3D Shape and Indirect Appearance by Structured Light Transport
Matthew O'Toole, John Mather, and Kiriakos N. Kutulakos. CVPR, 2014.
Abstract:
We consider the problem of deliberately manipulating the direct and indirect light flowing through a time-varying, fully-general scene in order to simplify its visual analysis. Our approach rests on a crucial link between stereo geometry and light transport: while direct light always obeys the epipolar geometry of a projector-camera pair, indirect light overwhelmingly does not. We show that it is possible to turn this observation into an imaging method that analyzes light transport in real time in the optical domain, prior to acquisition. This yields three key abilities that we demonstrate in an experimental camera prototype: (1) producing a live indirect-only video stream for any scene, regardless of geometric or photometric complexity; (2) capturing images that make existing structured-light shape recovery algorithms robust to indirect transport; and (3) turning them into one-shot methods for dynamic 3D shape capture.
3D Shape and Indirect Appearance by Structured Light Transport
1. 3D Shape and Indirect Appearance
by Structured Light Transport
Matthew O’Toole John Mather Kiriakos N. Kutulakos
Department of Computer Science
University of Toronto
http://www.dgp.toronto.edu/~motoole/slt
7. using epipolar geometry to analyze light transport
new live video channels for computer vision
• indirect-only and direct-only imaging
• indirect-invariant imaging of structured patterns
• one-shot, multi-pattern imaging
contributions
8. using epipolar geometry to analyze light transport
new live video channels for computer vision
• indirect-only and direct-only imaging
• indirect-invariant imaging of structured patterns
• one-shot, multi-pattern imaging
structured light transport for 3D shape acquisition
contributions
28. camera projector
mirror
1. open electronic shutter
2. for i = 1 to N
use random epipolar mask &
project complementary pattern
3. close electronic shutter
blocking epipolar paths with patterns & masks
complementary random
epipolar patterns
30. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
31. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
direct-only imaging
32. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
direct-only imaging
33. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
direct-only imaging
34. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
direct-only imaging
35. live indirect-only and direct-only video stream
conventional imaging indirect-only imaging
direct-only imaging
36. separating direct & indirect components [e.g. Achar et al. 2013]
• multiple shots, motion processing
• only low-frequency indirect light (i.e. diffuse)
transport-robust shape acquisition [e.g. Gupta et al. 2012]
• assumption: only high-freq patterns can be robust to indirect light
• no specular transport
structured light transport
• live video, fully-general transport, no motion processing
• any projection pattern can be made robust to indirect light
• new forms of one-shot 3D imaging
related work
39. transport matrixmask
illumin.
deriving mask and
illumination patterns
• exact derivation of patterns
is NP-hard [Zhong 2012]
• instead, use random codes,
and approximate imaging in
expectation
structured light transport: theory
dominance of non-
epipolar transport
57. • a new general imaging technique for computer vision
• combines epipolar geometry and light transport analysis
• theoretical analysis & a physical device for live video
generation
• SIGGRAPH 2014: extension to time-of-flight imaging
structured light transport
58. 3D Shape and Indirect Appearance
by Structured Light Transport
Matthew O’Toole John Mather Kiriakos N. Kutulakos
Department of Computer Science
University of Toronto
http://www.dgp.toronto.edu/~motoole/slt
Poster ID: O-4A-4