基于机器视觉的垃圾分拣机器人设计

Design of Garbage Sorting Robot Based on Machine Vision

设计了一款基于机器视觉的垃圾分拣机器人,其利用陀螺仪和编码器配合GPS进行路径规划、摄像头配合超声波模块进行物体检测、机械臂进行垃圾的抓取。为了准确识别垃圾种类,采用SIFT图像匹配算法利用逆运动学求解进行机械臂轨迹的规划。经测试,机器人能够正确识别大多数带有商标垃圾而不受光照等因素的影响,并能够完成对具有一定形状的垃圾的抓取,初步实现了垃圾分拣的目的。

A machine vision-based garbage sorting robot is designed.It uses gyroscope and encoder to cooperate with GPS for path planning,camera with ultrasonic module for object detection,and robot arm for garbage capture.In order to accurately identify the type of garbage,the SIFT image matching algorithm is used to solve the trajectory of the robot arm by inverse kinematics.After testing,the robot can correctly identify most of the trademarked garbage without the influence of light and other factors,and can complete the pair.The grasping of certain shaped objects has initially achieved the goal of garbage sorting.