基于嵌入式视觉的园区垃圾处理机器人设计

Embedded Vision-based Design of Campus Waste Disposal Robot

当前园区垃圾分拣以人力为主,受气候、时间和体力的限制,无法实现垃圾的连续大规模分拣,并且垃圾中存在大量有毒物质,对人体造成极大危害。基于此,针对园区垃圾处理的问题,结合机器视觉和嵌入式系统的发展,设计了一种嵌入式系统视觉的垃圾处理移动机器人。设计以STM32F103为全局控制核心,由若干模块(电源模块、显示模块、电机模块、超声模块、视觉模块等)构成硬件系统。机器人利用机器视觉精准定位垃圾桶,辅助机械臂完成垃圾收集、自主垃圾倾倒等功能。测试结果表明,机器人的PID调速响应速度快,正确识别标识垃圾桶并能够对园区垃圾的回收率达到93.3%,有效降低工作劳动强度,提高工作效率。

At present,waste sorting in the park is mainly manpower,which is limited by climate,time and physical strength,and fails to realize continuous large-scale sorting of garbage.Moreover,there are a lot of toxic substances in the garbage,causing great harm to human body.In this paper,aiming at the problem of waste disposal in parks,combined with the development of machine vision and embedded system,a mobile robot for waste disposal with embedded system vision is designed.The design takes STM32F103 as the global control core,and consists of several modules(power module,display module,motor module,ultrasonic module,vision module,etc.)to form the hardware system.The robot uses machine vision to accurately locate the trash can,and assists the robot arm to complete garbage collection,autonomous garbage dumping and other functions.The test results show that the robot's PID speed regulation response is fast,correctly identifies the marked garbage cans and is able to recycle 93.3%of the park's garbage,effectively reducing work labor intensity and improving work efficiency.