文档介绍：智能视频监控中目标跟踪算法研究及应用
摘要:视频监控中目标跟踪算法的研究及应用是实现视频监控系统的关键技术之一。实际视频监控系统中的场景往往是错综复杂、变化无常的,实现具有鲁棒性、准确性和实时性的运动目标检测和跟踪是当前智能视频监控技术努力的方向。本文研究了静态场景下运动目标跟踪的相关算法,给出了运动目标跟踪的实现方案,由于Lucas Kanade光流跟踪算法对目标仅仅跟踪其质心,存在跟踪目标容易丢失且鲁棒性不高的缺点;最小绝对差算法企图搜索目标模板、计算量大、实时性差;针对这两种算法在目标跟踪方面的不足,对其进行了相应的改进和完善。通过Lucas Kanade光流跟踪算法预测目标质心位置、并设定目标搜索区域减小最小绝对差算法计算量;将模板匹配定位之后更正的目标质心做为下一帧Lucas Kanade光流跟踪算法跟踪的特征点,可以减小单一光流法预测目标质心带来的误差,实现可靠的跟踪。
关键词:运动目标跟踪;Lucas Kanade光流法;模板匹配
Abstract:Video target tracking algorithm research and Application is one of the key techniques which implements intelligent video surveillance system(IVSS). The real video surveillance system applied environment is plex, full of variety. The realization of moving object detection and tracking with veracity, real-time performanceand robustness is the direction for current IVSS paper studied under static scene tracking algorithms movement are given, and the implementation scheme of moving object tracking Kanade light flow, because Lucas to target tracking algorithm only tracking its existing track the target centroid, easy to lose and robustness is not high weaknesses; Minimum absolute difference algorithm map search target template, large amount of calculation, real-time poor; In view of these two algorithm in target tracking deficiency, analyses the corresponding improvement and perfection. By Lucas Kanade light flow forecast target centroid position tracking algorithm, and set a target of reducing the search putation minimum absolute difference algorithm; Will the template matching after the target centroid position correction as the next frame Lucas Kanade light streaming tracking algorithm tracking feature points can reduce a single light flow method for forecasting the errors caused by target centroid, achieve reliable tracking.
Key words : Sports target tracking;Lucas Kanade light flow method;Background model
研究背景和意义
随着计算机技术、通信技术、图像处理技术的不断发展,目标检测与跟踪技术在移动机器人、战车与坦克、飞机、导弹、舰船等军事领域和航空航天、科学探测、天文观测以及视频监控等民用领域具有越来越广泛的应用,也