基于单目视觉的全自动捡球机器人控制算法设计  

Control algorithm design of automatic ball-picking robot based on monocular vision

作　　者：高长旭[1] 孙巧妍[2] 陈祥光 张德军[2] GAO Changxu;SUN Qiaoyan;CHEN Xiangguang;ZHANG Dejun(School of Humanities,Yantai Nanshan University,Yantai 265713,China;School of Engineering,Yantai Nanshan University,Yantai 265713,China;School of Chemistry and Chemical Engineering,Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]烟台南山学院人文学院,山东烟台265713 [2]烟台南山学院工学院,山东烟台265713 [3]北京理工大学化学与化工学院,北京100081

出　　处：《传感器与微系统》2022年第9期88-91,共4页Transducer and Microsystem Technologies

基　　金：山东省自然科学基金资助项目(ZR201709210161);烟台南山学院青年科研启动基金资助项目(NQ202014)。

摘　　要：为了解决乒乓球训练基地繁重的人工捡球工作,设计了一种全自动捡球机器人适用的控制算法。机械结构方面采用双齿轮逆向旋转形式的滚筒结构执行捡球动作,使用单目摄像头固定安装到捡球小车上周期性拍摄图片,通过无线传输或者网络云端传送到个人电脑(PC)。使用MATLAB软件通过识别并补偿干扰光斑、降噪、识别小球质心、规划最优路径等算法识别小球位置并规划行进路线。将识别出的质心坐标和规划路线顺序无线传送到安装在小车上的嵌入式系统,根据图像和实物的相应尺寸关系,控制小车按照规定路线行进,直至所有的球都捡起。通过仿真和实物制作,证明该方法切实可行。In order to solve the heavy work of manual picking up ball in table tennis training base, a control algorithm suitable for automatic picking ball robot is designed.In the mechanical structure, a double-gear roller structure with reverse rotation is adopted to perform the ball-picking action, use a monocular camera fixed to pick up the ball cart to periodically take pictures, through wireless transmission or network cloud to the PC.Using MATLAB software, the position of the ball is identified and the route is planned by a series of algorithms, such as identifying and compensating interference spot, reducing noise, identifying the center of mass of the ball, planning the optimal path and so on.The identified centroid coordinates and route planning sequence are then transmitted wirelessly to the embedded system mounted on the trolley, according to the size relationship between the image and the object, the trolley is controlled to move along the prescribed route until all the balls are picked up.Through the simulation and real production, it is proved that the scheme is feasible.

关 键 词：图像识别 去光斑干扰 质心识别 最优路径规划 双齿轮逆向旋转滚筒 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]