VISION CONTROLLED PICK AND PLACE OF MOVING OBJECT BY 3R ROBOT

robotics and automation

Vision is a quintessential sense for any system, be it human system or a mechanical system. Industrial automation involves robotic arms for performing various factory operations, pick and place operation being their basic and most common application. Performing this pick and place operation in real time with vision sensors is a step deeper into automation. Motion of the object adds further complexity to the problem. This paper presents a 3R robot in eye-to-hand orientation for the application of pick and place of a moving object using vision system. The methodology consists of real time video extraction, object detection using the motion of object as classification criterion and point based tracking to predict future state of the moving object using Kalman filter in MATLAB. The tracked estimate is transferred to the robot using Arduino platform. A 3R robot with three degrees of freedom is subjected to the implementation of the methodology. A USB webcam is used to provide the visual information to the computer, where image processing is done on the captured video and object is detected and the future position of the object is predicted using Kalman filter algorithm. The robot is controlled using Arduino IDE to reach the predicted position and grasp the object with maximum accuracy. The paper emphasizes on the control of 3R robot with visual parameters. That is established by integrating the computer vision techniques with mechanical structure. The Arduino platform is integrated with MATLAB to implement the technique on the 3R robot and tested to reach the final position with satisfactory accuracy.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.