CMSC
733
Pictorial Information
Fall
2012
Instructor: Yiannis Aloimonos (yiannis@cs.umd.edu)
Lecture: Tu-Th 11:00-12:15 - CSI
3118
Office Hours: Tu 12:30-1:30, Thu 10:00-11:00 or by appointment - AVW 4475
The best way to reach the instructor is through email
Teaching Assistants: Aleksandrs Ecins (aecins at cs dot umd dot edu), Austin Myers (amyers at cs dot umd dot edu)
Office Hours: Mo 10:00-11:00, We 10:00-11:00 - AVW 4426
Course Outline
In this class we will cover the following topics:
- 1. Multiple View Geometry:
- Image Formation, Geometric aspects, Radiometric Aspects, Digital Images, The Human Eye, Camera parameters, Camera Calibration, Intrinsic Parameters, Extrinsic Parameters, Stereo, The Correspondence Problem, Epipolar Geometry, 3D Reconstruction
- 2. Edge Detection:
- Gradient based edge Detectors, Laplacian, Parametric Models
- 3. Other Image Features:
- Hough Transform, Ellipse fitting, Deformable contours Sampling and Aliasing, Noise Reduction small
- 4. Lightness and Color:
- Surface Reflectance, Recovering Lightness, The Physics of Color, Human Color Perception, Color Representations
- 5. Filters:
- Linear Filters and Convolution, Spatial Frequency and Fourier Transform, Sampling and Aliasing, Noise Reduction
- 6. Motion and Video Processing:
- The Image Motion Field, Estimation of 3D Motion and Structure, Segmentation on the basis of different Motion, Image Compression
- 7.Shape from Single Image Cues:
- Surface Descriptions, Shape from Contours, Shape from Shading, Shape from Texture
- 8. Segmentaiton:
Recommended books
Multile View Geometry in Computer Vision by Hartley, R.~I. and Zisserman, A.
Robot Vision by Berthold K. P. Horn
Computer Vision: A Modern Approach by David A. Forsyth and Jean Ponce
Computer Vision: Algorithms and Applications by Richard Szeliski
Computer Vision: Models, Learning, and Inference by Simon J.D. Prince
Grading
Homework: 40%, Project: 30%, Exams: 30%
Lecture Notes
Lecture notes will appear as the semester goes on
- Topic 1: Introduction [pdf]
- Topic 2: Image Formation 1 [pdf] [ppt]
- Topic 3: Linear Algebra Review [Linear Albebra Tutorial]
- Topic 4: Filtering [ppt1] [ppt2]
- Topic 5: Edge detection [ppt] Canny edge detection [m-file]
- Topic 6: Resampling [ppt] (Slides from Univ. of Washington)
- Topic 7: Image motion [ppt]
- Topic 8: Statistics on image features: Review of statistical concepts [ppt] [Website on illusions]
- Topic 9: Stereopsis [ppt]
- Topic 10: Projective Geometry [ppt] (10 MB)
- Topic 11: Epipolar Geometry [ppt] (8 MB)
- Topic 12: Interpretation of image motion fields [ppt]
- Topic 13: 3D motion estimation from image derivatives [ppt] Updated!
- Topic 14: Shape from Shading [pdf] (from Daniel DeMenthon)
- Topic 15: Texture [ppt] (5.5 MB)
- Topic 16: Tracking with Kalman Filters [pdf] (from Daniel DeMenthon)
Quizzes
Quiz 1: Transform each of the images to images of frontoparallel surfaces, by applying an appropriate homography, as discussed in class. Due Thursday 9/13 in class [img1] [img2]
Quiz 2: Perform an affine rectification of the image i.e. parallel lines in the scene should be parallel in the image [img1]
Quiz 3, due 11/01 [imgs]
Quiz 4, due 11/01 [imgs]
Update: you can use [this code] to stitch your images in Matlab
Homework
Homework 1, due 10/1 [pdf]
Homework 2. Use the images provided to create a 3D model of the object [seq1] [seq2]. Due before Thanksgiving.
Update: [Calibration data]
Update 2: A project on structure from motion using Tomasi Kanade factorization [link]
Project
Here's the data for the project [data]. Refer to readme file inside.
Additional Resources
Additional material will be distributed in class or posted in the web site
http://www.youtube.com/watch?v=q8xsXFU7dK0
http://www.dsi.unive.it/~atorsell/Visione/02-Image%20Formation.pdf
Dialogue 2
Notes for dialogue 2