一种基于视频运动分析的智能码流方法  

An Intelligent Stream Method Based on Vedio Motion Analysis

作　　者：翟艳东[1] 于明[1] 阎刚[1] 

机构地区：[1]河北工业大学计算机科学与软件学院,天津300401

出　　处：《科学技术与工程》2013年第33期9987-9991,10012,共6页Science Technology and Engineering

摘　　要：采用迭代重加权最小二乘(IRLS)方法计算连续两帧图像的光流。根据光流计算的结果估计视频监控场景的运动状态。将场景分为静止、一般运动、剧烈运动,并分别制定不同的码流存储策略。实验表明,该方法能对监控场景的常见运动行为识别并分类,具有一定的鲁棒性。应用该方法可以减少冗余视频帧对存储资源的占用,扩充敏感监控数据的存储能力。Iteratively reweighted leassquare（IRLS） method iused to calculate two consecutive image optical flow, According to the results, the motion state of the video surveillance sceneiestimated, and divided them into static, general sportand strenuouexercise. Then differenbistream storage strategy iused to save the differenscenes. Experimentshow thathe method can recognize and classify the common motobehavioof the monitoring scene with robustness. Thimethod can reduce the redundanvideo frame storage resource occupation, and expand the storage capacity of sensitive video data.

关 键 词：光流 迭代重加权最小二乘 运动分析 智能码流 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]