Research On Target Track Algorithm Based On Radar Image Features

With the rapid development of China's shipping business,the number of ships in river channels is increasing,the management of ship channels is becoming more and more complicated,the number of ships is increasing,and the management of river channels is becoming more and more important.China's economic strength has developed rapidly in recent years,the electronic information industry has developed rapidly in the shipping industry,and the market demand is broader,but the civilian radar industry depends on imports and is monopolized by foreign companies.The domestic radar development starting point is low and the development is late.China's ability to own independent radar intellectual property has extremely high market value and research significance.The radar installed on the ship has an important role.It is the eyes of the captain.It is used to observe the situation around the island and the ship to determine the position of the ship,display the route,guide and supervise the ship's navigation route and status.Collision avoidance and other aspects play a more important role,playing an important role in civil and military fields.Radar tracking and identification of multiple targets.With the advancement and development of technology,the computer level and image processing technology continue to improve.Through image processing The method of processing the radar waveform target can quickly,efficiently and accurately check the target position.From this point of view,this paper studies algorithms suitable for image processing and target matching of civil ship radar.This paper first analyzes the characteristics of civil ship radar,studies the association algorithm of point track,and improves the existing problems based on the advantages and disadvantages of these algorithms to achieve the detection and tracking of the target in the image by the radar.In the researched image feature and target feature matching method,focusing on the gradient feature of the radar target,the target position,area,and moment invariant features are extracted as the target matching feature value,which is used for the multi-target radar image sequence Goal matching and goal tracking.Then the algorithm in the process of track data processing is studied.The commonly used methods of point agglomeration include amplitude weighted agglomeration algorithm and sorted point agglomeration algorithm.The existing point agglomeration method uses distance information to cluster bright spots with similar distances.,When two ships that are close to each other,the radar spots they reflect are very close or there is a certain degree of adhesion,the existing point trace will cluster it into a target,first propose a method that can effectively avoid the cause of point adhesion The target track is lost,the steps of the algorithm are introduced in detail,and then the common track association algorithms commonly used in radar are the nearest neighbor method(NN),the probabilistic data association method(PDAF),and the joint probabilistic data association method(JPDAF).Three algorithms are simulated to verify that the JPDAF tracking accuracy is better in multi-target cross motion.Compared with the new algorithm in this paper,the pixel gradient algorithm based on radar image assists the probabilistic data association method,and the point track navigation in the multi-target tracking clutter environment trace correlation algorithm works better.Finally,the software design is used to display the radar image.The radar video display uses the electronic chart as the background to display the target ship's trajectory.In the radar target point track track correlation method is to use image processing to achieve radar image target point track correlation to form a track,by simulating the movement of the ship in the channel such as cross motion,parallel,sparse and dense state The accuracy of the track association tests the practicability and effectiveness of the method,marking each target position,and displaying the target point positions of the radar image through the radar interface display system.