ROBOT NAVIGATION USING IMAGE PROCESSING AND ISOLATED WORD RECOGNITION

Abstract

Abstract – In this paper, a practical implementation of image processing and isolated word recognition has been explained. A modeled vehicle has been driven autonomously on the road and can navigate on a road on its own, stop at zebra-crossing, follow traffic lights and reach an end point or can be controlled through voice commands. The vehicle has been developed using an old scrap body of a Kinetic Sunny and its engine, Servo motors, microphone, PCB and cameras. The image processing is done with the help of a laptop using MATLAB image processing toolbox and speech processing is done using HTK (hidden markov model toolkit). The robot navigated through roads with marked lanes and traffic light using image processing after taking inputs from the environment with the help of CCD camera and webcam. However, on the roads with no standardized lane marking and traffic lights, the robot was controlled using voice commands only and processed by HTK toolkit.