基于视觉识别物料分拣机器人的设计

Design of material sorting robot based on visual recognition

为解决传统物料分拣投射机器人抓取结构和发射装置相分离、功能单一以及效率不高等问题。研究开发了一套以STM32H7为核心控制器,利用OpenMV摄像头对图像进行采集、处理与分析的分拣机器人,采用机器视觉图像处理算法、位置PID控制算法实现对物料颜色、形状、位置的识别及机器人行走控制。结构方面,由于该机器人采用抓取投篮一体化结构设计,抵达投射区瞄准目标区通过圆杆推动物料实现发射,能够实现高效智能化分拣。多次试验结果表明,该机器人在复杂环境下能保证98%以上的抓取率和97%以上的投篮准确率,快速地完成了分拣任务。为市场多样化的需求提供了简单有效的解决方案。

In order to solve the problems such as the separation of grasping structure and launching device, the singleness of function and the low efficiency of the traditional material sorting and projection robot.A sorting robot with STM32 H7 as the core controller and openmv camera for image acquisition, processing and analysis. The machine vision image processing algorithm and position PID control algorithm are used to recognize the color, shape and position of the material and control the robot walking.As for the structure of the robot, it can achieve high efficiency and intelligent sorting by using the round bar to push the material to be launched when it reaches the target area. The results of many experiments show that the robot can guarantee more than 98% grasping rate and more than 97% shooting accuracy in complex environment, and accomplish sorting task quickly. It provides a simple and effective solution to the diversified demands of the market.