Depth Image-based Distance Measurement for Two-wheeled Robots

In this paper, the two-wheeled robot is combined with Kinect sensor to measure obstacle distance. When the two-wheeled robot autonomously balances or moves, the ranging value will be unstable due to the shaking of the two-wheeled robots. This paper uses a smart prediction method to correct the ranging error value in order to increase the accuracy of ranging value and the obstacle avoidance decision. Finally, in order to verify the proposed method, some experiments are performed.