基于视频处理技术的运动轨迹实时跟踪研究  

Research on motion trajectory real⁃time tracking based on video processing technology

作　　者：向文红 XIANG Wenhong(Guizhou Normal University,Guiyang 550001,China)

机构地区：[1]贵州师范大学,贵州贵阳550001

出　　处：《现代电子技术》2021年第9期48-52,共5页Modern Electronics Technique

摘　　要：针对当前运动轨迹跟踪的实时性差、跟踪准确性低等难题,以提高运动轨迹跟踪整体性能为目标,提出基于视频处理技术的运动轨迹实时跟踪方法。首先对视频序列图像进行预处理,包括图像灰度化、图像滤波以及图像增强等;然后利用改进的背景差分法进行运动目标识别与检测;最后利用卡尔曼滤波法进行运动目标特征匹配,实现目标运动轨迹实时跟踪。结果表明,该方法运动目标检测准确性高,且运动轨迹跟踪误差小,证明了方法的有效性,具有很好的应用价值。In view of the poor real⁃time performance and low accuracy of the current motion trajectory tracking,a real⁃time tracking method based on video processing technology is proposed to improve the overall performance of motion trajectory tracking.The video sequence images are preprocessed,including image graying,image filtering and image enhancement.And then,the improved background difference method is used to recognize and detect the motion object.Finally,Kalman filtering method is used to perform feature matching of the motion object,so as to realize the real⁃time tracking of the object motion trajectory.The results show that,the detection accuracy of the motion object is high,and the tracking error of the motion trajectory is small,which prove the effectiveness of the proposed method.Therefore,it has a good application value.

关 键 词：跟踪方法 运动轨迹 视频处理技术 目标识别 目标检测 特征匹配 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP152.5[电子电信—信息与通信工程]