Distance Measurement of Objects using Stereo Vision

Abstract

Vision based distance measurement of objects is continues to be a popular research topic in computer and robot vision nowadays. Measuring distance of targets are important for robot navigation, obstacle detection, navigation of autonomous vehicles, surveillance monitoring, person localization and tracking, and many more. In this paper, we propose an efficient approach for distance measurement of objects using stereo vision. The distance is measured from the estimated depth information of the object from its two image sequences (called left and right image pair) captured by a stereo camera. In order to estimate the depth information we employ a fast stereo algorithm. Experimental evaluation proves the efficiency of our proposed approach comparable to other related methods.