3-D Reconstruction of Underwater Object Based on Extended Kalman Filter by Using Acoustic Camera Images

This paper presents a sensing method to measure three-dimensional (3-D) information in an underwater environment using an acoustic camera. Acoustic cameras can acquire clear information even in turbid water which is difficult to photograph with an optical camera. In addition, its detection range is extensive. Compared to traditional sensors, acoustic cameras with no restrictions on vision are the most powerful sensors for acquiring underwater information. In this paper, we propose a novel approach which enables 3-D measurement of underwater objects using arbitrary viewpoints based on an extended Kalman filter (EKF). By using the probabilistic method based on the EKF, 3-D reconstruction of underwater objects is possible even if the control input for camera movement has uncertainty. Furthermore, since the EKF based estimation is performed sequentially each time, our methodology can be adapted to realtime applications. Simulation and experimental results show the effectiveness of the proposed method.