Estimating the distance of a human from an object using 3D image reconstruction

Abstract

Obstacle detection, pedestrian detection, human motion detection, etc., are the recent technologies which are booming high in the computer vision industry. These technologies play a major role in various applications like surveillance, autonomous cars, driverless vehicles, etc. Our focus is on identifying and calculating the distance of a human from an object using 3D image reconstruction. The object can be a video camera or a sensor in case of surveillance or driverless vehicle. Using this distance, intelligent decisions could be taken. This methodology can help the computer vision researchers to detect any obstacle in their region of interest and also their distance from the particular point. This work could also detect the frame in which the person or the obstacle is detected along with the distance. The above said methodology could be incorporated in traffic monitoring system for identifying or detecting the pedestrians so that accidents could be avoided.