Research On Marine Ship Detection And Tracking System And Its Key Algorithms

Before the ship identification verification,the ship target detection and tracking technology is employed for the image detail information acquisition suspicious ship including the ship size,position and the corresponding contour information.It provides the basis for the accurate extraction of ship features to realize the accurate classification and recognition of naval ships.After completing the recognition of the ships,the target detection and tracking technology is used to track the abnormal behaviors of the suspicious ships to achieve real-time and accurate monitoring of suspicious ships.In this paper,the characteristics of the target detection and tracking algorithm involves the sea surveillance are system atically analyzed,and the theme how to effectively detect and track suspicious behaviors in complex and changeable environments and ship's abnormal behaviors in real time is emphasized.On this purpose,some improvements for current used methods and innovative algorithm are proposed,and a marine ship detection and tracking system which is used for marine monitoring and testing is designed and developed by using those proposed technologies.The main contribution of this paper is summarized as follows:1.To detect and track the target of marine ship accurately and provide the basis for ship tracking,a ship target detection algorithm based on sea antenna is proposed in this paper.First,the maximum inter class variance and the Hough transform are used to extract sea antenna.Then,the Shi-Tomasi corner detection algorithm is used to find the approximate location of ships near Haitian line.Finally,the detection of ships on the sea is detected by setting the detection area and the corresponding threshold.This part lays a foundation for the accurate tracking research of the subsequent ships.2.To improve the real-time and robustness of the multi-target visual tracking algorithm in tracking ship.An improved automatic tracking algorithm for multi-target ships based on TLD(Tracking-Learning-Detection)is proposed in this paper.The goal of automatically real-time tracking ships at sea have been achieved in the algorithm.Firstly,the detection and extraction method are proposed to split sea-sky-line.By using Otsu threshold segmentation and Hough transform,the sea-sky-line is easily extracted;Secondly,through using the principle of Kalman filter,the target ships near the sea-sky-line has been positioned and detected in the algorithm,then segregate the target ships.Finally,the initial target tracking frame has been generated with its minimum bounding rectangle by the extracted target,the tracker uses the position coordinate information of the initial tracking frame to automatically track the target in real time.During the running of the algorithm,scanning detection is carried out near the sea-sky-line,which reduces the traversing range and improves the real-time performance of the algorithm.Experimental analysis shows,by comparing with the traditional Mean-shift algorithm and the previous target tracking algorithm called TLD,the proposed algorithm is more accurate and more efficient in ship tracking.3.To further detect and tracking the abnormal behavior of personnel on the near-by ships,an improved hybrid Gauss based scene anomaly detection and tracking algorithm is proposed in this paper.Firstly,the improved hybrid Gauss modeling idea is used to detect the target in the sensitive area initially,and then the optical flow method and corner detection method is combined to accurately locate and detect the sensitive area personnel.And track the abnormality of the ship's personnel.Achieve the effect of accurate detection and tracking of abnormal behavior of suspicious ship personnel at close range.It is convenient for our ship personnel to understand the activities of suspected ship personnel in a timely manner by tracking the abnormal behavior of the suspected warship.Based on the above research,the marine ship detection and tracking system is designed and developed in this paper.The design thought,overall structure,function modules and key technologies are introduced in detail.The experimental results show that the proposed system can accurately and effectively detect,track and identify marine vessels and track abnormal behavior of ships.