中文摘要：

视频序列中的运动对象检测和跟踪算法在视频分析、计算机视觉、侦察、监控等诸多应用领域起着非常重要的作用.文中从低层图像特征角度对该问题进行了概述.对于运动对象检测,分析了基于结构化特征的进展、基于生物学研究、自底向上、激励驱动的可视化关注模型、综合时域信息的复合时空关注模型、DCT系数分析和运动场分析;对于运动对象跟踪,强调了粒子过滤系统、尺度不变特征变换（SIFT）,均值偏移,以及几者的组合应用.对压缩域运动对象跟踪也作为一个重要方面进行了分析,介绍了一种仅利用运动矢量的压缩域运动对象跟踪算法.

英文摘要：

Moving object detection and tracking in video sequences plays an important role in many applications such as video analysis,computer vision,reconnaissance,surveillance,etc.This paper gives a brief overview on this topic from the viewpoint of the low level image features.For object detection,structural feature-based approaches,stimuli-driven bottom-up visual attention models based on psychology studies,combined spatial-temporal attention model using temporal cues,DCT coefficients analysis and motion field analysis,etc.,are analyzed.For object tracking,particle filtering,scale invariant feature transform（SIFT）,mean shift and their hybrid combinations are emphasized.Since moving object tracking in compressed video stream is still a challenging problem,a novel idea for motion-based object tracking in compressed domain is introduced.