Refractive Height Fields from Single and Multiple Images

Qi Shan , Sameer Agarwal, and  Brian Curless

CVPR12 Paper [pdf (5M)]
CVPR12 Poster [pdf (41M)]
ICCP12 Fast forward slides [ppt (2M)]

Abstract: We propose a novel framework for reconstructing homogenous, transparent, refractive height-fields from a single viewpoint. The height-field is imaged against a known planar background, or sequence of backgrounds. Unlike existing approaches that do a point-by-point reconstruction – which is known to have intractable ambiguities – our method estimates and optimizes for the entire height-field at the same time. The formulation supports shape recovery from measured distortions (deflections) or directly from the images themselves, including from a single image. We report results for a variety of refractive height-fields showing significant improvement over prior art.

From multiple images

A photo of a flower shaped refractive object.   Mutiple images are taken from a same viewpoint and with different background patterns.   A sparse set of deflection vectors computed from these captured patterns.   The reconstructed shape of the flower glass.

From a single image

A single image is taken with a specifically designed color pattern. The reconstructed shape.


*** End ***