3D reconstruction of transparent objects exploiting surface fluorescence caused by UV irradiation

In this paper, we present a novel approach exploiting fluorescence imaging to estimate the shape of transparent objects. Classical inspection systems require users to coat transparent objects with some powder before measurement. Methods suggested in literature through non contact measurement do not effectively deal with the refraction problem, thus, providing inaccuracies. The proposed method handles the scanning of transparent objects without using any powder and solving the refraction problem using UV environment. A classical triangulation method based on stereovision scheme using fixed stereoscopic visible range cameras with a fixed UV (Ultra Violet) laser source is implemented. Transparent object surface irradiated by UV laser emits fluorescence i.e. a visible white light in a diffused manner. Images consisting of fluorescent points are thereafter analyzed to extract 3D information of the transparent object surface with a stereovision scheme. The object is moved for scanning different parts and the above procedure is repeated.