Detection of pedestrians in road context for intelligent vehicles and advanced driver assistance systems

Abstract

Pedestrian detection is one of the key issues of the intelligent vehicles and advanced driver assistance systems (ADAS) used in the daily urban traffic. This paper addresses a system designed for finding the pedestrians in the road context, which can enhance the pedestrian detection performance based on the contextual correlations. More specifically, stereo vision is employed to seek the free road space based on a Markov Random Field (MRF). Such information is then used for correlation with the pedestrian detection procedure, which is based on a deformable part-based model with histogram of oriented gradient (HOG) features. Experimental results in various typical but challenging scenarios show the effectiveness of the proposed system.