The Research Of Visual Tracking Control System Based On AIS For Vessels In Controlled Waterway

With the rapid development of inland shipping industry, the number of ships risecontinually, and navigation pressure increase greatly. In addition, more curved andnarrow controlled waterway together with illegal sailing of some vessels easily causeinland congestion and traffic accidents frequently. The command and control of ships’navigation mainly rely on lookout and VHF communications, both lacking of vessels’navigation status and image information, and consuming of manpower, material andinefficient. According to the construction requirement of digital waterway, theluxuriant vessels image information is conducive to making decision for navigation,improving navigation efficiency, reducing accidents and strengthening lawenforcement under the construction of Intelligent Assistant Command System forsignal station in controlled waterway. Therefore, this paper researches an intelligentvessels tracking and monitoring system that merge AIS and CCTV technology.The visual tracking and control system for vessels is proposed to solve the aporiaof visual tracking for target ships in controlled waterway and realize the intellengce ofvessels visual tracking system. The voyage status information was collected anddecoded automatically by the system, and the local visual coordinate system wasestablished. In the process of target ship tracking, it’s firstly needs to complete initialalignment for target ship, make dynamic prediction on the basis of navigation statusparameters. In addition, image recognition method is supplemented to improve theaccuracy. Meanwhile, according to the magnification of target image and objectdistance, the focus is adjusted automatically. The main research contents of this paperas follows:(1) The development status of vessels visual tracking technology in inlandwaterway was analysed. The platform of target ship tracking control is designed for thejoint actions of AIS and digital PTZ cameras, and the equipment selection wascompleted.(2) On the basis of local visual coordinate system, visual tracking controlalgorithm was proposed. The track angle forecasted in equal time intervals is used forsmoothing image and eliminating lag while static alignment tracking, and make effortsto solve the error problem. The effectiveness of the algorithm is validated bysimulation. (3) According to the higher real-time requirement of visual tracking andmonitoring system, the specific workflow of threads is designed after constructing asoftware framework based on multithreading. The acquisition and decoding of AISinformation is made an in-depth research. Meanwhile, the real-time video capturebased on network video servicer has already been implemented.