基于高效二阶最小化的运动模糊目标跟踪算法  

Motion Blur Target Tracking Method Based on Efficient Second-Order Minimization

作　　者：赵越[1] 李晶皎[1] 李贞妮[1] 马技[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110819

出　　处：《东北大学学报（自然科学版）》2013年第12期1678-1681,共4页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(60970157)

摘　　要：针对运动模糊严重时容易导致目标跟踪失败的问题,提出了基于高效二阶最小化(ESM)的模板匹配目标跟踪算法.首先提出一种运动模糊模板匹配的图像构造模型.然后引入ESM算法,在ESM算法基础上,用改进的高效二阶最小化(ESM-MB)算法跟踪运动模板.引入摄像头快门估计时间作为参数,提出了自适应不同快门速度所引起的不同的运动模糊的ESM-MB-ST跟踪算法.最后,通过真实视频序列的跟踪实验,验证了提出的ESM-MB算法及ESM-MB-ST算法具有更强的鲁棒性与实时性.As severe motion blur will lead to the failure of target tracking, an ESM-based template matching target tracking algorithm was proposed. First, an image structure model of motion-blurred template matching was presented. Then, the ESM algorithm was introduced and motion template was tracked based on ESM-MB. Subsequently, camera shutter estimated time was introduced as a parameter and self-adaptive ESM-MB-ST tracking algorithm of motion blur caused by different shutter speed was proposed. Finally, the real video sequence tracking experiments indicated that the proposed ESM-MB algorithm and the ESM-MB-ST algorithm had a better robustness and real-time ability.

关 键 词：高效二阶最小化 运动模糊 图像构造模型 快门时间 模板匹配 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]