Determining surface orientations of transparent objects based on polarization degrees in visible and infrared wavelengths

Abstract

Techniques for modeling an object through observation are very important in object recognition and virtual reality. A wide variety of techniques have been developed for modeling objects with opaque surfaces, whereas less attention has been paid to objects with transparent surfaces. A transparent surface has only surface reflection; it has little body reflection. We present a new method for obtaining surface orientations of transparent surfaces through analysis of the degree of polarization in surface reflection and emission in visible and far-infrared wavelengths, respectively. This parameter, the polarization degree of reflected light at the visible wavelengths, is used for determining the surface orientation at a surface point. The polarization degree at visible wavelengths provides two possible solutions, and the proposed method uses the polarization degree at far-infrared wavelengths to resolve this ambiguity.

Shape and Refractive Index from Single-View Spectro-Polarimetric Images

“…Moreover, the mapping is not necessarily consistent across different material refractive indices. The related work in [27] and [26] employed the degree of polarisation in the visible and far-infrared regions to resolve ambiguities in determining the surface orientation of transparent objects. The main drawback of the method in [26] is the need for an omni-directional diffuse illumination source, which limits its applicability in real-world settings.…”

“…The related work in [27] and [26] employed the degree of polarisation in the visible and far-infrared regions to resolve ambiguities in determining the surface orientation of transparent objects. The main drawback of the method in [26] is the need for an omni-directional diffuse illumination source, which limits its applicability in real-world settings. Moreover, the method requires measurements of the energy emitted in the far-infrared spectrum, which is susceptible to thermal noise due to heating of the object under study.…”

A Combinatorial Transparent Surface Modeling from Polarization Images

“…Despite recent advances in opaque surface modeling, transparent surface modeling relatively has not received much attention. Only recently, some prospective techniques for modeling transparent or specular surface based on polarization images have emerged [2][3][4][5]. These techniques, however, commonly face two fundamental difficulties.…”

“…Such light setting (referred to as photometric sampler firstly proposed by Nayar, et al, [6]) has two limitations: it restricts the object's diameter to be sufficiently small compared to the diffuser's diameter and it suffered much from undesirable inter-reflections. The second difficulty, namely the ambiguity problem, is previously solved by introducing other sources of information such as thermal radiation [3], or new view image [2]. The necessity of such additional information that is not readily available leads to even more impractical and time consuming implementation.…”

Surface normal estimation of black specular objects from multiview polarization images

“…Drbohlav and Sara 10 and Ngo et al 11 solved the ambiguity problem of uncalibrated photometric stereo via polarization analysis and estimated both the light direction and the surface normal of a nonspecular object. Miyazaki et al 12 estimated the surface normal of a transparent object by analyzing the polarization state of the thermal radiation from the object. Miyazaki et al 13 attempted to estimate the surface normal of a diffuse object from a single view.…”