Pedestrians' Detection Methods in Video Images: A Literature Review

Abstract

This paper reviews current algorithms for pedestrian detection using image processing, where used images have been obtained from video surveillance or conventional cameras. In most of reviewed papers, detection is done to verify whether or not a pedestrian is identified, and if so, his or her position in the scene is calculated. A variety of novel algorithms are studied, which are based from the traditional detection techniques until the state of the art deep learning techniques. Therefore, we are reporting those pedestrian's detection methods performance, which have been experimented in controlled and uncontrolled environments.