Human Gesture Based Wireless Robotic Arm Using Matlab

Abstract

Wireless motion control using human gestures is a prominent field. Once this field gets completely explored, there are a lot of advantages. The main advantage of this paper is to reduced hardware, thus making user-friendly motion control. The primary goal of the project is to control every real-time motion of an arm based 4-wheel drive using wearable colored finger bands of suitable intensity. Each color movement corresponds to a particular motion on the machine[4]. This facilitates programming in MATLAB to a great extent. Through rigorous real-time image processing techniques, extra hardware components and wireless communication between the computer and the machine, it is possible to control the movement of the machine. This project uses two different colored bands namely red and green. They are sometimes used individually and sometimes simultaneously as per the application required. This technique has been used for achieving the following task: Detection of color and shape of the finger for motion control. This paper presents the implementation and analysis of real-time finger tracking system used for interaction purposes. As this technique works over the continuously grabbed images, it appears as if colored finger band (i.e. object) is being tracked in a live video. Consecutive centroids are continuously compared and a decision is made for movement in UP, DOWN, LEFT, RIGHT direction of the motor. Also, arm movements of the 4-wheel drive in UP and DOWN directions are controlled using fingers.