Stereovision based obstacle detection system for unmanned surface vehicle

This paper presents the stereo based obstacle detection system for unmanned surface vehicle (USV). The system is designed and developed toward the aim of real-time and robust obstacle detection and tracking at sea. Stereo vision methods and techniques have been employed to offer the capacity of detecting, locating and tracking multiple obstacles in the near field. Field test in the real scenes has been conducted, and the obstacle detection system for USV is proven to provide stable and satisfactory performance. The valid range with good accuracy of depth estimation is from 20 to 200 meters for high speed USV.