CN110956132B - Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence - Google Patents

Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence Download PDF

Info

Publication number
CN110956132B
CN110956132B CN201911198341.3A CN201911198341A CN110956132B CN 110956132 B CN110956132 B CN 110956132B CN 201911198341 A CN201911198341 A CN 201911198341A CN 110956132 B CN110956132 B CN 110956132B
Authority
CN
China
Prior art keywords
aerial vehicle
unmanned aerial
edge computing
weed
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201911198341.3A
Other languages
Chinese (zh)
Other versions
CN110956132A (en
Inventor
施连敏
张满
侯亮
冯蓉珍
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou Institute of Trade and Commerce
Original Assignee
Suzhou Institute of Trade and Commerce
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Suzhou Institute of Trade and Commerce filed Critical Suzhou Institute of Trade and Commerce
Priority to CN201911198341.3A priority Critical patent/CN110956132B/en
Publication of CN110956132A publication Critical patent/CN110956132A/en
Application granted granted Critical
Publication of CN110956132B publication Critical patent/CN110956132B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/10Terrestrial scenes
    • G06V20/188Vegetation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/241Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02Agriculture; Fishing; Forestry; Mining
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/25Determination of region of interest [ROI] or a volume of interest [VOI]
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00Road transport of goods or passengers
    • Y02T10/10Internal combustion engine [ICE] based vehicles
    • Y02T10/40Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • Agronomy & Crop Science (AREA)
  • Primary Health Care (AREA)
  • Animal Husbandry (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Mining & Mineral Resources (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Evolutionary Biology (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Medical Informatics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Catching Or Destruction (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a method for constructing a weed control model based on unmanned aerial vehicle cooperative intelligence. According to the invention, an edge computing platform is constructed, and the edge computing unmanned aerial vehicle carrying the platform carries out real-time processing and analysis on the collected rice field image data, so that flight protection decision information is quickly obtained and transmitted to the plant protection unmanned aerial vehicle, and the plant protection unmanned aerial vehicle commands the unmanned aerial vehicle to carry out the water and straw pest flight protection operation, thereby greatly improving the flight protection efficiency of the water and straw pest.

Description

一种基于无人机协同智能构建草害防治模型的方法A method to build a grass damage prevention and control model based on drone collaborative intelligence

技术领域Technical field

本发明涉及无人机灾害防治平台技术领域,具体涉及一种基于无人机协同智能构建草害防治模型的方法。The invention relates to the technical field of UAV disaster prevention and control platforms, and specifically relates to a method for constructing a grass damage prevention and control model based on UAV collaborative intelligence.

背景技术Background technique

植保无人机低空喷洒药剂、种子、粉剂是一项适应现代农业、现代植保需求的新型技术。与传统人工喷雾器相比,农用植保无人机具有操作简单、速度快、效率高、低成本、喷洒均匀、雾化效果好、可规模化作业等特点,可有效降低农户劳动强度,提高粮食质量和产量,减少农药对田间环境的污染,确保生态和粮食生产安全。Low-altitude spraying of pesticides, seeds, and powders by plant protection drones is a new technology that adapts to the needs of modern agriculture and modern plant protection. Compared with traditional manual sprayers, agricultural plant protection drones have the characteristics of simple operation, fast speed, high efficiency, low cost, uniform spraying, good atomization effect, and large-scale operation. They can effectively reduce the labor intensity of farmers and improve grain quality. and yield, reduce pesticide pollution to the field environment, and ensure ecological and food production safety.

而针对更大规模化作业,构建基于无人机的边缘计算平台,对无人机进行更加高效的管理和规划,对水稻杂草识别模型和草害严重等级进行更精确和优化的评价,则显得越来越重要。For larger-scale operations, build an edge computing platform based on drones, conduct more efficient management and planning of drones, and conduct more accurate and optimized evaluations of rice weed identification models and weed damage severity levels. appears more and more important.

发明内容Contents of the invention

本发明的目的在于克服现有技术存在的问题,提供一种基于无人机协同智能构建草害防治模型的方法。The purpose of the present invention is to overcome the problems existing in the existing technology and provide a method for constructing a grass damage prevention and control model based on collaborative intelligence of drones.

为实现上述技术目的,达到上述技术效果,本发明通过以下技术方案实现:In order to achieve the above technical objectives and achieve the above technical effects, the present invention is implemented through the following technical solutions:

一种基于无人机协同智能构建草害防治模型的方法,该方法包括以下步骤:A method for constructing a grass damage control model based on UAV collaborative intelligence. The method includes the following steps:

步骤1)训练数据模型:获取大量常见水稻杂草的视频和图像数据,学习并训练水稻杂草识别模型,以及采样杂草株数和鲜质量数据,学习并训练草害严重等级评价模型;Step 1) Training data model: Obtain a large amount of video and image data of common rice weeds, learn and train a rice weed identification model, and sample weed number and fresh quality data, learn and train a weed damage severity evaluation model;

步骤2)搭建边缘计算无人机:将步骤1)中训练出来的模型进行数据精简和优化,迁移到计算资源和存储资源都受限制的嵌入式边缘计算平台中,并将嵌入式边缘计算平台搭载到对应无人机上形成边缘计算无人机;Step 2) Build an edge computing drone: Streamline and optimize the data of the model trained in step 1), migrate it to an embedded edge computing platform with limited computing resources and storage resources, and integrate the embedded edge computing platform with Mounted on the corresponding drone to form an edge computing drone;

步骤3)无人机协同飞防作业:边缘计算无人机获取稻田区域信息并生成飞防决策信息,植保无人机与边缘计算无人机建立通信,接收并根据边缘计算无人机的飞防决策信息进行水稻草害的飞防任务;Step 3) UAV cooperative flight prevention operation: The edge computing UAV obtains rice field area information and generates flight prevention decision information. The plant protection UAV establishes communication with the edge computing UAV, receives and based on the flight control of the edge computing UAV. Use prevention decision-making information to carry out aerial prevention tasks of rice grass damage;

步骤4)模型重新优化:将步骤3)中飞防作业采集的图像数据导入云服务器上,对水稻杂草识别模型和草害严重等级评价模型进行重新优化,在下一次飞防作业前,将优化后的模型重新迁移到到嵌入式边缘计算平台中,实现对整个过程的闭环执行。Step 4) Model re-optimization: Import the image data collected during the aerial control operation in step 3) to the cloud server, and re-optimize the rice weed identification model and weed damage severity evaluation model. Before the next aerial control operation, the optimization will be carried out. The final model is re-migrated to the embedded edge computing platform to realize closed-loop execution of the entire process.

进一步的,所述步骤3)中,边缘计算无人机获取稻田区域信息包括以下步骤:Further, in step 3), the edge computing drone obtains rice field area information including the following steps:

步骤3.1)边缘计算无人机绕着整个稻田区域飞行,确定稻田区域范围;Step 3.1) The edge computing drone flies around the entire rice field area to determine the scope of the rice field area;

步骤3.2)将整个稻田区域划分为若干虚拟的网格区域并编号;Step 3.2) Divide the entire rice field area into several virtual grid areas and number them;

步骤3.3)边缘计算无人机再次起飞,拍照并通过其搭载的边缘计算平台计算该网格区域内草害情况,确定杂草类型和草害严重等级;Step 3.3) The edge computing drone takes off again, takes pictures and calculates the weed damage situation in the grid area through its edge computing platform to determine the weed type and weed damage severity level;

步骤3.4)边缘计算无人机将网格编号、网格区域位置、杂草类型和草害严重等级作为飞防决策信息传输给植保无人机,确定植保无人机的飞行轨迹,植保无人机携带相应除草剂飞到指定位置按照草害严重等级、飞防轨迹喷洒除草剂;Step 3.4) The edge computing drone transmits the grid number, grid area location, weed type and weed damage severity level as flight control decision-making information to the plant protection drone, and determines the flight trajectory of the plant protection drone. The machine carries the corresponding herbicide and flies to the designated location to spray the herbicide according to the severity level of the weed damage and the flight control trajectory;

步骤3.5)边缘计算无人机按路径飞到下一网格区域继续进行拍照和计算,依此循环执行步骤3.4),直至执行完成全部网格区域。Step 3.5) The edge computing drone flies to the next grid area according to the path to continue taking pictures and calculations. Step 3.4) is executed in this loop until all grid areas are completed.

进一步的,所述步骤1)中,利用TensorFlow机器学习框架,在云服务器上学习并训练水稻杂草识别模型和草害严重等级评价模型。Further, in step 1), the TensorFlow machine learning framework is used to learn and train a rice weed identification model and a weed damage severity evaluation model on the cloud server.

进一步的,所述步骤1)中,首先完成作业地区常见杂草的识别,利用9点取样法,对每个虚拟的网格区域取9个点,每点调查0.25m2,统计杂草株数和鲜质量,利用TensorFlow机器学习框架,训练水稻草害严重等级评价模型。Further, in step 1), first complete the identification of common weeds in the working area, use the 9-point sampling method, take 9 points for each virtual grid area, survey 0.25m 2 at each point, and count the number of weeds and fresh quality, using the TensorFlow machine learning framework to train a rice weed damage severity evaluation model.

进一步的,所述步骤2)中,通过MCU级AI推理框架Tengine-Lite,将训练出来的模型进行数据精简和优化,并迁移至基于开源边缘计算框架OpenEdge的嵌入式边缘计算平台中,再将该边缘计算平台搭载到开源pixhawk四旋翼无人机上形成边缘计算无人机。Further, in step 2), the trained model is data streamlined and optimized through the MCU-level AI inference framework Tengine-Lite, and migrated to an embedded edge computing platform based on the open source edge computing framework OpenEdge, and then The edge computing platform is mounted on the open source pixhawk quadcopter drone to form an edge computing drone.

本发明的有益效果是:The beneficial effects of the present invention are:

本发明构建了边缘计算平台,搭载该平台的边缘计算无人机对采集的稻田图像数据进行实时处理和分析,从而快速得到飞防决策信息,并传输给植保无人机,指挥其进行水稻草害飞防作业,大大提高水稻草害的飞防效率。The invention constructs an edge computing platform. The edge computing drone equipped with the platform processes and analyzes the collected rice field image data in real time, thereby quickly obtaining flight prevention decision-making information and transmits it to the plant protection drone to instruct it to carry out rice straw operations. Aerial pest control operations have greatly improved the efficiency of aerial pest control of rice straw.

附图说明Description of the drawings

图1为本发明的流程示意图;Figure 1 is a schematic flow diagram of the present invention;

图2为本发明的无人机协同飞防作业执行过程图。Figure 2 is a process diagram of the execution process of UAV coordinated flight prevention operation according to the present invention.

具体实施方式Detailed ways

下面将参考附图并结合实施例,来详细说明本发明。The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

如图1所示,一种基于无人机协同智能构建草害防治模型的方法,该方法包括以下步骤:As shown in Figure 1, a method for constructing a grass damage control model based on collaborative intelligence of drones includes the following steps:

步骤1)训练数据模型:获取大量常见水稻杂草的视频和图像数据,学习并训练水稻杂草识别模型,以及采样杂草株数和鲜质量数据,学习并训练草害严重等级评价模型;Step 1) Training data model: Obtain a large amount of video and image data of common rice weeds, learn and train a rice weed identification model, and sample weed number and fresh quality data, learn and train a weed damage severity evaluation model;

步骤2)搭建边缘计算无人机:将步骤1)中训练出来的模型进行数据精简和优化,迁移到计算资源和存储资源都受限制的嵌入式边缘计算平台中,并将嵌入式边缘计算平台搭载到对应无人机上形成边缘计算无人机;Step 2) Build an edge computing drone: Streamline and optimize the data of the model trained in step 1), migrate it to an embedded edge computing platform with limited computing resources and storage resources, and integrate the embedded edge computing platform with Mounted on the corresponding drone to form an edge computing drone;

步骤3)无人机协同飞防作业:边缘计算无人机获取稻田区域信息并生成飞防决策信息,植保无人机与边缘计算无人机建立通信,接收并根据边缘计算无人机的飞防决策信息进行水稻草害的飞防任务;Step 3) UAV cooperative flight prevention operation: The edge computing UAV obtains rice field area information and generates flight prevention decision information. The plant protection UAV establishes communication with the edge computing UAV, receives and based on the flight control of the edge computing UAV. Use prevention decision-making information to carry out aerial prevention tasks of rice grass damage;

步骤4)模型重新优化:将步骤3)中飞防作业采集的图像数据导入云服务器上,对水稻杂草识别模型和草害严重等级评价模型进行重新优化,在下一次飞防作业前,将优化后的模型重新迁移到到嵌入式边缘计算平台中,实现对整个过程的闭环执行。Step 4) Model re-optimization: Import the image data collected during the aerial control operation in step 3) to the cloud server, and re-optimize the rice weed identification model and weed damage severity evaluation model. Before the next aerial control operation, the optimization will be carried out. The final model is re-migrated to the embedded edge computing platform to realize closed-loop execution of the entire process.

所述步骤3)中,边缘计算无人机获取稻田区域信息包括以下步骤:In step 3), obtaining rice field area information by edge computing drone includes the following steps:

步骤3.1)边缘计算无人机绕着整个稻田区域飞行,确定稻田区域范围;Step 3.1) The edge computing drone flies around the entire rice field area to determine the scope of the rice field area;

步骤3.2)将整个稻田区域划分为若干虚拟的网格区域并编号;Step 3.2) Divide the entire rice field area into several virtual grid areas and number them;

步骤3.3)边缘计算无人机再次起飞,拍照并通过其搭载的边缘计算平台计算该网格区域内草害情况,确定杂草类型和草害严重等级;Step 3.3) The edge computing drone takes off again, takes pictures and calculates the weed damage situation in the grid area through its edge computing platform to determine the weed type and weed damage severity level;

步骤3.4)边缘计算无人机将网格编号、网格区域位置、杂草类型和草害严重等级作为飞防决策信息传输给植保无人机,确定植保无人机的飞行轨迹,如图2所示,本实施例中,采用MG-1P植保无人机,植保无人机携带相应除草剂飞到指定位置按照草害严重等级、飞防轨迹喷洒除草剂;Step 3.4) The edge computing drone transmits the grid number, grid area location, weed type and weed damage severity level as flight control decision-making information to the plant protection drone, and determines the flight trajectory of the plant protection drone, as shown in Figure 2 As shown, in this embodiment, the MG-1P plant protection drone is used. The plant protection drone carries the corresponding herbicide and flies to the designated location to spray the herbicide according to the severity level of the weed damage and the flight control trajectory;

步骤3.5)边缘计算无人机按路径飞到下一网格区域继续进行拍照和计算,依此循环执行步骤3.4),直至执行完成全部网格区域。Step 3.5) The edge computing drone flies to the next grid area according to the path to continue taking pictures and calculations. Step 3.4) is executed in this loop until all grid areas are completed.

所述步骤1)中,利用TensorFlow机器学习框架,在云服务器上学习并训练水稻杂草识别模型和草害严重等级评价模型。In step 1), the TensorFlow machine learning framework is used to learn and train the rice weed identification model and the weed damage severity evaluation model on the cloud server.

所述步骤1)中,首先完成作业地区常见杂草的识别,具体可以根据地理位置不同,来确定识别种类,比如苏州地区常见的杂草有32种,当用在苏州地区时,可以优先对这32种杂草进行识别,利用9点取样法,对每个虚拟的网格区域取9个点,每点调查0.25m2,统计杂草株数和鲜质量,利用TensorFlow机器学习框架,训练水稻草害严重等级评价模型。In step 1), first complete the identification of common weeds in the working area. The identification types can be determined according to different geographical locations. For example, there are 32 common weeds in Suzhou. When used in Suzhou, priority can be given to identifying weeds. To identify these 32 kinds of weeds, use the 9-point sampling method to take 9 points in each virtual grid area, survey 0.25m 2 at each point, count the number of weeds and fresh quality, and use the TensorFlow machine learning framework to train water Rice straw damage severity grade evaluation model.

所述步骤2)中,通过MCU级AI推理框架Tengine-Lite,将训练出来的模型进行数据精简和优化,并迁移至基于开源边缘计算框架OpenEdge的嵌入式边缘计算平台中,再将该边缘计算平台搭载到开源pixhawk四旋翼无人机上形成边缘计算无人机,OpenEdge框架将计算能力拓展至网络边缘终端,提供低延时的计算服务,包括AI推断、智能分析、函数计算等,通过OpenEdge和现有的水稻草害飞防云端管理套件配合使用,可以实现云端对水稻杂草识别模型和草害严重等级评价模型管理和下发更新,在边缘计算无人机端对采集的稻田图像数据进行实时处理和分析,从而快速得到飞防决策信息。In step 2), the trained model is data streamlined and optimized through the MCU-level AI inference framework Tengine-Lite, and migrated to an embedded edge computing platform based on the open source edge computing framework OpenEdge, and then the edge computing The platform is mounted on the open source pixhawk quadcopter drone to form an edge computing drone. The OpenEdge framework extends computing capabilities to network edge terminals and provides low-latency computing services, including AI inference, intelligent analysis, function computing, etc., through OpenEdge and Used in conjunction with the existing rice weed control cloud management suite, the cloud can manage and issue updates to the rice weed identification model and weed damage severity evaluation model, and the collected rice field image data can be processed on the edge computing drone side. Real-time processing and analysis to quickly obtain flight defense decision-making information.

以上所述仅为本发明的优选实施例而已,并不用于限制本发明,对于本领域的技术人员来说,本发明可以有各种更改和变化。凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims (4)

1. The method for constructing the grass damage control model based on unmanned aerial vehicle cooperative intelligence is characterized by comprising the following steps of:
step 1) training a data model: acquiring video and image data of a large number of common rice weeds, learning and training a rice weed identification model, sampling plant number and fresh quality data of the weeds, and learning and training a weed severity level evaluation model;
step 2) building an edge computing unmanned aerial vehicle: the data of the model trained in the step 1) is reduced and optimized, the model is migrated to an embedded edge computing platform with limited computing resources and storage resources, and the embedded edge computing platform is carried on a corresponding unmanned aerial vehicle to form an edge computing unmanned aerial vehicle;
step 3) unmanned aerial vehicle cooperation flight protection operation: the edge computing unmanned aerial vehicle acquires the paddy field area information and generates the flight protection decision information, the plant protection unmanned aerial vehicle establishes communication with the edge computing unmanned aerial vehicle, and receives and carries out the flight protection task of the water and straw injury according to the flight protection decision information of the edge computing unmanned aerial vehicle;
in the step 3), the step of obtaining the paddy field area information by the edge computing unmanned aerial vehicle comprises the following steps:
step 3.1), calculating the flight of the unmanned aerial vehicle around the whole paddy field area by the edge, and determining the range of the paddy field area;
step 3.2) dividing the whole paddy field area into a plurality of virtual grid areas and numbering the virtual grid areas;
step 3.3) taking off the edge computing unmanned aerial vehicle again, photographing and computing the weed conditions in the grid area through an edge computing platform carried by the edge computing unmanned aerial vehicle, and determining the weed types and the weed severity level;
step 3.4), the edge computing unmanned aerial vehicle transmits the grid number, the grid area position, the weed type and the grass damage severity level as flight protection decision information to the plant protection unmanned aerial vehicle, determines the flight track of the plant protection unmanned aerial vehicle, and sprays the herbicide according to the grass damage severity level and the flight protection track when the plant protection unmanned aerial vehicle flies to a designated position with the corresponding herbicide;
step 3.5), the edge computing unmanned aerial vehicle flies to the next grid area according to the path to continue photographing and computing, and the step 3.4) is executed in a circulating way until all the grid areas are completed;
step 4) model re-optimization: and 3) importing the image data acquired by the flying prevention operation in the step 3) into a cloud server, re-optimizing the rice weed identification model and the weed severity level evaluation model, and re-migrating the optimized model to an embedded edge computing platform before the next flying prevention operation to realize closed-loop execution of the whole process.
2. The method for constructing a weed control model based on unmanned aerial vehicle collaborative intelligence according to claim 1, wherein in the step 1), a rice weed recognition model and a weed severity level evaluation model are learned and trained on a cloud server by using a TensorFlow machine learning framework.
3. The method for constructing a weed control model based on unmanned aerial vehicle collaborative intelligence according to claim 2, wherein in the step 1), the identification of common weeds in an operation area is completed first, 9 points are taken for each virtual grid area by using a 9-point sampling method, and each point is investigated by 0.25m 2 Counting the number of plants and fresh quality of weeds, and training a water straw pest severity level evaluation model by using a TensorFlow machine learning framework.
4. The method for constructing the weed control model based on the unmanned aerial vehicle collaborative intelligence according to claim 1, wherein in the step 2), the trained model is subjected to data reduction and optimization through an MCU-level AI reasoning framework tennine-Lite, and is migrated to an embedded edge computing platform based on an open source edge computing framework OpenEdge, and then the edge computing platform is carried on the open source pixhawk quadrotor unmanned aerial vehicle to form the edge computing unmanned aerial vehicle.
CN201911198341.3A 2019-11-29 2019-11-29 Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence Active CN110956132B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911198341.3A CN110956132B (en) 2019-11-29 2019-11-29 Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911198341.3A CN110956132B (en) 2019-11-29 2019-11-29 Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence

Publications (2)

Publication Number Publication Date
CN110956132A CN110956132A (en) 2020-04-03
CN110956132B true CN110956132B (en) 2023-12-29

Family

ID=69978985

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911198341.3A Active CN110956132B (en) 2019-11-29 2019-11-29 Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence

Country Status (1)

Country Link
CN (1) CN110956132B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111783968B (en) * 2020-06-30 2024-05-31 山东信通电子股份有限公司 Power transmission line monitoring method and system based on cloud edge cooperation
CN112528912A (en) * 2020-12-19 2021-03-19 扬州大学 Crop growth monitoring embedded system and method based on edge calculation
CN113126650A (en) * 2021-03-03 2021-07-16 华南农业大学 Automatic weeding operation method for unmanned aerial vehicle
CN113255932A (en) * 2021-06-01 2021-08-13 开放智能机器(上海)有限公司 Federal learning training platform and method based on terminal equipment
CN115390504A (en) * 2022-09-29 2022-11-25 重庆电子工程职业学院 Wisdom farming system based on 5G thing networking
CN116048115A (en) * 2022-12-29 2023-05-02 中联智慧农业股份有限公司 Control method for unmanned aerial vehicle, group cooperation system and processor

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104392228A (en) * 2014-12-19 2015-03-04 中国人民解放军国防科学技术大学 Unmanned aerial vehicle image target class detection method based on conditional random field model
WO2017074966A1 (en) * 2015-10-26 2017-05-04 Netradyne Inc. Joint processing for embedded data inference
JP2017159750A (en) * 2016-03-08 2017-09-14 国立大学法人京都大学 Unmanned aircraft position estimation method and system
CN109446987A (en) * 2018-10-29 2019-03-08 北京麦飞科技有限公司 Method based on PCA and PNN algorithm detection rice pest grade
CN109871029A (en) * 2019-02-21 2019-06-11 华南农业大学 An optimization method for plant protection UAV route planning based on image processing technology

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104392228A (en) * 2014-12-19 2015-03-04 中国人民解放军国防科学技术大学 Unmanned aerial vehicle image target class detection method based on conditional random field model
WO2017074966A1 (en) * 2015-10-26 2017-05-04 Netradyne Inc. Joint processing for embedded data inference
JP2017159750A (en) * 2016-03-08 2017-09-14 国立大学法人京都大学 Unmanned aircraft position estimation method and system
CN109446987A (en) * 2018-10-29 2019-03-08 北京麦飞科技有限公司 Method based on PCA and PNN algorithm detection rice pest grade
CN109871029A (en) * 2019-02-21 2019-06-11 华南农业大学 An optimization method for plant protection UAV route planning based on image processing technology

Also Published As

Publication number Publication date
CN110956132A (en) 2020-04-03

Similar Documents

Publication Publication Date Title
CN110956132B (en) Method for constructing weed control model based on unmanned aerial vehicle cooperation intelligence
Singh et al. An intelligent WSN-UAV-based IoT framework for precision agriculture application
Idoje et al. Survey for smart farming technologies: Challenges and issues
Yépez-Ponce et al. Mobile robotics in smart farming: current trends and applications
CN105173085B (en) Unmanned plane variable farm chemical applying automatic control system and method
JP2022523836A (en) Control methods, devices and storage media for pesticide application
CN106227075A (en) Unmanned plane aquaculture based on wireless sense network precisely feeds intake operating system and method
CN113920474B (en) Internet of things system and method for intelligently supervising citrus planting situation
CN106428564A (en) Unmanned aerial vehicle sprinkling method and system based on growth cycles of crops
US20240276971A1 (en) Targeted treatment of specific weed species with multiple treatment devices
Faiçal et al. Fine-tuning of UAV control rules for spraying pesticides on crop fields: An approach for dynamic environments
Khuzaimah et al. Application and potential of drone technology in oil palm plantation: Potential and limitations
Radić et al. New technologies as a driver of change in the agricultural sector
Das et al. Drone-based intelligent spraying of pesticides: current challenges and its future prospects
WO2025097545A1 (en) Unmanned aerial vehicle flight height control method and apparatus, electronic device, and storage medium
CN117761077A (en) A drone-based rice pest and disease detection system and detection method
Chen et al. Novel intelligent grazing strategy based on remote sensing, herd perception and UAVs monitoring
CN117837366B (en) Agricultural supervision platform based on agricultural Internet of things
Hong et al. Adaptive target spray system based on machine vision for plant protection UAV
Yadav et al. Computer Vision for Volunteer Cotton Detection in a Corn Field with UAS Remote Sensing Imagery and Spot Spray Applications
Soman et al. Applications of UAV-AD (unmanned aerial vehicle-agricultural drones) in precision farming
Lawrence et al. Autonomous Agriculture: The Role of IoT, AI, and ML in Drone Operations
Shehab et al. Precision Farming with Autonomous Drones: Real-time Data and Efficient Fertilization Techniques
Kumar et al. An affordable multitasking drone for smart farming with the artificial intelligence feature
Gopal et al. Role of Digital Solutions in Agricultural Sector

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant