摘要
以割草机器人作业过程为研究对象,分析了机器人作业的环境特点,建立特定的环境参数模型,并运用计算机视觉算法进行作业环境的自动化识别。试验结果表明:割草机器人视觉系统所包含的目标识别算法和边界特征导航算法,对割草机器人作业过程中的目标识别准确率达到95%以上,控制指令发出至作业过程的平均反应距离不大于20mm,能够实时准确地进行割草作业过程自动控制。
This paper takes the working process of mowing robot as the research object,analyzes the characteristics of robot working environment,establishes the specific environmental parameter model,and uses computer vision algorithm to automatically identify the working environment.The experimental results show that the target recognition algorithm and boundary feature navigation algorithm contained in the vision system of the mowing robot can achieve the target recognition accuracy of more than 95%,and the average reaction distance from the control command to the operation process is not more than 20 mm,which shows that the mowing robot can realize real-time and accurate automatic control of the mowing process.
作者
徐晨
Xu Chen(Chongqing Technology and Business Institute, Chongqing 401520, China)
出处
《农机化研究》
北大核心
2022年第12期235-238,共4页
Journal of Agricultural Mechanization Research
基金
重庆市教委科学研究项目(KJQN202004008)。
关键词
割草机器人
计算机视觉
图像采集
特征比对
mowing robot
computer vision
image acquisition
feature comparison
