Polarized 3D: High-Quality Depth Sensing with Polarization Cues

Abstract

Coarse depth maps can be enhanced by using the shape information from polarization cues. We propose a framework to combine surface normals from polarization (hereafter polarization normals) with an aligned depth map. Polarization normals have not been used for depth enhancement before. This is because polarization normals suffer from physics-based artifacts, such as azimuthal ambiguity, refractive distortion and fronto-parallel signal degradation. We propose a framework to overcome these key challenges, allowing the benefits of polarization to be used to enhance depth maps. Our results demonstrate improvement with respect to state-of-the-art 3D reconstruction techniques.

Linear Depth Estimation from an Uncalibrated, Monocular Polarisation Image

“…However, this approach requires at least 4 light directions in contrast to the single direction required by our method. Very recently, Kadambi et al [3] proposed an interesting approach in which a single polarisation image is combined with a depth map obtained by an RGBD camera. The depth map is used to disambiguate the normals and provide a base surface for integration.…”

“…Hence, previous work focuses on developing heuristics to locally disambiguate the surface normals. Even having done so, surface orientation is only a 2.5D shape cue and so the estimated normal field must be integrated in order to recover surface depth [2] or used to refine a depth map captured using other cues [3]. This two step approach of disambiguation followed by integration means that the surface integrability constraint is not enforced during disambiguation and also that errors accumulate over the two steps.…”

Surface normal estimation of black specular objects from multiview polarization images

“…For example, Miyazaki et al, 14 Kadambi et al, 31 and several other researchers used DOP for estimating the surface normal from specular reflection. However, DOP depends on the refractive index and surface roughness.…”

“…The dense surface normal of a specular black object cannot be obtained using the Helmholtz stereo method because of the discretized sampling of the light source. Kadambi et al 31 combined the 3-D geometry obtained by a time-of-flight (ToF) sensor and the surface normal obtained from the DOP. Unlike space carving, which can be applied to a completely black object, a ToF sensor cannot measure such objects because the laser does not reflect at a black surface.…”

Data-Fusion for a Vision-Aided Radiological Detection System: Sensor dependence and Source Tracking

“…The Velodyne uses Light Detection and Ranging (LiDAR) technology, which measures distances using Time of Flight (TOF) and short laser pulses. Massachusetts Institute of Technology is currently working on depth sensing with polarization cues, which will drastically improve cheap 3D imaging sensors [5]. This paper will only cover the results using the Velodyne HDL-32E and the EJ-309 liquid scintillators.…”