基于图像处理技术的目标跟踪研究

基于视频的交通信息采集目前已成为计算机视觉领域新兴的研究方向,准确的运动目标跟踪是其关键技术并为智能交通系统奠定基础。详细介绍了几种常用的目标跟踪算法,并对目前算法相对成熟、匹配效果较好的Hausdorff算法深入研究,随后针对其计算量大的问题,采取了一种新策略,即Hausdorff算法与Kalman滤波器对目标运动进行预测估计相结合的方法进行运动目标的跟踪。

Research on Target Tracking Based on Image Processing

Video-based traffic information collection has become a new and rising research direction,and accurate moving target tracking technique is one significant foundation of intelligent traffic system.In the paper,some main tracking algorithms are introduced first,then the Hausdorff algorithm which is mature and of good matching effect is studied thoroughly.In order to reduce the amount of calculation of tracking algorithm based on Hausdorff distance,the paper adopts a new method for moving target tracking with Hausdorff algorithm combined with Kalman filter to predict the target moving location.