Top Banner
Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft Research, Redmond, USA
79

Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Dec 19, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Reconstructing Building Interiors from Images

Yasutaka Furukawa Brian Curless Steven M. SeitzUniversity of Washington, Seattle, USA

Richard SzeliskiMicrosoft Research, Redmond, USA

Page 2: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Reconstruction & Visualizationof Architectural Scenes

• Manual (semi-automatic)– Google Earth & Virtual Earth– Façade [Debevec et al., 1996]– CityEngine [Müller et al., 2006, 2007]

• Automatic– Ground-level images [Cornelis et al., 2008, Pollefeys et al., 2008]

– Aerial images [Zebedin et al., 2008]

Google Earth Virtual Earth Zebedin et al.Müller et al.

Page 3: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Reconstruction & Visualizationof Architectural Scenes

• Manual (semi-automatic)– Google Earth & Virtual Earth– Façade [Debevec et al., 1996]– CityEngine [Müller et al., 2006, 2007]

• Automatic– Ground-level images [Cornelis et al., 2008, Pollefeys et al., 2008]

– Aerial images [Zebedin et al., 2008]

Google Earth Virtual Earth Zebedin et al.Müller et al.

Page 4: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Reconstruction & Visualizationof Architectural Scenes

• Manual (semi-automatic)– Google Earth & Virtual Earth– Façade [Debevec et al., 1996]– CityEngine [Müller et al., 2006, 2007]

• Automatic– Ground-level images [Cornelis et al., 2008, Pollefeys et al., 2008]

– Aerial images [Zebedin et al., 2008]

Google Earth Virtual Earth Zebedin et al.Müller et al.

Page 5: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Reconstruction & Visualizationof Architectural Scenes

Google Earth Virtual Earth Zebedin et al.Müller et al.

Little attention paid to indoor scenes

Page 6: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Our Goal• Fully automatic system for indoors/outdoors– Reconstructs a simple 3D model from images– Provides real-time interactive visualization

Page 7: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

What are the challenges?

Page 8: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Challenges - Reconstruction

• Multi-view stereo (MVS) typically produces a dense model

• We want the model to be– Simple for real-time interactive visualization of a

large scene (e.g., a whole house)– Accurate for high-quality image-based rendering

Page 9: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Challenges - Reconstruction

• Multi-view stereo (MVS) typically produces a dense model

• We want the model to be– Simple for real-time interactive visualization of a

large scene (e.g., a whole house)– Accurate for high-quality image-based rendering

Simple mode is effective for compelling visualization

Page 10: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Challenges – Indoor Reconstruction

Texture-poor surfaces

Page 11: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Challenges – Indoor Reconstruction

Texture-poor surfaces Complicated visibility

Page 12: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Challenges – Indoor Reconstruction

Texture-poor surfaces Complicated visibility

Prevalence of thin structures(doors, walls, tables)

Page 13: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Outline

• System pipeline (system contribution)• Algorithmic details (technical contribution)

• Experimental results• Conclusion and future work

Page 14: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images

Images

Page 15: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Structure-from-Motion

Images

Bundler by Noah SnavelyStructure from Motion for unordered image collectionshttp://phototour.cs.washington.edu/bundler/

Page 16: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM

Page 17: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM

PMVS by Yasutaka Furukawa and Jean PoncePatch-based Multi-View Stereo Softwarehttp://grail.cs.washington.edu/software/pmvs/

Multi-view Stereo

Page 18: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Page 19: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 20: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 21: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 22: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 23: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 24: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS

Manhattan-world Stereo[Furukawa et al., CVPR 2009]

Page 25: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS MWS

Page 26: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS MWS

Axis-aligned depth map merging(our contribution)

Page 27: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

System pipeline

Images SFM MVS MWS Merging

Rendering: simple view-dependent texture mapping

Page 28: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Outline

• System pipeline (system contribution)• Algorithmic details (technical contribution) • Experimental results• Conclusion and future work

Page 29: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 30: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 31: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 32: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 33: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 34: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Basic framework is similar to volumetric MRF [Vogiatzis 2005, Sinha 2007, Zach 2007, Hernández 2007]

Page 35: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Page 36: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & data

Page 37: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

Page 38: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

• Weak regularization at interesting places

• Focus on a dense model

Page 39: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

• Weak regularization at interesting places

• Focus on a dense model

• We want a simple model

Page 40: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

Page 41: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

Page 42: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Unary encodes data

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

Page 43: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Binary is smoothnessUnary encodes data

Binary penalty

Binary encodes smoothness & dataUnary is often constant (inflation)

Page 44: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 1 - Penalty terms

Regularization becomes weakDense 3D model

Regularization is data-independentSimpler 3D model

Binary penalty

Page 45: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

Page 46: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)

Page 47: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)

Page 48: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)

Page 49: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)• Smoothness (binary)

Page 50: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)• Smoothness (binary)

Page 51: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Axis-aligned Depth-map Merging

• Align voxel grid withthe dominant axes

• Data term (unary)• Smoothness (binary)• Graph-cuts

Page 52: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

• For large scenes, data info are not complete

Page 53: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

• For large scenes, data info are not complete

• Typical volumetric MRFs bias to general minimal surface [Boykov and Kolmogorov, 2003]

• We bias to piece-wise planar axis-aligned for architectural scenes

Page 54: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Page 55: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Page 56: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Page 57: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Page 58: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Page 59: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Same energy (ambiguous)

Page 60: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Same energy (ambiguous)Data penalty: 0

Page 61: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Same energy (ambiguous)Data penalty: 0 Smoothness penalty: Data penalty: 0 Smoothness penalty: 24Data penalty: 0

Page 62: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

shrinkage

Page 63: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 2 – Regularization

Axis-aligned neighborhood + Potts model

Ambiguous

Break ties with the minimum-volume solution

Piece-wise planar axis-aligned model

Page 64: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 3 – Sub-voxel accuracy

Page 65: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 3 – Sub-voxel accuracy

Page 66: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 3 – Sub-voxel accuracy

Page 67: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Key Feature 3 – Sub-voxel accuracy

Page 68: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Summary of Depth-map Merging

• For a simple and axis-aligned model– Explicit regularization in binary– Axis-aligned neighborhood system & minimum-

volume solution• For an accurate model– Sub-voxel refinement

Page 69: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Outline

• System pipeline (system contribution)• Algorithmic details (technical contribution)• Experimental results• Conclusion and future work

Page 70: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Kitchen - 22 images1364 triangles

hall - 97 images3344 triangles

house - 148 images8196 triangles

gallery - 492 images8302 triangles

Page 71: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.
Page 72: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Demo

Page 73: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Running Time

Kitchen (22 images) Hall (97 images) House (148 images) gallery (492 images)

SFM 13 76 92 716

MVS 38 158 147 130

MWS 39.6 281.3 843.6 5677.4

Merging 0.4 0.4 3.6 22.4

Running time of 4 steps [min]

Page 74: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Conclusion & Future Work

• Conclusion– Fully automated 3D reconstruction/visualization

system for architectural scenes– Novel depth-map merging to produce piece-wise

planar axis-aligned model with sub-voxel accuracy• Future work– Relax Manhattan-world assumption– Larger scenes (e.g., a whole building)

Page 75: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

For More Details

Please refer to the paper andour project websitehttp://grail.cs.washington.edu/projects/interior/

3D viewer and dataset available

Page 76: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

For More Details• Come to a demo session this afternoon (1:15pm)

Page 77: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

For More Details• Come to a demo session this afternoon (1:15pm)

• Open-source SFM and MVS software

BundlerNoah Snavelyhttp://phototour.cs.washington.edu/bundler

PMVS Version2Yasutaka Furukawa and Jean Poncehttp://grail.cs.washington.edu/software/pmvs

Page 78: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Acknowledgements

• Sameer Agarwal and Noah Snavely for support on SFM and discussion

• Funding sources– National Science Foundation grant IIS-811878– SPAWAR– The Office of Naval Research– The University of Washington Animation Research Labs

• Datasets– Christian Laforte and Feeling Software for Kitchen– Eric Carson and Henry Art Gallery for gallery

Page 79: Reconstructing Building Interiors from Images Yasutaka Furukawa Brian Curless Steven M. Seitz University of Washington, Seattle, USA Richard Szeliski Microsoft.

Any Questions?

Images SFM MVS MWS Merging


Related Documents
Motion estimation Introduction to Computer Vision CS223B, Winter 2005 Richard Szeliski.

Motion estimation Introduction to Computer Vision CS223B,...

Category: Documents
Yasutaka Nagai (Toyota Central R&D Labs., Inc, Japan)

Yasutaka Nagai (Toyota Central R&D Labs., Inc, Japan)

Category: Documents
By Yasutaka IHARA May 2011 · 2011. 5. 30. · Comparison of some quotients of fundamental groups of algebraic curves over p-adic elds By Yasutaka IHARA ... the different exponent

By Yasutaka IHARA May 2011 · 2011. 5. 30. · Comparison.....

Category: Documents
Computer Vision (CSE 576) Staff Steve Seitz (seitz@cs)seitz@cs Rick Szeliski ( szeliski@microsoft.com ) szeliski@microsoft.com Ian Simon? ( iansimon@cs.

Computer Vision (CSE 576) Staff Steve Seitz...

Category: Documents
Mosaics Today’s Readings Szeliski, Ch 5.1, 8.1 StreetView.

Mosaics Today’s Readings Szeliski, Ch 5.1, 8.1 StreetView.

Category: Documents
Shape and Appearance from Images and Range Dataseitz/course/Sigg00/slides/curless-shape.pdf · Shape and Appearance from Images and Range Data Brian Curless University of Washington

Shape and Appearance from Images and Range...

Category: Documents
Object Recognition Szeliski Chapter 14. Recognition.

Object Recognition Szeliski Chapter 14. Recognition.

Category: Documents
Visual motion Many slides adapted from S. Seitz, R. Szeliski, M. Pollefeys.

Visual motion Many slides adapted from S. Seitz, R....

Category: Documents
Richard Szeliski Image Stitching 2 - University of Washington

Richard Szeliski Image Stitching 2 - University of...

Category: Documents
Sachie Kanada, Yasutaka Wakatsuki and Kazuhisa Tsuboki · ＊Sachie Kanada, Yasutaka Wakatsuki and Kazuhisa Tsuboki ... raingauge stations network of the ... time-longitude cross

Sachie Kanada, Yasutaka Wakatsuki and Kazuhisa Tsuboki ·.....

Category: Documents
Surface Reconstruction Using Volumetric Implicit Functions Some figures by Curless et al.

Surface Reconstruction Using Volumetric Implicit Functions.....

Category: Documents
Shape and Appearance from Images and Range Dataseitz/course/SIGG99/slides/curless-shape.pdf · Shape and Appearance from Images and Range Data Brian Curless University of Washington

Shape and Appearance from Images and Range...

Category: Documents