Achuta Kadambi, Vage Taamazyan, Boxin Shi, Ramesh Raskar

Abstract

Coarse depth maps can be enhanced by using the shape information from polarization cues. We propose a fusion of polarization and depth, a synergistic technique to exploit the high lateral resolution and compensate for the low depth resolution of traditional sensors. Although the fusion of depth and photometric normals is not new, existing frameworks cannot handle the unique challenge of fusing normals from polarization. Specific challenges we address include azimuthal ambiguity, refractive distortion, fronto-parallel surfaces, and depth discontinuities. In this project we validate our algorithmic framework for depth enhancement through a low-cost prototype that compares favorably with other depth sensors.