Research On Obstacle Avoidance Of Unmanned Surface Vessel Based On Binocular Stereo Vision Technology

With the continuous growth of human demand for resources and energy,the ocean with rich mineral resources has become the targets of competition among countries,and the exploration and research of the oceans has incessantly desired.USV(unmanned surface vehicle),as an important platform in the future marine operation,the research of related technologies is particularly required.Binocular stereo vision technology is one of the important forms of machine vision.It is a method to obtain 3D geometric information of the target object by calculating the positional deviation between the corresponding points of the images,based on the basic parallax principle and using the binocular vision camera to acquire two different images of the measured object from different positions.This paper mainly focuses on the obstacle avoidance problem of USV based on binocular stereo vision technology during voyage.Firstly,the front video data is acquired by the binocular stereo vision system installed on the USV,and the real-time calculation processing is performed by the onboard industrial computer,including binocular stereo calibration,stereo matching,depth reconstruction,target information detection,etc.The purpose is to obtain the information in front of the USV and ensure safe navigation.At the same time,a collision avoidance algorithm called "grid" is proposed innovatively.The parallax map is grid-divided to obtain the steering angle required for the collision avoidance operation of the USV by analyzing and processing the obstacle target.Then the PD steering controller designed by closed-loop shaping filter algorithm makes the USV to complete the operating of collision avoidance.The experimental verification of USV’s binocular stereo vision system and the designed PD steering controller was carried out.The experimental results show that the binocular stereo vision system can detect dangerous targets ahead of the USV in real time.Furthermore,the system can detect and identify the obstacle targets that affect navigation safety quickly and accurately,and perform collision avoidance operation in time.At the same time,the proposed USV PD steering controller in this paper is tested by MATLAB simulation tool,and the results show that it performs good control effect,which meets the requirements of the USV obstacle avoidance experiment.