A novel low false alarm rate pedestrian detection framework based on single depth images

作者：

Highlights：

• Proposed a low false alarm rate pedestrian detection method

• Extracted foregrounds with the employment of depth sensor

• Designed a pre-evaluation stage to save computation time and reduce false alarm rate

• Built a pedestrian database with 673 depth images collected from 11 different scenes

摘要

•Proposed a low false alarm rate pedestrian detection method•Extracted foregrounds with the employment of depth sensor•Designed a pre-evaluation stage to save computation time and reduce false alarm rate•Built a pedestrian database with 673 depth images collected from 11 different scenes

论文关键词：Pedestrian detection,Histogram of Oriented Gradients,Shape context,Chamfer matching

论文评审过程：Received 20 October 2014, Revised 9 October 2015, Accepted 6 November 2015, Available online 3 December 2015, Version of Record 28 December 2015.

论文官网地址：https://doi.org/10.1016/j.imavis.2015.11.002