VESTRO: Velocity Estimation using Stereoscopic Vision

Abstract

This paper presents a real-time stereoscopic vision system that estimates the speed and direction of a moving object. Stereoscopic is applied in a real-time image processing system to re-construct a moving object in 3D, and further estimate the object’s velocity vector. The developed system, which is called Vestro, can be operated to measure the speed of any moving object, ranging from remote controlled vehicles, snooker balls to human beings, and determine where it is going. We show how 3D analytical geometry and vector operations are utilized to find the velocity magnitude, and the corresponding angular compass direction for a particular motion. In-lab experiments prove that the developed Vestro system produces velocity estimation with an accuracy of more than 94%.