Date of Publication :17th August 2017

Abstract: Vision based pick and place operation using colour identification is one of the most researched topic. Machine vision is supposed to be the pioneer future technology. It gives flexibility to identify and manipulate any surrounding object .This thesis focuses on pick and place operation of a 4-axis articulate robot using colour segmentation technique.Various colour segmentation techniques are discussed and a general comparison between K-mean clustering technique and Fuzzy C mean clustering technique is made.The vision sensor used here is a Kinect v2 camera which is a RGB-D camera for depth measurement and coordinate determination .In the experimental setup a set of coloured markers are placed, from which the robot pick the required segmented coloured marker. An important part of industrial robot manipulators is to achieve desired position and orientation of end effector or tool so as to complete the pre-specified task. To achieve the above stated goal one should have the sound knowledge of inverse kinematic problem. The robot kinematics is carried out considering both forward and inverse kinematics. Finally the accuracy of the robot is measured using Kinect sensor

Reference :