Object detection and tracking method of AUV based on acoustic vision

In spite of the recent advances in AUV intelligence techniques, robustly solving their vision perception in ocean environment is still a challenging problem. In this paper, we describe a new framework for object detection and tracking of an AUV, which includes software system and hardware system. It is designed for Vision-based AUV carrying forward looking sonar. The raw data from the sensors are processed and fused in-line with a Gaussian particle filter (GPF), which estimates the positions of moving objects. The proposed method has been tested in the sea on a certain AUV. The results show it is feasible.