复杂背景下运动目标检测与实时跟踪

为准确追踪复杂背景下视频序列待处理目标的位置，提出一种基于互相关信息和Mean Shift相结合的目标追踪算法。将待处理帧和背景帧图像从RGB空间转化为GRAY空间，利用背景剪除法获取待匹配帧的基准图像，对基准图像用快速傅里叶变换得到基准图象和模版图像的相关性，结合Mean Shift最终确定目标图像的位置。实验结果表明，该方法在遮挡、环境干扰等复杂情况下均能稳定有效地追踪目标。

Motion Object Detection and Tracking on Complex Background

In order to track the moving object on a complex background ,this article proposes an algorithm of object track-ing based on the cross-correlation and Mean Shift .Firstly ,we convert a pending image and background image to Gray space from RGB space .Then ,making use of the background subtraction ,we get the source image .Thirdly ,wield fast fourier transform we get the cross-correlation properties .Finally ,we get the position of the target by cross-correlation properties and Meanshift .The experiments have shown that the algorithm efficiently and accurately track object on occlu-sion and environmental disturbing case .