Dynamic path planning algorithm for wearable visual navigation system based on the improved A𢈗

Abstract

Path planning is one important component of obstacle avoidance system for the blind. According to the requirements of wearable visual navigation system (WVNS), a combinatorial planner combining the static global and the dynamic local planning is proposed in this paper. In each frame of the video captured by the binocular camera, the aiming-tracking mechanism for tracking global path is used to determine the local target. Then with the local path planning carried out based on the real-time environmental information observed by the camera and the positions of the dynamic obstacles predicted by the autoregressive model, the deviation angles are calculated to instruct the blind to the destination. This research proposed an improved dynamic weighted A𢈗 algorithm in which “potential field” is introduced to dynamically adjust the weights of the heuristic function. The results of simulations and experiments show that the proposed algorithm act more effectively and it is reasonable in both static and dynamic environment.