Introducing a novel marker-based geometry model in monocular vision

Abstract

A spherical marker-based distance capture concept using monocular vision is presented in this paper. A novel method is explored, within the concept of a virtual sphere, which shows how to improve the reading measurements of the distance of a moving object from low resolution digital images, and from a single viewpoint. The aim here is to be able to track accurately the object at a furthest possible position. A conclusion with experimental simulations carried out using 3D modeling of markers and representing the real world showing the potency of the marker's geometry on improving the accuracy of the measurements. A potential application field of the proposed method is the implementation of tracking object in mobile robots, marker-based localization, and field of topography.