Robust visual tracking via augmented kernel SVM

作者：

Highlights：

• Augmented Kernel Matrix (AKM) is applied to combine complementary features.

• AKM clustering is utilized to group the tracking results into a few aspects.

• Representative patches are selected to learn the appearance model.

• Adding representative patches our tracker more robust to abrupt appearance changes.

摘要

•Augmented Kernel Matrix (AKM) is applied to combine complementary features.•AKM clustering is utilized to group the tracking results into a few aspects.•Representative patches are selected to learn the appearance model.•Adding representative patches our tracker more robust to abrupt appearance changes.

论文关键词：Feature representation,Appearance model,Augmented Kernel Matrix (AKM)

论文评审过程：Received 5 March 2013, Revised 2 March 2014, Accepted 12 April 2014, Available online 22 April 2014.

论文官网地址：https://doi.org/10.1016/j.imavis.2014.04.008