Research on Collision Avoidance Method Based on Image Recognition Technology

Abstract

This paper describes the development of unmanned surface vessel (USV) collision avoidance technology, introduces the basic principles and key technologies of image recognition technology, analyzes the shortcomings of traditional collision avoidance technology, and explains the research status of the image recognition technology in the collision avoidance of unmanned surface vessel. The design ideas of the existing research and development results are summarized. Advantages and disadvantages are analyzed and compared. This paper focuses on the techniques of stereo vision, deep learning, multi-sensor fusion, and the method of unmanned surface vessel collision avoidance based on image recognition technology. The future development trend of this field is also discussed.