Vision-Guided Mobile Robot Navigation

Abstract

In this paper, the vision-based navigation problem for mobile robots in indoor environments is discussed. Corridor guidelines between the walls and the floor are used to guide the robot to move in the middle of the hallway. A neural network method is proposed to calibrate the camera and then the corridor guidelines are transformed from image coordinates to robot coordinates. The corridor guidelines are extracted using the least square estimator. The backstepping method is employed to design the controller that guarantees the robot to move in the desired path. Experimental and simulation results show the efficiency of the proposed method.