the weeding robot at all time points. The expression of the mechanical data set is: , to They are the first time point to the second time point of the weeding robot. The operation data at each time point includes the moving speed and cutting speed. Indicates the time point of obtaining the operation data. The weeding robot includes an agricultural machine and a weeder. The agricultural machine drives the weeder to weed in the orchard. The moving speed is the driving speed of the agricultural machine, measured in m / min, and the cutting speed is the cutting speed of the weeder, measured in m² / min;

[0046] The sensor data module transmits the image data set and the mechanical data set to the intelligent control module through the network;

[0047] The intelligent control module consists of a positioning analysis unit, a path analysis unit and a weeding management unit. The positioning analysis unit establishes a two-dimensional coordinate system in proportion to the image data set. , and its establishment process is as follows:

[0048] According to the image dataset, the length of all paths from west to east is counted and marked as , to Articles 1 to 2 The length of the paths from west to east is counted, and the length of all the paths from south to north is counted and marked as , to Articles 1 to 2 The length of the path from south to north;

[0049] according to to , filter out the longest east-west path;

[0050] according to to , select the longest north-south path;

[0051] In an orchard with a square geometry, the cross path in the center is usually used as the main passage. The longest east-west path intersects with the longest north-south path. The intersection of the two paths is used as the origin, and a two-dimensional coordinate system is established in equal proportion. ;

[0052] In an orchard with irregular edges, the longest east-west path does not intersect with the longest north-south path. to as well as to , filter out any intersecting east-west paths and north-south paths, usually using the intersecting paths at the edge of the orchard as a reference, and using the intersection of the intersecting paths as the origin to establish a two-dimensional coordinate system in equal proportion , which helps to accurately locate the location coordinates of each fruit tree;

[0053] Analyze the distance between each fruit tree and generate the corresponding distance between plants , and transmitted to the path analysis unit through the network. The calculation process is as follows:

[0054] Extract the image data set The image data of the paths are marked as ;

[0055] Through image processing technology, recognition All fruit trees in the two-dimensional coordinate system Mark the coordinate points of the outer diameter of all fruit trees, where: The coordinate points of the outer diameter of the fruit tree are , The coordinate points of the outer diameter of the fruit tree are ;

[0056]

[0057] In the formula, represents the distance between plants, According to the Euclidean distance calculation formula, we get The outer diameter of the fruit tree The straight-line distance between the outer diameters of the fruit trees;

[0058] The path analysis unit is based on a two-dimensional coordinate system Perform grid processing to generate The grid is combined with the inter-plant data set , calculate the trunk diameter of each fruit tree , analyze the trajectory of the weeding robot , and transmitted to the weed control management unit through the network. The calculation formula is as follows:

[0059]

[0060] In the formula, represents the trunk diameter, represents the standard distance of a single grid, represents the standard distance minus The outer diameter of the fruit tree The straight-line distance between the outer diameters of the fruit trees is the difference between The trunk diameter of fruit trees, real-time understanding of the growth status of all fruit trees;

[0061]

[0062] In the formula, Indicates the running track. Indicates the safe distance between the blade and the tree trunk. Indicates based on The trunk diameter and safety distance of the fruit tree, and the calculated safety semi-circumference, It represents the sum of the distance between plants and the safety semi-circumference, which is the distance around the weeding robot. The running trajectory of fruit trees during operation, with precise planning of running trajectory and strong applicability;

[0063] The weed control management unit analyzes the coverage of weed control operations based on the machinery data set , the calculation process is as follows:

[0064] According to the mechanical dataset, the moving speed of the weeding robot at all time points is marked as , to They are the first time point to the second time point of the weeding robot. The moving speed at each time point is marked as , to They are the first time point to the second time point of the weeding robot. Cutting speed at each time point;

[0065]

[0066]

[0067]

[0068]

[0069] In the formula, represents the coverage rate, Indicates that the average moving speed of the weeding robot is obtained according to the average calculation formula , measured in m / min, Indicates that the average cutting speed of the weeding robot is obtained according to the average calculation formula , measured in m² / min, It represents the total length of the running track of the weeding robot when it is operating around all the fruit trees. Indicates the estimated duration obtained by dividing the total length of the running track by the average moving speed , It means the estimated time multiplied by the average cutting speed to get the weeding area. It represents the percentage of weeding area to grid area, which is coverage rate;

[0070] The weed control unit is set with a fixed monitoring period , combined with the distance between plants , trunk diameter and coverage , judge the growth status of fruit trees and weeds, if the monitoring cycle Internal, interplant distance Gradually shortening and trunk diameter When it increases gradually, it means that the fruit tree is growing well. Estimated duration If the coverage rate remains unchanged, When reduced, the average cutting speed A slight decrease indicates that weeds are growing vigorously and it is necessary to increase the frequency of weeding operations and accurately evaluate the weeding effect.

[0071] Example 1: In this experiment, the outer diameter coordinate is selected as and the outer diameter coordinates are Two fruit trees were used as experimental objects. After testing, the standard distance of a single grid in the orchard area was 1.2 meters, and the distance between the two fruit trees was The calculation formula is as follows:

[0072]

[0073] In the formula, represents the distance between plants, According to the Euclidean distance calculation formula, the straight-line distance between the outer diameters of the two fruit trees is rice;

[0074] The outer diameter coordinates are Trunk diameter of fruit trees , the calculation formula is as follows:

[0075]

[0076] In the formula, represents the trunk diameter, represents the standard distance of a single grid, It means the standard distance minus the straight-line distance between the outer diameters of the two fruit trees. That is, the outer diameter coordinate is The trunk diameter of a fruit tree.

[0077] Example 2: In this experiment, a fruit tree with a trunk diameter of 0.2 meters was selected as the experimental object. After testing, the safe distance between the blade and the trunk was 0.1 meters. The distance between the fruit tree and the adjacent fruit trees was 1.5 meters. The corresponding running trajectory of the fruit tree , the calculation formula is as follows:

[0078]

[0079] In the formula, Indicates the running track. It represents the safe semi-circumference calculated based on the trunk diameter and the safe distance. It represents the sum of the distance between plants and the safety semi-circumference. That is the running trajectory of the weeding robot when it operates around the fruit tree.

[0080] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An orchard intelligent inter-row weeding system with multi-sensor information fusion, characterized by: Including sensor data module and intelligent control module; The sensor data module is composed of an image data unit and a mechanical data unit. The image data unit is connected to a panoramic camera through a network to collect an image data set, and the image data set includes image data of all paths between orchard plants. The mechanical data unit is connected to a weeding robot through a network to collect a mechanical data set, and the mechanical data set includes operation data of the weeding robot at all time points. The sensor data module transmits the image data set and the mechanical data set to the intelligent control module through the network. The intelligent control module is composed of a positioning analysis unit, a path analysis unit and a weeding management unit. The positioning analysis unit establishes a two-dimensional coordinate system in proportion to the image data set. , analyze the distance between each fruit tree and generate the corresponding distance between plants , and transmitted to the path analysis unit through the network. The path analysis unit is based on the two-dimensional coordinate system Perform grid processing to generate The grid is combined with the inter-plant data set , calculate the trunk diameter of each fruit tree , analyze the trajectory of the weeding robot and transmitted to the weeding management unit via the network, wherein the weeding management unit analyzes the coverage rate of the weeding operation based on the mechanical data set. The weed control unit is provided with a monitoring cycle of fixed duration. , combined with the distance between plants , trunk diameter and coverage , judge the growth status of fruit trees and weeds, and output the corresponding judgment results.

2. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 1, characterized in that: The expression of the image dataset is: , to They are the first to the second rows between the orchard plants. The image data of the path, Indicates the time point for obtaining image data of a single path.

3. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 2, characterized in that: The expression of the mechanical data set is , to The weeding robot is from the first time point to the The operation data at each time point includes the moving speed and cutting speed. Indicates the time point at which the running data is obtained.

4. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 3, characterized in that: The two-dimensional coordinate system The establishment process is as follows: According to the image dataset, the length of all paths from west to east is counted and marked as , to Articles 1 to 2 The length of the paths from west to east is counted, and the length of all the paths from south to north is counted and marked as , to Articles 1 to 2 The length of the path from south to north; according to to , filter out the longest east-west path; according to to , select the longest north-south path; If the longest east-west path intersects with the longest north-south path, a two-dimensional coordinate system is established with the intersection of the two paths as the origin and in equal proportion. ; If the longest east-west path does not intersect the longest north-south path, then to as well as to , select any intersecting east-west paths and north-south paths, and establish a two-dimensional coordinate system with the intersection of the intersecting paths as the origin. .

5. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 4, characterized in that: The distance between plants The calculation process is as follows: Extract the image data set The image data of the paths are marked as ; Through image processing technology, recognition All fruit trees in the two-dimensional coordinate system Mark the coordinate points of the outer diameter of all fruit trees, where: The coordinate points of the outer diameter of the fruit tree are , The coordinate points of the outer diameter of the fruit tree are ; In the formula, represents the distance between plants, According to the Euclidean distance calculation formula, we get The outer diameter of the fruit tree The straight-line distance between the outer diameters of fruit trees.

6. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 5, characterized in that: The trunk diameter The calculation formula is as follows: In the formula, represents the trunk diameter, represents the standard distance of a single grid, represents the standard distance minus The outer diameter of the fruit tree The straight-line distance between the outer diameters of the fruit trees is the difference between The trunk diameter of a fruit tree.

7. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 6, characterized in that: The running track The calculation formula is as follows: In the formula, Indicates the running track. Indicates the safe distance between the blade and the tree trunk. Indicates based on The trunk diameter and safety distance of the fruit tree, and the calculated safety semi-circumference, It represents the sum of the distance between plants and the safety semi-circumference, which is the distance around the weeding robot. The running track of fruit trees during operation.

8. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 7, characterized in that: The coverage The calculation process is as follows: According to the mechanical dataset, the moving speed of the weeding robot at all time points is marked as , to The weeding robot is from the first time point to the The moving speed at each time point is marked as , to The weeding robot is from the first time point to the Cutting speed at each time point; In the formula, represents the coverage rate, Indicates that the average moving speed of the weeding robot is obtained according to the average calculation formula , Indicates that the average cutting speed of the weeding robot is obtained according to the average calculation formula , It represents the total length of the running track of the weeding robot when it is operating around all the fruit trees. Indicates the estimated duration obtained by dividing the total length of the running track by the average moving speed , It means the estimated time multiplied by the average cutting speed to get the weeding area. It represents the percentage of weeding area to grid area, which is the coverage rate.

9. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 8, characterized in that: If the monitoring cycle Internal, interplant distance Gradually shortening and trunk diameter When it grows gradually, it means the fruit tree is in good growth condition.

10. The orchard intelligent inter-row weeding system based on multi-sensor information fusion according to claim 9, characterized in that: If the monitoring cycle Estimated duration If the coverage rate remains unchanged, When it decreases, it means that weeds are growing vigorously and the frequency of weeding operations needs to be increased.

Images