| The ability of the unmanned ship to achieve autonomous berthing depends on its control system being able to smoothly control the unmanned ship during the berthing,and the control effect mainly depends on the precise self-state information obtained by the unmanned ship during the berthing and surrounding environment information.In order to meet the requirements of distance measurement accuracy in the berthing of unmanned ship,this paper proposes an autonomous berthing environment perception system for unmanned ship based on the fusion of vision and lidar.The relative distance to ensure the safety and stability of the unmanned ship berthing.The environment perception system uses vision to quickly identify and locate the shore mooring equipment,and integrates the mooring equipment information obtained by vision and lidar through space and improve the ability of acquiring distance information between unmanned ship and target berth in the process of autonomous berthing.The main research contents of this paper are as follows:1.Overall design of the environment perception system during the autonomous berthing of unmanned ships.Research the use of mooring equipment to realize the autonomous berthing of the unmanned ships.According to the requirements of environmental information during the autonomous berthing,design and build an environmental perception system based on vision and lidar,and complete the selection of hardware in the system.and layout,determine the data transmission and control method of the environment perception system.2.Recognition of mooring equipment based on improved YOLOv4.In order to quickly and accurately identify mooring equipment,the images of mooring equipment are preprocessed based on Laplacian edge detection and median filtering method,and the CNN convolution in the YOLOv4 algorithm is replaced by Depthwise Separable convolution to reduce the model size.calculation amount.The experimental results show that,compared with the original model,the improved model shortens the recognition time and improves the recognition accuracy.3.Positioning of mooring equipment based on the fusion of stereo vision and lidar.The working principle of depth camera and lidar is studied,combined with the practical application scenario of unmanned ship berthing,the use of two depth cameras and the introduction of a rotating pan/tilt head will expand the scope of environmental perception.Through spatial geometry and coordinate transformation,the spatial unification of perception information obtained by vision and laser ranging system is completed,and the acquisition time of the two sensors is synchronized by multi-thread processing,realizing the fusion of visual information and laser information,and improving accuracy of the mooring equipment distance measurement.4.Build an unmanned ship berthing environment perception platform to verify its reliability in the autonomous berthing.Experiments are carried out in the range of short distance and long distance,obtain the distance between the unmanned ship and the mooring equipment,and compare the measurement result with the actual distance to determine error range.The experimental results show that the distance measurement accuracy of the environmental perception system designed in this paper is significantly better than the visual measurement results,and the error of the obtained distance value is always controlled within the range of 10 mm,which meets the requirements of the measurement distance accuracy for autonomous berthing of unmanned ships. |
