CMPT 414   Model-based Computer Vision
Spring 2017

Lecture Time: MWF 12:30-1:20
Classroom: RCB8100   Lab: ASB 9804

Instructor:   Dr. Ze-Nian Li, TASC-1 8223, Phone: 778-782-3761,
Office Hours: Mondays and Wednesdays 1:30-2:20
TA: Feitong Tan,
Office Hours: Fridays 1:30-2:30

Objectives: This course covers various topics in computer vision with the emphasis on model-based approaches. Main subjects include image processing techniques, Hough transforms, 2-D and 3-D modeling and matching, neural networks for computer vision, and stereo vision. State-of-the-art model-based vision systems will be used as study cases. Students are expected to complete a term project.

  • Low-level Image Processing
  • Hough Transforms
  • 2D and 3D Representations - Modeling
  • Neural Networks for Computer Vision
  • Matching Using Invariant Local Features (SIFT)
  • Stereo Correspondence Algorithms
  • Other Topics
Prerequisites: MATH 152 and nine credits in CMPT upper division courses, or permission of the instructor.
Grading: Assignments 20%, Midterm 20%, Term project 25%, Final exam 35%.
Late Penalties:Assignments are due at the beginning of the specified class time.   For each day late, 10% of the total possible points will be deducted. No work will be accepted after two days late.

Recommended Textbooks:
  • R. Szeliski, "Computer Vision: Algorithms and Applications", Springer, 2010.   Pre-publication version on line:
  • D.A. Forsyth and J. Ponce, "Computer Vision: A Modern Approach", 2nd ed., Prentice Hall, 2012.
  • L.G. Shapiro and G.C. Stockman, "Computer Vision", Prentice Hall, 2001.
  • E. Trucco and A. Verri, "Introductory Techniques for 3-D Computer Vision", Prentice Hall, 1998.
  • D.H. Ballard and C.M. Brown, "Computer Vision", Prentice Hall, 1982.