Simultaneous localization and mapping of autonomous underwater vehicle using looking forward sonar

A method of underwater simultaneous localization and mapping (SLAM) based on on-board looking forward sonar is proposed. The real-time data flow is obtained to form the underwater acoustic images and these images are pre-processed and positions of objects are extracted for SLAM. Extended Kalman filter (EKF) is selected as the kernel approach to enable the underwater vehicle to construct a feature map, and the EKF can locate the underwater vehicle through the map. In order to improve the association efficiency, a novel association method based on ant colony algorithm is introduced. Results obtained on simulation data and real acoustic vision data in tank are displayed and discussed. The proposed method maintains better association efficiency and reduces navigation error, and is effective and feasible.