cs682

CS 682
Computer Vision
Jana Kosecka
http://cs.gmu.edu/~kosecka/cs682.html

Week 1: Introduction, Review of linear algebra, 2D-3D transformations, image formation process
(intro.pdf, linear-algebra.pdf (.mov), image-formation.pdf)
Week 2: Geometry and photometry of the image formation process, camera/rigid-body motion
gray-level images, point-wise image operations, linear filtering
(image-filtering.pdf, threshold-selection.pdf)
Week 3: Filtering, Pointwise operations, thresholding, edge detection
(edge-detection.pdf, canny-edge-detector.pdf )
Week 4: Image features, point/line features; ( line-fitting.pdf, pyramids.pdf, image-features.pdf )
Week 5: Feature correpondences and matching (more-linear-algebra.pdf (.mov), feature-correspondences.pdf)
Week 6: Two view geometry (two-view-calibrated.pdf (.mov))
Week 7: Two view uncalibrated geometry (two-view-uncalibrated-1.pdf two-view-uncalibrated-2.pdf cs682-stereo.pdf )
Week 9: Thresholding, Segmentation (thresholding.pdf segmentation.pdf)
Week 10: Segmentation, EM (segmentation-part2.pdf)
Week 11: Recognition (recognition.pdf)
Week 12: Template/Shape Matching (template-matching.pdf)
Week 13: Texture (texture.pdf)
Week 14: Multiview (multiview.pdf)
Week 14: Step by Step (step_by_step.pdf)

Additional Readings

M. Turk and A. Pentland. Eigenfaces for Recognition.
Journal of Cognitive Neuroscience, 3(1), 1991. (.pdf) , excercise
D. Lowe. Distintive Image Features from scale-invariant keypoints.
International Journal of Computer Vision, 2004. (.pdf)

Project Resourcess
Object recognition database
Background subtraction

Tutorials, Data, Code
Book code
Sample sequences (coutesy of http://www.cs.cmu.edu/~cil/vision.html , tsukuba university)
Fundamentals of Image Processing (.pdf)
Quick introduction to Matlab
Tracker , Harris Corner Detector

Stereo pairs

Other Resources

Matlab Primer (use online help or help "function name", when in trouble)
Computer Vision Compendium
Optical Illusions

Project Ideas

Wide-Baseline Matching and Computation of Epipolar Geometry
Interactive Model Construction Using Prior Knoweldge of the Scene
Dense Stereo Matching (expoliting ordering constraints)
Feature Tracking and Estimation of Motion and Structure from multiple views
Structure and motion recovery from line segments
Image Based Mosaicing (automated mosaic construction)
Face Detection and/or Tracking
Image Morphing
Object Detection in cluttered environments
Motion Detection via background subtraction
Independent Motion Detection

References

Book web site see the sample code: http://cs.gmu.edu/~kosecka/bookcode.html
Face/car/object detection/recognition. Here are few papers. More references can be tracked down in them.

Jones, Regh. Statistical Color models with application to skind detection, CVPR 1999.
Schniederman, Kanade. A statistical Method for 3D Object detection applied to cars and faces.
CVPR200 and earlier 1998 paper.
Weber, Welling, Perona Unsupervised learning of models for recognition, ECCV 2000
Turk and Pentland, Eigenfaces, 1991
Nene, S. Nayar. PCA applied to object recognition+ COIL object recognition database
www.cs.columbia.edu/CAVE/ research/softlib/coil-100.html
FERET face database, YALE face database more information of faces at
VISMOD Face Recognition Home Page