What does a single light-ray reveal about a transparent object?

We address the following problem in refractive shape estimation: given a single light-ray correspondence, what shape information of the transparent object is revealed along the path of the light-ray, assuming that the light-ray refracts twice. We answer this question in the form of two depth-normal ambiguities. First, specifying the surface normal at which refraction occurs constrains the depth to a unique value. Second, specifying the depth at which refraction occurs constrains the surface normal to lie on a 1D curve. These two depth-normal ambiguities are fundamental to shape estimation of transparent objects and can be used to derive additional properties. For example, we show that correspondences from three light-rays passing through a point are needed to correctly estimate its surface normal. Another contribution of this work is that we can reduce the number of views required to reconstruct an object by enforcing shape models. We demonstrate this property on real data where we reconstruct shape of an object, with light-rays observed from a single view, by enforcing a locally planar shape model.

[1]  Ding Liu,et al.  Frequency-Based 3D Reconstruction of Transparent and Specular Objects , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Jingyi Yu,et al.  Reconstructing Gas Flows Using Light-Path Approximation , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Ohad Ben-Shahar,et al.  Toward a Theory of Shape from Specular Flow , 2007, 2007 IEEE 11th International Conference on Computer Vision.

[4]  Feng Li,et al.  Dynamic fluid surface acquisition using a camera array , 2011, 2011 International Conference on Computer Vision.

[5]  Gordon Wetzstein,et al.  Hand-held Schlieren Photography with Light Field probes , 2011, 2011 IEEE International Conference on Computational Photography (ICCP).

[6]  Kiriakos N. Kutulakos,et al.  A Theory of Refractive and Specular 3D Shape by Light-Path Triangulation , 2005, ICCV.

[7]  G. Settles.  Schlieren and shadowgraph techniques , 2001 .

[8]  Svenja Lowitzsch,et al.  Full-Field Shape Measurement of Specular Surfaces , 2006 .

[9]  Mathieu Salzmann,et al.  Mirror Surface Reconstruction from a Single Image , 2015, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[10]  Hans-Peter Seidel,et al.  3D acquisition of mirroring objects using striped patterns , 2005, Graph. Model..

[11]  Cong Phuoc Huynh,et al.  Shape and refractive index recovery from single-view polarisation images , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[12]  Pau Gargallo,et al.  General Specular Surface Triangulation , 2006, ACCV.

[13]  Frédo Durand,et al.  Refraction Wiggles for Measuring Fluid Depth and Velocity from Video , 2014, ECCV.

[14]  Wolfgang Heidrich,et al.  Stochastic tomography and its applications in 3D imaging of mixing fluids , 2012, ACM Trans. Graph..

[15]  Gordon Wetzstein,et al.  Refractive shape from light field distortion , 2011, 2011 International Conference on Computer Vision.

[16]  Aswin C. Sankaranarayanan,et al.  Image Invariants for Smooth Reflective Surfaces , 2010, ECCV.

[17]  David J. Kriegman,et al.  The Bas-Relief Ambiguity , 2004, International Journal of Computer Vision.

[18]  Michael Weinmann,et al.  Multi-view Normal Field Integration for 3D Reconstruction of Mirroring Objects , 2013, 2013 IEEE International Conference on Computer Vision.

[19]  Visesh Chari,et al.  A Theory of Refractive Photo-Light-Path Triangulation , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[20]  Qionghai Dai,et al.  Transparent Object Reconstruction via Coded Transport of Intensity , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[21]  Aswin C. Sankaranarayanan,et al.  Specular surface reconstruction from sparse reflection correspondences , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.