Obstacle avoidance for mobile robot using RGB-D camera

Abstract

Autonomous navigation of mobile robots in indoor cluttered environment is always a challenging task. This research paper represents the effectiveness of an RGB-D camera in order to achieve the task. With the help of the Kinect as a vision sensor and utilizing the open source Robot Operating System (ROS), this research aims to evaluate effectiveness of the system for Mobile robots. Research results show that within the typical range of the Kinect sensor, it performed very well with high accuracy. Experiment performed on a Raspberry Pi system shows minimum lag of the robot for avoiding obstacles, and verified the feasibility of the implemented algorithm.