Low resolution pedestrian detection using light robust features and hierarchical system

作者：

Highlights：

• The proposed PPMPF is utilized to initially filter out most of the non-pedestrian regions.

• Independent orientation and magnitude features are adopted as descriptors for pedestrians.

• A superior performance can be achieved compared to the former HOG and the edgelet.

摘要

•The proposed PPMPF is utilized to initially filter out most of the non-pedestrian regions.•Independent orientation and magnitude features are adopted as descriptors for pedestrians.•A superior performance can be achieved compared to the former HOG and the edgelet.

论文关键词：Pedestrian detection,Computer vision,Pattern recognition,Intelligent vehicle highway systems,AdaBoost

论文评审过程：Received 5 March 2013, Revised 8 September 2013, Accepted 3 November 2013, Available online 16 November 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.11.008