Advanced Algorithms in 3D Computer Vision
  Fall 2011 Tentative Schedule
It can change substantially




August 31


Introduction

Introduction to Open Inventor Libraries.

Reading:
Luc Van Gool etal., 3D Challanges and Non-In-Depth Overview of Recent Progress, 3D Imaging and Modlineg Conference, 2007.
Fausto Bernardini and Holly Rushmeier, The 3D Model Acquisition Pipeline, Computer Graphics Forum V.21, No.2, pp. 149-172.
Fausto Bernardini etal. Building a Digital Model of Michelangelo's Florentine Pieta, Computer Graphics and Applications, 2002

Pieta Project web site.

Slides:

Introduction, Sensors




Libraries and Sample Code
Coin3D library

Reading and Visualizing Points in Scan Order (code presented in class), Input File (Park Avenue)

Sample Code (to read input files)
Code EigenVector
Computing local planes (section 3.1)
Coin3D_Sample_Code

Input File 0 (Small Example)
Input File 1 (Real Data)
Input File 2 (Real Data)



September 7


Laser Scanning and
Segmentation

Reading:

Review of 20 Years of Range Sensor Development by Francois Blais, Journal of Electronic Imaging, 13(1): 231-240. January 2004.

Szymon Rusinkiewicz, Olaf Hall-Holt and Marc Levoy, Real-Time 3D Model Acquisition, SIGGRAPH 2002.

Range Image Segmentation for Modeling and Object Detection in Urban Scenes, C. Chen, I. Stamos, The 6th International Conference on 3-D Digital Imaging and Modeling, Montreal, Canada, August 21-23 2007. PDF

Slides:

RangeSegmentation, RegionGrowing, Hough

Project1

Other references:

Yizhou Yu, Andras Ferencz, and Jitendra MalikExtracting Objects from Range and Radiance Images, IEEE Transactions on Visualization and Computer Graphics, Vol. 7, No. 4, 2001, pp.351-364.
J. Shi and J. Malik, Normalized Cuts and Image Segmentation, IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2000.  <Helpful references>

J. Shi and J. Malik, Normalized Cuts and Image Segmentation, IEEE Conf. in Computer Vision and Pattern Recognition, Puerto Rico, 1997 [shorter version of the journal paper].
Besl and Jain, Segmentation through Variable-Order Surface Fitting, IEEE Transactions on PAMI, 1988 (Vol. 10, number 2).






September 14



Alignment of 3D Point Sets (Range Registration).

Reading:

Efficient variants of the ICP algorithm

I. Stamos and M. Leordeanu, Automated Feature-Based Range Registration of Urban Scenes of Large Scale, IEEE International Conference of Computer Vision and Pattern Recognition 2003, pp. 555-561, Vol. II, Madison, WI.

Registration notes - Peter K. Allen, Columbia University

4-points Congruent Sets for Robust Surface Registration

A. Makadia, A. Patterson, and K. Daniilidis. Fully automatic registration of 3d point clouds. In IEEE Conf. Computer Vision and Pattern Recognition, New York, NY, June 18-20, 2006. pdf

Andrew E. Johnson and Martial Hebert, Using Spin Images for Efficient Object Recognition in Cluttered 3D Scenes,IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 21, NO. 5, MAY 1999.

Slides+:

ICP_slides, ICP_point_to_plane_linear_version,Line-based Registration

Original ICP papers: BeslMcKay,ChenMedioni

Horn's Closed Form Solution

Tutorials+code

A tutorial on k-d trees, Approximate Nearest Neighbor Code



September 21


3D Modeling on the Mesh Level.

Reading:

Mesh representations (slides needed).

Mesh Data Structures - Intro to Decimation (2nd lecture)  by Prof. Szymon Rusinkiewicz, Princeton University

The "Zipper" algorithm

The "BPA" algorithm

BPASlides, ZipperSlides1,SurfaceReconstructionSlides,MeshDataStructuresSlides




September 28 -- No class






October 5


3D Modeling on the Voxel Level.

Reading:

Michael Reed and Peter Allen 3-D Modeling from Range Imagery: An Incremental Method with a Planning Component, Image and Vision Computing, V. 17, pp. 99-111, 1999.

Brian Curless and Marc Levoy, A Volumetric Method for Building Complex Models from Range Images, Proc. SIGGRAPH '96.

Brian Curless and Marc Levoy, Better optical triangulation through spacetime analysis, Proc. ICCV '95.


Brian Curless: PhD Thesis



MichaelReed_Slides, VRIP_slides

Project2




October 12


Mesh Simplification

Reading:

Mesh simplification using quadric error measures

Variational Shape Approximation

Slides:
quadric_error_measures, vsa

October 19



VSA (continued from Oct. 12)

Presentation by Allan: Aleksey Golovinskiy, Vladimir Kim, and Thomas Funkhouser,
Shape-based Recognition of 3D Point Clouds in Urban Environments
International Conference on Computer Vision (ICCV), September 2009.




October 26



Passive Techniques (Structure From Motion)

Reading:

Detailed Real-Time Urban 3D Reconstruction from Video.

PhotoSynth

S.M. Seitz, etal. A comparison and evaluation of multi-view stereo reconstruction algorithms, IEEE Conference on Computer Vision and Pattern Recognition, volume I, pages 519-526, 2006.




  November 2


Fusing 3D and 2D techniques

Reading:

Integrating Automated Range Registration with Multiview Geometry for the Photorealistic Modeling of Large-Scale Scenes,I. Stamos, L. Liu, C. Chen,G. Wolberg, G. Yu, S. Zokai, International Journal of Computer Vision [Special Issue on Modeling and Representation of Large-Scale 3D Scenes], 2008. PDF -- The original publication is available at www.springerlink.com (DOI 10.1007/s11263-007-0089-1). To get the original publication click here

A systematic approach for 2D-image to 3D-range registration in urban environments, L. Liu, I. Stamos, VRML Workshop, 11th International Conference on Computer Vision, Rio de Janeiro, Brazil, October 14-20 2007. PDF

Estimating the Location of a Camera with Respect to a 3D Model Gehua Yang, Jacob Becker, Charles V. Stewart, 3DIM 2007.

Slides:

CameraCalibrationSlides, DistortionCorrectionNotes, 2D_3D_AutomatedMethods

More data:

SomeScansToTryWithICP


Presentation by Ben:
D. Munoz, J. A. Bagnell, N. Vandapel, M. Hebert, 
Contextual Classification with Functional Max-Margin Markov Networks, 
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2009 

Link

<<Projects Formed>>

A) The final project will be carried out by Xiaodong Yang and Chucai Yi in a group. Our topic is "Human Action Recognition based on Kinect-Captured Visual and Depth Channels". We will model the human action by both appearance features such as Histogram Orientation of Gradient (HOG), and the 3-D skeleton structure including the positions of head, body and limb. Through the extracted features from different samples of human actions (stand static, sit, squat, etc.), we train SVM-based classifiers to recognize the involved actions.
B) Kwoo Wai: Kinect based modeling.
C) Allan Zelener: Object classification.
D) Ben Hixon: Object classification.
E) Juan Liu: Morphology operations







November 9


City Modeling

Reading:

An Automated Method for Large-Scale, Ground-Based City Model Acquisition, C Früh, A Zakhor - International Journal of Computer Vision, 2004 - Springer
Towards Wiki based Dense City Modeling, Arnold Irschara, Christopher Zach and Horst Bischof, ICCV VRML Workshop 2007 (need link!)

Student Presention: Tom Flynn, Spin Images.

The MIT City Scanning Project.

Seth Teller and Matthew Antone Scalable Rotational Registration of Omni-Directional Image Networks, IJCV 2001. City Scanning Project.

Seth Teller and Matthew Antone Scalable, Absolute Position Recovery for Omni-Directional Image Networks IJCV, 2001

J. Diebel and S. Thrun. An application of markov random fields to range sensing. In Proceedings of Conference on Neural Information Processing Systems (NIPS), Cambridge, MA, 2005. MIT Press.

Presentation by Joseph Samleo: Accelerated patch-based plane clustering of noisy range images in indoor environments for robot mapping ", Kaushik, Ravi; Xiao, Jizhong; Journal of Robotics and Autonomous systems, 2012 (accepted for publication)



November 16


Symmetry

Main Reading: 

Discovering Structural Regularity in 3D Geometry, SIGGRAPH 2008 [related: Partial and Approximate Symmetry Detection for 3D Geometry, SIGGRAPH 2006 and Symmetrization, SIGGRAPH 2007]


A Planar-Reflective Symmetry Transform for 3D shapes, SIGGRAPH 2006 [related: A reflective symmetry descriptor for 3D models, Algorithmica 2003]

Student Presentation: Sam Friedman, Symmetry.

Other:

Image-based procedural modeling of facades, SIGGRAPH 2007.

Detecting and Matching Repeated Patterns for Automatic Geo-tagging in Urban Environments

Presentation by Juan Liu: Transform Invariant Low-rank Textures.



November 23


Texture Mapping

Reading:

Acquiring, Stitching and Blending Appearance Attributes on 3D Models, by Rocchini, P. Cignoni, C. Montani and R. Scopigno The Visual Computer, Springer International, 18 (2002) 3, 186-204, 2002.

Building illumination coherent 3D models of large-scale outdoor scenes, A. Troccoli and P.K. Allen, IJCV Special Issue, 2008.

"High-Quality Texture Reconstruction from Multiple Scans" by Fausto Bernardini, Ioana M. Martin and Holly Rushmeier, IBM T. J. Watson Research Center.

Presentation by Kwoo Wai: Kinect based modeling.

Presentation by Yang Xiadong:  "Real-Time Human Pose Recognition in Parts from Single Depth Images", CVPR 11.



November 30


Alternative Methods

Reading:

Steven J. Gortler, Radek Grzeszczu, Richard Szeliski, Michael F. Cohen, The Lumigraph, SIGGRAPH 1996.

Marc Levoy and Pat Hanrahan LightField Rendering , SIGGRAPH, 1996. More recent work: Marc Levoy, IEEE Computer Society 2006: Light Fields and Computational Imaging.

Chris Buehler etal. Unstructured Lumigraph, SIGGRAPH 2000.

The Facade Method

Paul E. Debevec, Camillo J. Taylor, and Jitendra Malik. Modeling and Rendering Architecture from Photographs. In SIGGRAPH '96, August 1996.

Presentation by Chucai Yi: Parametric shaping of Human bodies in images.




December 7


Student Presentations