基于OpenMV的投篮机器人设计  

Design of Shooting Robot Based on OpenMV

作　　者：黄淑芬[1] 罗炜杰 洪世杰 高志涛 Huang Shufen;Luo Weijie;Hong Shijie;Gao Zhitao(School of Mechanical and Electrical Engineering,Huali College of Guangdong University of Technology,Guangzhou 511325,China)

机构地区：[1]广东工业大学华立学院机电工程学院,广州511325

出　　处：《机电工程技术》2020年第8期121-123,共3页Mechanical & Electrical Engineering Technology

基　　金：广东省普通高校特色创新项目(编号:2018KTSCX311)。

摘　　要：为满足成本低、识别精确和动作灵活等需求,开发出一款基于Python语言,以OpenMV为控制核心的投篮机器人。OpenMV摄像头采集目标图像,其自带的STM32单片机可以进行视觉识别和定位,并控制机械臂、手爪、小车和弹射器动作。依据Lab颜色空间,由OpenMV阈值编辑器调试颜色阈值,库函数image.find_blobs()识别目标颜色。测试表明,该投篮机器人能正确识别目标和抓取小球,但投篮命中率不高。采用OpenMV及其Python机器视觉库,制造成本低,程序开发效率高。In order to meet the requirements of low cost,accurate identification and flexible action,a shooting robot based on Python language and taking openmv as the control core was developed.The target image is collected by the OpenMV camera,and the STM32 single-chip microcomputer can be used for visual recognition and positioning,and control the movements of the manipulator,gripper,car and catapult.According to the lab color space,the openmv threshold editor debugs the color threshold and library functions image.find_blobs()recognize the target color.The test results show that the robot can correctly identify the target and grasp the small ball,but the shooting percentage is not high.Using OpenMV and its Python machine vision library,the manufacturing cost is low and the program development efficiency is high.

关 键 词：OpenMV 投篮机器人 视觉识别 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]