Abstract
Many mobile robot navigation methods utilize laser scanners, ultrasonic sensors, vision cameras, and so on for detecting obstacles and path following. However, human utilizes only vision (e.g. eye) information for navigation. In this paper, we propose a mobile robot control method based on machine learning algorithms which use only the camera vision. To efficiently define the state of the robot from raw images, our algorithm has the image processing and feature selection steps to choose the feature subset for a neural network and uses the output of the neural network learned by the supervised learning. The output of the neural network is utilized the state of reinforcement learning algorithm to learn the obstacle avoiding and path following strategy from the camera vision image. The algorithm is verified by two experiments which are the line tracking and the obstacle avoidance.
