Detection And Tracking Of Moving Ship In Bridge Area

With the rapid development of inland water transportation and the increase in thenumber of inland bridges, the risk of Ship-Bridge Collision is increasing. How to avoidthe occurrence of accidents is a very important topic in inland navigation. The thesis, fromthis point, the paper used the technique on image preprocessing, moving objects, targettracking and other related techniques to realize the real-time analysis of the captured videoimages. With that, the paper presented an algorithm of moving target detection andtracking based on the bridge area.Taking the example for Moving Ship in Bridge Area, The paper tries to research ondetection and tracking method. The main elements are as follows:First of all, in the target detection, the paper studies a variety of the conventionalmoving target detection methods, which based on the gray scale of pixels, including thecoterminous frames difference method, optical flow method, Gaussian mixture modelmethod. Experiments show that they are unsuitable for Application Environment in thispaper.Secondly, related to the Region Correlation, the paper studied the visual backgroundextraction moving object detection algorithm, studied of the Spatial-Temporal Saliencyfeature, local texture pattern feature and the features based on the Region Correlation indepth, which provides some effective candidate features to resolve this problem.Experiments show that local texture pattern feature is suitable for ApplicationEnvironment in this paper.At last, base on characteristic of conditions in Bridge Area. A Moving ship detectionand tracking method based on the adaptive texture features of the bridge area is presented.The background model is established by replacing brightness of LBP texture features toestablish the background model, and updated for every coming frame in real time.Foreground areas are extracted as moving ships by adaptive threshold. Finally, the movingships are tracked using simple nearest neighboring method.This paper, through the video surveillance Technology, can sufficiently acquire the position information of the ship in the bridge area and simultaneously prompt thefoundation on occlusion judgment and occlusion avoidance.