Robust visual tracking based on hierarchical appearance model

详细信息    查看全文

• 作者：Hua Bao^aAuthor Vitae ; Mingqiang Lin^bAuthor Vitae ; Zonghai Chen^a ; ^{chenzh@ustc.edu.cn}Author Vitae
• 关键词：Visual tracking ; Hierarchical appearance model ; Bayesian framework
• 刊名：Neurocomputing
• 出版年：2017
• 出版时间：19 January 2017
• 年：2017
• 卷：221
• 期：Complete
• 页码：108-122
• 全文大小：4500 K
• 卷排序：221

文摘

A hierarchical appearance model is proposed to represent the target effectively. An effective criteria is proposed to measure how likely a patch belong to the target. An adaptively model update mechanism is exploited to handle occlusion and etc. Visual saliency feature is introduced into the tracking framework.