Agricultural laser weeding safety decision method and system
Patent Information
- Application Number
- CN202610407657.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-31
- Publication Date
- 2026-08-28
AI Technical Summary
[0003]依赖除草剂喷洒实现杂草灭杀的化学除草，具有除草剂易流失造成的水体污染（全球每年因除草剂流失造成的水体污染超200万吨），同时造成土壤板结、除草剂残留，破坏土壤生态系统的问题；
[0033]The agricultural laser weeding safety decision-making method provided by this invention can achieve high-precision plant identification by using hyperspectral imaging data in conjunction with a pre-trained plant phenotypic classifier. It then combines the plant three-dimensional model constructed based on the plant's three-dimensional point cloud data to generate weed location results and operation path planning results, solving the problems of inaccurate weed identification and location and high false kill rate in traditional technologies. By calculating the minimum lethal energy through an energy optimization algorithm, it can achieve precise control of laser energy, which significantly reduces the energy consumption of laser weeding and thus reduces agricultural operation costs.
Smart Images
Figure CN122642392A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of smart agriculture technology, and in particular to a method and system for safe decision-making in agricultural laser weeding. Background Technology
[0002] In agricultural production, weed control is a crucial step in ensuring crop yields. Existing weeding technologies mainly include chemical weeding, mechanical weeding, and traditional laser weeding, but all have significant technical shortcomings and cannot meet the demands of smart agriculture for precise, green, and safe weed control. Specifically:
[0003] Chemical weed control, which relies on herbicide spraying to kill weeds, has problems such as water pollution caused by the easy loss of herbicides (more than 2 million tons of water are polluted globally each year due to herbicide loss), soil compaction, herbicide residues, and damage to soil ecosystems.
[0004] Mechanical weeding, which removes or uproots weeds by physical means, has problems such as low operational precision (e.g., the measured seedling damage rate in cornfields reaches 18%-25%), serious damage to crop seedlings, and inability to adapt to complex plant types and planting densities.
[0005] While traditional laser weeding achieves chemical-free weed control, it suffers from insufficient positioning accuracy (positioning error of ±3cm), a weed false kill rate of >15%, unreasonable laser energy control, high energy consumption, and unstable weed control effect.
[0006] In summary, existing technologies have not yet solved the multi-objective collaborative control challenges of accurate weed identification, ecological safety during weeding, and laser energy optimization. There is an urgent need for a laser weeding safety decision-making method that can simultaneously improve weeding efficiency, crop protection, and ecological safety. Summary of the Invention
[0007] To address the aforementioned problems, this invention provides an agricultural laser weeding safety decision-making method, along with a corresponding agricultural laser weeding safety decision-making system and operational equipment for implementing the agricultural laser weeding safety decision-making method.
[0008] Therefore, according to a first aspect, the present invention provides a method for making safe decisions regarding agricultural laser weed control, comprising the following steps:
[0009] S100: Acquires hyperspectral imaging data, ecological environment data, and three-dimensional point cloud data of plants in the target weeding area; hyperspectral imaging data includes visible light imaging data, near-infrared imaging data, and thermal imaging data;
[0010] S200: Based on hyperspectral imaging data, weed identification is performed using a pre-trained plant phenotypic classifier;
[0011] S300: Based on hyperspectral imaging data, ecological environment data, and plant 3D point cloud data, a pre-trained energy optimization model is used to calculate the minimum lethal energy of laser for killing weeds, and laser emission parameter results are generated based on the minimum lethal energy. The minimum lethal energy is the product of the laser base lethal energy, weed species coefficient, water content correction coefficient, growth status correction coefficient, and environmental safety coefficient. The minimum lethal energy of laser is not less than the laser base lethal energy.
[0012] S400: Based on the three-dimensional point cloud data of the plant, a three-dimensional model of the plant is constructed, and the weed location result and operation path planning result are generated; the operation equipment positions the laser emitter to the target weed according to the location result and operation path planning result, and controls the laser emitter to emit a laser corresponding to the laser emission parameter result.
[0013] Optionally, the agricultural laser weeding safety decision-making method also includes the following steps:
[0014] S500: Generates real-time temperature of weed leaves and crop leaves in the laser-irradiated area based on real-time acquired hyperspectral imaging data;
[0015] S600: Generates laser power adjustment results for the laser emitter based on the real-time temperature of weed leaves and crop leaves.
[0016] Optionally, the agricultural laser weeding safety decision-making method also includes the following steps:
[0017] S700: Real-time acquisition of laser emitter positioning data, hyperspectral imaging data, and plant 3D point cloud data to obtain real-time comparison results of laser operation deviation; the real-time comparison results of operation deviation refer to the comparison results of the deviation between the target position and the actual position of the laser operation.
[0018] S800: When the real-time comparison result of the operation deviation reaches the preset threshold, a laser operation stop signal is generated.
[0019] Optionally, the agricultural laser weeding safety decision-making method also includes the following steps:
[0020] S900: Acquires the open status monitoring data of the safety shield on the laser emitter, and generates a laser operation stop signal when the open status monitoring data is abnormal.
[0021] Optionally, the ecological environment data includes wind speed data, humidity data, light parameter data, and soil herbicide residue data; the soil herbicide residue data is obtained using a three-electrode soil electrochemical sensor.
[0022] Optionally, step S300 specifically includes the following steps:
[0023] Based on hyperspectral imaging data, environmental data, and plant 3D point cloud data, a pre-trained energy optimization model was used to construct an energy gradient field for weed plants that includes spatial location, tissue type, water content, and growth status.
[0024] The minimum lethal energy of laser at each point in the three-dimensional space of weeds is calculated point by point based on the energy gradient field.
[0025] The optimal laser irradiation path and corresponding laser emission parameters are planned based on the minimum lethal energy of the laser at each point; the laser emission parameters include laser wavelength, laser power and laser irradiation duration.
[0026] According to a second aspect, embodiments of the present invention provide an agricultural laser weeding safety decision-making system, comprising:
[0027] The ecological sensing layer is used to acquire hyperspectral imaging data, ecological environment data, and three-dimensional point cloud data of plants in the target weeding area; the hyperspectral imaging data includes visible light imaging data, near-infrared imaging data, and thermal imaging data;
[0028] A plant phenotypic classification layer is used for weed identification based on hyperspectral imaging data using a pre-trained plant phenotypic classifier.
[0029] The lethal energy calculation layer is used to calculate the minimum lethal energy of laser for killing weeds based on hyperspectral imaging data, environmental data, and three-dimensional point cloud data of plants using a pre-trained energy optimization model, and to generate laser emission parameter results based on the minimum lethal energy. The minimum lethal energy is the product of the laser base lethal energy, weed species coefficient, water content correction coefficient, growth status correction coefficient, and environmental safety coefficient. The minimum lethal energy of laser is not less than the laser base lethal energy.
[0030] The path planning layer is used to construct a three-dimensional model of the plant based on the three-dimensional point cloud data of the plant, and generate weed location results and operation path planning results; the operation equipment positions the laser emitter to the target weed according to the location results and operation path planning results, and controls the laser emitter to emit laser corresponding to the laser emission parameter results.
[0031] According to a third aspect, an embodiment of the present invention provides a working device, including: an ecological sensing module, a walking module, a robotic arm and a laser emitting end disposed on the robotic arm, wherein a control module executes the method described in the first aspect or any embodiment of the first aspect.
[0032] The technical solution provided by this invention has the following advantages:
[0033] The agricultural laser weeding safety decision-making method provided by this invention can achieve high-precision plant identification by using hyperspectral imaging data in conjunction with a pre-trained plant phenotypic classifier. It then combines the plant three-dimensional model constructed based on the plant's three-dimensional point cloud data to generate weed location results and operation path planning results, solving the problems of inaccurate weed identification and location and high false kill rate in traditional technologies. By calculating the minimum lethal energy through an energy optimization algorithm, it can achieve precise control of laser energy, which significantly reduces the energy consumption of laser weeding and thus reduces agricultural operation costs. Attached Figure Description
[0034] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0035] Figure 1 A flowchart of a method for a safety decision-making method for agricultural laser weeding provided in an embodiment of the present invention;
[0036] Figure 2 Another flowchart of the agricultural laser weeding safety decision-making method provided in the embodiments of the present invention;
[0037] Figure 3 This is a schematic diagram of an agricultural laser weeding safety decision-making system provided in an embodiment of the present invention. Detailed Implementation
[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0039] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Example 1
[0040] Figure 1 A flowchart of an agricultural laser weeding safety decision-making method according to an embodiment of the present invention is shown, specifically, as follows: Figure 1 As shown, the method may include the following steps:
[0041] S100: Acquire hyperspectral imaging data, ecological environment data, and three-dimensional point cloud data of plants in the target weeding area.
[0042] In this embodiment, the hyperspectral imaging data includes visible light imaging data, near-infrared imaging data, and thermal imaging data.
[0043] Specifically, the hyperspectral imaging data was acquired using a hyperspectral imager that integrates an infrared thermal imaging channel; the three-dimensional point cloud data of the plant was obtained based on a three-dimensional lidar.
[0044] In this embodiment, the ecological environment data can include wind speed data, humidity data, light parameter data, and soil herbicide residue data. The wind speed data, humidity data, and light parameter data can be collected using a micro-meteorological sensor, while the soil herbicide residue data is detected using a three-electrode soil electrochemical sensor.
[0045] S200: Based on hyperspectral imaging data, weed identification is performed using a pre-trained plant phenotypic classifier.
[0046] S300: Based on hyperspectral imaging data, environmental data, and three-dimensional point cloud data of plants, a pre-trained energy optimization model is used to calculate the minimum lethal energy of laser for killing weeds, and laser emission parameter results are generated based on the minimum lethal energy of laser.
[0047] In this embodiment, the minimum lethal energy is the product of the laser base lethal energy, the weed species coefficient, the water content correction coefficient, the growth state correction coefficient, and the environmental safety factor. The minimum lethal energy of the laser is not less than the laser base lethal energy. It should be noted that the base lethal energy is the lowest energy density threshold at a laser wavelength (e.g., 980nm) that can permanently inactivate and prevent the regeneration of weed target tissue; if the energy is below this threshold, effective killing cannot be achieved, and only surface burns will occur.
[0048] The minimum lethal energy of this invention is not lower than the basic lethal energy. Through dynamic matching of laser power and action time, the power is reduced and the action time is optimized while ensuring that the total energy reaches the lethal threshold, so as to achieve an equivalent killing effect. At the same time, excessive energy loss and crop thermal damage are avoided, and the invention strictly complies with the ISO 11553 laser safety requirements standard.
[0049] In specific implementation, step S300 can be set to include the following steps:
[0050] S310: Based on hyperspectral imaging data, ecological environment data, and plant 3D point cloud data, a pre-trained energy optimization model is used to construct an energy gradient field for weed plants that includes spatial location, tissue type, water content, and growth status.
[0051] Specifically, water content can be obtained from hyperspectral imaging data acquired by a hyperspectral imager (400–2500 nm). Specifically, image data of vegetation water-sensitive bands (970 nm, 1200 nm, 1450 nm, 1940 nm) are extracted, and the relative water content of leaves / whole plants is retrieved by inverting the characteristics of water absorption peaks.
[0052] In practice, a temperature-water content relationship curve can be constructed using thermal infrared images (the lower the plant water content, the higher the leaf temperature), and this curve can be used to further correct the above water content results.
[0053] Specifically, growth status includes parameters such as the plant's growth stage, degree of lignification, and robustness; growth status data can be obtained from the output of the plant phenotypic classifier, hyperspectral imaging data, and plant 3D point cloud data.
[0054] S320: Calculate the minimum lethal laser energy at each point in the three-dimensional space of weeds based on the energy gradient field.
[0055] S330: Based on the minimum lethal energy of the laser at each point, the optimal laser irradiation path and the corresponding laser emission parameters are planned.
[0056] In this embodiment, the laser emission parameters include laser wavelength, laser power, and laser irradiation duration.
[0057] S400: Based on the three-dimensional point cloud data of the plants, a three-dimensional model of the plants is constructed, and the weed location results and operation path planning results are generated.
[0058] In this embodiment, the operating equipment positions the laser emitter to the target weeds based on the positioning results and the operating path planning results, and controls the laser emitter to emit lasers corresponding to the laser emission parameter results.
[0059] As an optional implementation of this embodiment, in order to further improve the weed control effect of the method in this embodiment and reduce the possibility of crop damage, such as Figure 2 As shown, the method in this implementation may also include the following steps:
[0060] S500: Generates real-time temperature of weed leaves and crop leaves in the laser-irradiated area based on real-time acquired hyperspectral imaging data.
[0061] S600: Generates laser power adjustment results for the laser emitter based on the real-time temperature of weed leaves and crop leaves.
[0062] Specifically, several leaf temperature thresholds can be set, and laser power adjustment results can be generated by comparing the real-time temperatures of grass leaves and crop leaves with these temperature thresholds. For example, a lethal temperature range (including an upper and lower threshold) and a safe temperature threshold can be set. When the real-time temperature of weed leaves is lower than the lower threshold of the lethal temperature range, the laser power is increased; when the real-time temperature of weed leaves is within the lethal temperature range, the power is maintained; when both the real-time temperatures of weed leaves and crop leaves are higher than the safe temperature threshold, a laser operation stop signal is generated.
[0063] As an optional implementation of this embodiment, in order to further improve the accuracy of weed control in the method of this embodiment, such as Figure 2 As shown, the method in this implementation may also include the following steps:
[0064] S700: Real-time acquisition of laser emitter positioning data, hyperspectral imaging data, and plant 3D point cloud data to obtain real-time comparison results of laser operation deviations.
[0065] In this embodiment, the real-time comparison result of the operation deviation refers to the comparison result of the deviation between the target position and the actual position of the laser operation.
[0066] S800: When the real-time comparison result of the operation deviation reaches the preset threshold, a laser operation stop signal is generated.
[0067] Specifically, a preset threshold can be set to be between ±(0.5~1) mm.
[0068] As an optional implementation method of this embodiment, in order to further improve the implementation security of the method in this embodiment, such as... Figure 2 As shown, the method in this embodiment may further include the following steps:
[0069] S900: Acquires the open status monitoring data of the safety shield on the laser emitter, and generates a laser operation stop signal when the open status monitoring data is abnormal.
[0070] Specifically, the open status monitoring data can be obtained by the protective cover status sensor equipped with the safety cover. The protective cover status sensor can be a micro switch, Hall sensor or photoelectric position sensor, etc.
[0071] This invention also provides an agricultural laser weeding safety decision-making system, which is used to implement the above-mentioned agricultural laser weeding safety decision-making method. For example... Figure 3 As shown, the system includes: an ecological perception layer 100, a plant phenotypic classification layer 200, a lethal energy calculation layer 300, and a path planning layer 400.
[0072] Among them, the ecological sensing layer 100 is used to acquire hyperspectral imaging data, ecological environment data and plant three-dimensional point cloud data of the weeding target area; the hyperspectral imaging data includes visible light imaging data, near-infrared imaging data and thermal imaging data.
[0073] Among them, the plant phenotypic classification layer 200 is used to identify weeds based on hyperspectral imaging data using a pre-trained plant phenotypic classifier.
[0074] The lethal energy calculation layer 300 is used to calculate the minimum lethal energy of laser for killing weeds based on hyperspectral imaging data, environmental data, and three-dimensional point cloud data of plants using a pre-trained energy optimization model, and to generate laser emission parameter results based on the minimum lethal energy of laser. The minimum lethal energy is the product of the laser base lethal energy, weed species coefficient, water content correction coefficient, growth status correction coefficient, and environmental safety coefficient. The minimum lethal energy of laser is not less than the laser base lethal energy.
[0075] The path planning layer 400 is used to construct a three-dimensional model of the plant based on the three-dimensional point cloud data of the plant, and generate weed positioning results and operation path planning results. The operation equipment positions the laser emitter to the target weed according to the positioning results and operation path planning results, and controls the laser emitter to emit lasers corresponding to the laser emission parameter results.
[0076] The agricultural laser weeding safety decision-making method and system in this embodiment can achieve high-precision plant identification by using hyperspectral imaging data in conjunction with a pre-trained plant phenotypic classifier. Then, it generates weed location results and operation path planning results by combining the plant three-dimensional model constructed based on the plant's three-dimensional point cloud data, which solves the problems of inaccurate weed identification and location and high false kill rate of traditional technology. By calculating the minimum lethal energy through energy optimization algorithm, it can achieve precise control of laser energy, which greatly reduces the energy consumption of laser weeding and thus reduces agricultural operation costs. Example 2
[0077] This invention also provides an operating device, which includes an ecological sensing module, a walking module, a robotic arm, and a laser emitting end disposed on the robotic arm. The control module executes the agricultural laser weeding safety decision-making method described in any one of the embodiments of Embodiment 1 above, or includes the agricultural laser weeding safety decision-making device described in any one of the embodiments of Embodiment 1 above.
[0078] Those skilled in the art should understand that the ecological sensing module is a data acquisition device corresponding to step S100 or the ecological sensing layer; specifically, the ecological sensing module may include a micro-meteorological sensor, a soil electrochemical sensor, a hyperspectral imager with an infrared thermal imaging channel, and a three-dimensional lidar.
[0079] Those skilled in the art should understand that the structural connection relationships between the above modules are unrelated to the technical solution in this embodiment, and this embodiment does not impose any limitations.
[0080] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.
[0081] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.
Claims
1. A safety decision-making method for agricultural laser weeding, characterized in that, Includes the following steps: S100: Acquire hyperspectral imaging data, ecological environment data, and three-dimensional point cloud data of plants in the target area for weeding; the hyperspectral imaging data includes visible light imaging data, near-infrared imaging data, and thermal imaging data; S200: Based on the hyperspectral imaging data, weed identification is performed using a pre-trained plant phenotypic classifier; S300: Based on the hyperspectral imaging data, the ecological environment data, and the plant 3D point cloud data, a pre-trained energy optimization model is used to calculate the minimum lethal energy of laser for killing weeds, and laser emission parameter results are generated according to the minimum lethal energy; the minimum lethal energy is the product of the laser base lethal energy, weed species coefficient, water content correction coefficient, growth state correction coefficient, and environmental safety coefficient, and the minimum lethal energy of laser is not less than the laser base lethal energy; S400: Based on the plant's three-dimensional point cloud data, construct a three-dimensional model of the plant, and generate weed location results and operation path planning results by combining the weed identification results; The operating equipment positions the laser emitter to the target weeds based on the positioning results and the operating path planning results, and controls the laser emitter to emit a laser corresponding to the laser emission parameter results.
2. The agricultural laser weeding safety decision-making method according to claim 1, characterized in that, It also includes the following steps: S500: Based on the real-time acquired hyperspectral imaging data, generate the real-time temperature of weed leaves and crop leaves in the laser-irradiated area; S600: Based on the real-time temperature of the weed leaves and the real-time temperature of the crop leaves, generate the laser power adjustment result of the laser emitter.
3. The agricultural laser weeding safety decision-making method according to claim 1 or 2, characterized in that, It also includes the following steps: S700: Real-time acquisition of the laser emitter positioning data, the hyperspectral imaging data, and the plant three-dimensional point cloud data to obtain the real-time comparison result of the laser operation deviation; the real-time comparison result of the operation deviation refers to the comparison result of the deviation between the target position and the actual position of the laser operation. S800: When the real-time comparison result of the operation deviation reaches the preset threshold, a laser operation stop signal is generated.
4. The agricultural laser weeding safety decision-making method according to any one of claims 1-3, characterized in that, It also includes the following steps: S900: Acquire the opening status monitoring data of the safety shield on the laser emitter, and generate a laser operation stop signal when the opening status monitoring data is abnormal status monitoring data.
5. The agricultural laser weeding safety decision-making method according to claim 1, characterized in that, The ecological environment data includes wind speed data, humidity data, light parameter data, and soil herbicide residue data; the soil herbicide residue data is obtained using a three-electrode soil electrochemical sensor.
6. The agricultural laser weeding safety decision-making method according to claim 1, characterized in that, Step S300 specifically includes the following steps: Based on the hyperspectral imaging data, the environmental data, and the plant 3D point cloud data, a pre-trained energy optimization model is used to construct an energy gradient field for weed plants that includes spatial location, tissue type, water content, and growth status. The minimum lethal energy of laser at each point in the three-dimensional space of the weeds is calculated point by point based on the energy gradient field. Based on the minimum lethal energy of the laser at each location, the optimal laser irradiation path and the corresponding laser emission parameters are planned; the laser emission parameters include laser wavelength, laser power, and laser irradiation duration.
7. An agricultural laser weeding safety decision-making system, characterized in that, include: The ecological sensing layer is used to acquire hyperspectral imaging data, ecological environment data, and three-dimensional point cloud data of plants in the target weeding area. The hyperspectral imaging data includes visible light imaging data, near-infrared imaging data, and thermal imaging data; A plant phenotypic classification layer is used to identify weeds based on the hyperspectral imaging data using a pre-trained plant phenotypic classifier. The lethal energy calculation layer is used to calculate the minimum lethal energy of laser for killing weeds based on the hyperspectral imaging data, the ecological environment data, and the plant three-dimensional point cloud data, using a pre-trained energy optimization model, and to generate laser emission parameter results based on the minimum lethal energy. The minimum lethal energy is the product of the laser base lethal energy, weed species coefficient, water content correction coefficient, growth state correction coefficient, and environmental safety coefficient, and the minimum lethal energy is not less than the laser base lethal energy. The path planning layer is used to construct a three-dimensional model of the plant based on the three-dimensional point cloud data of the plant, and generate weed location results and operation path planning results. The operating equipment positions the laser emitter to the target weeds based on the positioning results and the operating path planning results, and controls the laser emitter to emit a laser corresponding to the laser emission parameter results.
8. A working device, characterized in that, include: An ecological sensing module, a walking module, a robotic arm, and a laser emitting end mounted on the robotic arm; the control module executes the method described in any one of claims 1-7.