Implement of a 6-DOF manipulator with machine vision and machine learning algorithms

This paper explores the application of a Machine Vision and Machine Learning Algorithm to a Manipulator with six degrees of freedom (6-DOF). A Kinect sensor were used to extract images from a screen and obtain the relevant target information. Image processing was accomplished using a Scale-invariant feature transform (SIFT) Algorithm to capture image of the target object. The processed visual is rendered on the computer controller and Manipulator Learning is accomplished using a Reinforcement Learning Algorithm. Markov Decision Processes (MDP) were used to train the Manipulator to move to the location of the target object. Experimental results showed the Reinforcement Learning Algorithm proposed in this paper is effective and can be utilized on a 6-DOF Manipulator with reproducible results.