Roboic recognition for unstructered 2-D parts to pick and place objects

Abstract

Robot grasping of unstructured objects to pick and place in a home environment is a problem in autonomous assembly. Current investigation focus on technology related with machine vision for image manipulation. Desired grasping location is found out from 2-D vision by the bounding rectangle and bounding rectangle diagonal to calculate total number of pixel on parts. Various steps are adopted especially tools and built-in functions of LabVIEW for image acquisition, manipulations, and processing. These tools and functions helped us to create efficient algorithm for object recognition, grasping of sensitive and hard objects without fail. The system contains routine functions for database creation, learning, training, searching, monitoring, and controlling of autonomous mobile robot. Communication between hardware to software and viz versa are configured by VISA resource through COM port. Experiments analysis is done for evaluation of robot in real time.