摘要
针对现有分拣搬运机器人不可移动、操作单一、无法智能分拣搬运等特点,设计了单目视觉智能分拣搬运机器人。通过对摄像头获取的图像进行中值滤波、去除阴影、图像分割等预处理,结合形状和颜色特征对物料进行识别。提出了基于摄像头角度反馈测距法的目标定位算法,并采用增量式PID控制,使机器人动作更稳定和快速。实验表明该机器人在复杂环境下可准确高效完成物料分拣搬运。
Aiming at the problems that the sorting and handing robots are immovable,singleoperation,cannot be sorted and transported intelligently,this paper designs a monocular vision intelligent sorting and handing robot.The robot obtains information through the camera,performs median filtering,shadowremoval,image segmentation processing on the image,combines the shape and color characteristics to identify the material.A target location algorithm based on camera angle feedback ranging method is proposed,and the incremental PID control is adopted to make the robot move more stable and fast.The experiment shows that the robot can complete the sorting and handling of materials accurately and efficiently in the complex environment.
作者
朱颖
黎伟健
郭国辉
严宗灿
Zhu Ying;Li Weijian;Guo Guohui;Yan Zongcan(Huali College Guangdong University of Technology,Guangzhou 511325,China)
出处
《国外电子测量技术》
2020年第10期138-142,共5页
Foreign Electronic Measurement Technology
基金
广东省普通高校青年创新人才项目(2019KQNCX202)
广东省大学生创新训练计划项目(201813656019)资助。
关键词
搬运机器人
目标定位
角度反馈测距
阴影去除
handling robot
target recognition
angle feedback ranging
shadow remove
