Obstacle Tracking on Unmanned Surface Vehicle Using Kalman Filter

Abstract

Unmanned Surface Vehicles (USV) are self-driving vehicles that operate on the water surface. In order to be operated autonomously, USV has a guidance system designed for path planning to reach its destination. The ability to detect obstacles in its paths is one of the important factors to plan a new path in order to avoid obstacles and reach its destination optimally. This research designed an obstacle tracking system which integrates USV perception sensors such as camera and Light Detection and Ranging (LiDaR) to gain information of the obstacle’s relative position in the surrounding environment to the ship. To improve the relative position estimation of the obstacles to the ship, Kalman filter is applied to reduce the measurements noises. The results of the system design are simulated using MATLAB software so that results can be analyzed to see the performance of the system design. Results obtained using the Kalman filter show 12% noise reduction. Keywords: filter kalman, obstacle tracking, unmanned surface vehicle.