海事执法船视频稳像与目标跟踪的仿真研究

Modeling and Simulation of Video-image Stabilization and Moving-object Tracking for Maritime Authority Ships

根据海事执法船舶的新需求,提出了一种将视频稳像与跟踪有机结合的新方法:利用小波变换对海天线进行辨识、提取;再以海天线为参照,通过图像几何变换和云台反馈实现视频的双重补偿稳像;同时,利用截断中心点与全局坐标的位置关系,通过Kalman滤波器对截断中心点进行估计,实现对目标的截断跟踪。该方法不同于已有的稳像、跟踪方法,它充分地利用了海天线的运动参数,在解决输出稳定视频信号的同时实现了对目标的自动锁定跟踪。

A technique of video-image stabilization and moving-object tracking plays an important role in maritime law enforcement specially for maritime authority ships. According to the new requirements of maritime authority ships, a new algorithm that could combine video-image stabilization with moving-object tracking was proposed. First, the technique identifies and extracts the horizontal line of the video-image using wavelet transformation. Then, referring to horizontal line, it adopts image transformation and video platform＇s feedback compensation to realize the principle of dual compensatory video-image stabilization （DCVS）. It also adopts Kalman filter to estimate the ubiety of truncation-center （TC） in the global coordinate system and finish the so-called truncation tracking （TT） simultaneously. Because this algorithm makes full use of hydraulic parameters of horizontal line to achieve video-image stabilization and moving-object tracking synchronously, the technique performs well. Simulation results show that the technique is effective.