Vision-based Real-time Pedestrian Detection for Autonomous Vehicle
Xin Liu, Dai Bin, HE Han-gen · 2007
This paper presents a real-time single-frame pedestrian detection approach. Combining efficient interesting regions selection and proper SVM classifier, the method is applicable to the autonomous vehicles running on urban roads. Experiment results with test dataset extracted from real driving on urban roads are presented to illustrate the performance of this approach.