Design And Implementation Of Target Recognition System For Pigeon-borne Surface Fishing Vessels

In recent years,coastal and coastal fishery management,especially fishing vessel management,still faces problems such as illegal fishing and overfishing,resulting in the continuous decline of coastal and coastal fishery resources.With the introduction of innovative models such as "Marine Information+" and "Marine Internet of Things+",it has become the trend of coastal and offshore fishery management to build a more comprehensive and intelligent fishing vessel supervision system to maintain the sustainable use of coastal fishery resources.In terms of intelligent video monitoring of coastal fishing boats,the current difficulties are the miniaturized design and carrying of terminal equipment,and how to quickly identify the targets of fishing boats on the sea surface under various sea conditions.key technology.Therefore,this paper studies the target recognition algorithm of surface fishing boats,and designs and implements a target recognition system for pigeon-borne sea fishing boats.The system uses the camera sensor of the pigeon-borne equipment to collect the video data of the fishing boat.The image data is first processed by the pigeon-borne equipment,and then transmitted to the PC through the 4G module and server.Finally,the target of the fishing boat is identified on the PC through the target recognition algorithm,and the upper The machine displays the target recognition results.The main research contents of this paper are as follows:(1)The overall framework of the target recognition system for pigeon-borne surface fishing boats is designed.The specific requirements such as the volume,weight and working time of the pigeon-borne equipment of the system are analyzed,and the functional requirements and performance indicators of the system are determined.The target recognition system of pigeon-borne surface fishing boats composed of the host computer software for positioning display.(2)The software and hardware design of the target recognition system of the pigeon-borne sea surface fishing vessel is realizedFirst of all,according to the functional requirements of the device’s image acquisition,image processing and data communication,a hardware device integrating MCU control system and data transmission system is realized,and then it is realized through the program of the device’s 4G communication software,server transfer software and host computer software.A data link from the device side→the server side → the PC side is constructed,and the data of the pigeon-borne equipment is transmitted to the PC side to complete the final processing.(3)Developed the target recognition algorithm of surface fishing boats,and realized the target recognition function of sea surface fishing boats in the pigeon-borne system.Aiming at the shortcomings of the original YOLOv5 model in the multi-scale fishing boat target and long-distance small target scene in the constructed sea surface fishing boat target data set,an improved YOLOv5 model fishing boat target recognition algorithm is proposed.By adding a micro-scale detection layer to the detection layer of YOLOv5,the problem of missing detection of small targets in the model under multiple targets is reduced,and the CBAM attention module is added to the neck network of YOLOv5 to improve the model’s ability to recognize small target fishing boats in the sea background.The improved YOLOv5 model is trained and tested on the constructed surface fishing boat target data set.Compared with other algorithms,it reduces the missed detection of small fishing boats and improves the classification confidence score of small target fishing boats.(4)The test of the target recognition system for the pigeon-borne sea surface fishing vessel is completed,which verifies the effectiveness of the system for target recognition of the fishing vessel in the sea scene.The simulation test and in-body test were completed for the target recognition system of pigeon-borne sea fishing boats.First,in the simulation test of the system,the system function test was completed using the data communication function test and the target recognition function test,and the special test method was used to complete the test.Testing of system design metrics.The simulation test results show that the functions and design indicators of the system meet the expected design requirements.Then,the actual scene flight recognition test of the pigeon-borne system is completed through the in-body test,which verifies that the pigeon-borne system can effectively identify the target of sea fishing boats in the sea scene.