Design of intelligent sorting system for industrial parts based on machine vision
This article presents the design of an intelligent sorting system that integrates machine vi-sion with robotics to meet the extensive application demands in industrial production.Specifically,this system is engineered to identify and locate moving target objects on a conveyor belt through machine vi-sion.It ensures precision in the robot's identification and grasping positioning by calibrating the camera and coordinate systems.The system employs the YOLOv5 detection algorithm to recognize target objects on the conveyor belt and utilizes the centroid method to determine the central pixel coordinates of the ob-jects.Subsequently,an affine transformation approach is applied for the precise localization of the target objects.Experimental outcomes indicate that during the industrial part-sorting process,this intelligent sorting system achieves an accuracy rate of over 98%in identifying specific target workpieces.Moreover,the positioning error remains within less than 1 mm when the robot performs the task of grasping these workpieces.Consequently,the designed intelligent sorting system can carry out high-precision identifi-cation,positioning,and grasping of parts in industrial production.The system is capable of effectively meeting the precise requirements for automatic part sorting in industrial production,demonstrating its broad potential applications in the field of industrial automation.
machine visionindustrial robotobject sortingcalibrationrecognition and grasping
万方数据
维普
