Real-time tracking-by-learning with high-order regularization fusion for big video abstraction》.pdf

Signal Processing ] (]]]]) ]]]–]]] Contents lists available at ScienceDirect Signal Processing journal homepage: /locate/sigpro Real-time tracking-by-learning with high-order regularization fusion for big video abstraction Peng Zhang a,n, Tao Zhuo a, Yanning Zhang a, Lei Xie a, Dapeng Tao b a School of Computer Science, Northwestern Polytechnical University, Xi an, PR China b School of Information Science and Engineering, Yunnan University, Kunming, PR China a r t i c l e i n f o a b s t r a c t Article history: Visual tracking is a key technique used by video abstraction to achieve efficient post- Received 24 May 2015 analysis for big video surveillance. In order to tackle the problem of constantly changing Received in revised form scenarios during online tracking, additional factors such as motion can be incorporated by 27 July 2015 utilizing a fusion strategy to improve the final performance. Unfortunately, straightfor- Accepted 31 July 2015 ward output fusion is difficult to be synchronized due to the diversity in model regression. Therefore, a widely cited problem for learning based fusion is to incorporate regularizers Keywords: for label assignment of unlabeled samples, which is one of the major research focuses on Video abstraction semi-supervised learning. In this paper, a novel tracking strategy based on semi- Learni