期刊文献+

结合帧差法与Mean Shift的抗遮挡跟踪算法 被引量:7

Anti-occlusion tracking algorithm combining frame difference method and Mean Shift
下载PDF
导出
摘要 针对目标严重遮挡后,运动状态发生改变时,传统的基于运动预测的算法无法有效跟踪的问题,提出一种基于帧差法的改进算法。引入巴氏系数(Bhattacharyya)作为目标是否发生遮挡的判据;当发生遮挡时,帧差法检测目标,再次检测到目标时将此位置作为Mean Shift迭代的起始位置;最后正常跟踪时采用卡尔曼滤波预测目标位置,减少迭代次数。实验结果表明,当目标在严重遮挡后,运动状态改变时,基于运动预测的算法将无法跟踪目标,改进算法能够重新跟踪目标。 Since the traditional motion prediction based algorithm cannot conduct effective tracking when the moving state of the target changes after severe occlusion,an improved algorithm based on the frame difference method is proposed. Bhattacharyya is introduced as the judgment criterion of whether the target is occluded or not. The frame difference method is used to detect the target when occlusion occurs. The position where the target is again detected is taken as the starting position of the Mean Shift iteration. The Kalman filtering is used to predict the target position during normal tracking,so as to reduce iteration times. The experimental results show that the improved algorithm can track the target again while the motion prediction based algorithm cannot track the target when the moving state of the target changes after severe occlusion.
作者 岳昊恩 袁亮 吕凯 YUE Hao’en;YUAN Liang;Lü Kai(School of Mechanical Engineering,Xinjiang University,Urumqi 830047,China)
出处 《现代电子技术》 北大核心 2019年第12期180-182,186,共4页 Modern Electronics Technique
基金 新疆维吾尔自治区重点研发任务专项(2018B02011) 国家自然科学基金项目(31460248) 国家自然科学基金项目(61662075) 新疆维吾尔自治区科技支疆项目(2017E0284) 乌鲁木齐市科技人才计划(P151010006)~~
关键词 目标检测 跟踪算法 卡尔曼滤波 Mean SHIFT 帧差法 严重遮挡 target detection tracking algorithm Kalman filtering Mean Shift frame difference method severe occlusion
  • 相关文献

参考文献7

二级参考文献85

共引文献355

同被引文献59

引证文献7

二级引证文献52

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部