Research On Ship Moving Target Tracking Method Based On Improved Camshift Algorithm

The detection and tracking of moving ships can effectively monitor the situation at sea and improve the safety of the sea area.However,the accuracy of target location will be seriously disturbed by the complex and changeable sea conditions,severe wind weather,low visibility and sea light refraction.The shelter of ships also increases the difficulty to detect and track targets as well.Therefore,further optimization of moving ship target detection and tracking algorithm is the most important thing to improve the capability of marine monitoring.Aiming at the problems and difficulties in ship target detection and tracking technology,this paper mainly studies the detection and tracking of ship targets under occlusion,and analyzes and designs related algorithms to improve the accuracy of ship target detection and tracking.In the study,this article first demonstrated the preprocessing of the ship's target image,mainly including image graying,filtering and denoising,edge detection and morphological calculation.Then,it analyzed the basic methods used in moving target detection.In order to deal with the complex light changes at sea,a three-frame difference method based on the inter-frame difference method is researched and implemented,which can improve the accuracy of ship moving target detection based on the insensitivity to light changes.In the area of ship tracking,this paper proposes a "prediction-tracking" method.It selects a Kalman prediction model to predict the position of the ship's moving target at the next moment and performing target tracking,which effectively reduces the target search range.Camshift algorithms could effectively solved the problem that the tracking window cannot be adjusted.This article puts emphasis on the detection and tracking of targets in the case of shelter,and proposes a fusion algorithm with the ORB algorithm-CSO algorithm,which solves the problem that the Camshift algorithm cannot accurately track the ship target again after the targetocclusion occurs.Compared with the experimental results,the CSO algorithm proposed in this paper has a lower follow-up rate and higher accuracy which meets the real-time tracking requirements.Based on the design and improved algorithms,this project uses PTZ cameras and related software platforms to develop and design a ship moving target detection and tracking test system,which implements functions such as image acquisition,processing,and human interaction.The results show that the system can track realtime marine moving targets accurately,deal with sheltered target effectively and verify the feasibility of the system.