Vision Based Vehicle-Pedestrian Detection and Warning System

Abstract

Road Sense must be respected and obeyed by both the pedestrian and the driver. Moreover, urbanization has led to a steadfast rise in the fleet of vehicles, their speed, as well as non-compliance with road safety measures, and other such factors have provoked an inescapable increase of accidents in road traffic involving pedestrians. Pedestrian collisions can be predicted and prevented. At the very basic, there has to be vehicle and pedestrian detection along with speed estimation, which can be further applied to Vehicle-Pedestrian Collisions and various emerging fields like Industrial Automation, Transportation, Automotive, Security/Surveillance, or in Dangerous environments. This paper reviews the literature on vehicle and pedestrian detection based on two significant categories: pre-processing phase and detection phase, with a detailed comparative analysis. The papers reviewed cover video-based surveillance systems.