Pedestrian detection in single frame by edgelet-LBP part detectors

This paper proposes a method for human detection in crowded scene from static images. We introduce to combine edgelet and LBP features to obtain more discriminative representations for local area. To cope with partial occlusion, part detectors are learned using real AdaBoost in bootstrap way. Responses of part detectors are combined to form the final results. We test our approach on several common datasets and compare the proposed method with others. The experimental results prove that our method is comparable to the state-of-the-art method and performs well on crowded scenes.