Detect foreground objects via adaptive fusing model in a hybrid feature space

作者：

Highlights：

• Proposed a novel spatiotemporal scale invariant ternary pattern descriptor.

• Adaptively fusing the color and texture features calculates the pixel׳s probability.

• Using the lateral inhibition filter to smooth the pixel׳s probability score.

摘要

Highlights•Proposed a novel spatiotemporal scale invariant ternary pattern descriptor.•Adaptively fusing the color and texture features calculates the pixel׳s probability.•Using the lateral inhibition filter to smooth the pixel׳s probability score.

论文关键词：Foreground detection,ST-SILTP,Adaptive fusion model,Lateral inhibition filter

论文评审过程：Received 14 June 2013, Revised 19 January 2014, Accepted 18 March 2014, Available online 29 March 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.03.016