Digital Video Processing
Course Outline - Winter 2017

| [Background material] | [Course schedule] [Assignments] | [Final project] | [Test Images] | [Course notes] | [Useful links (matlab, tutorials, etc.)]
Course communication: Instructor: Maria A. Amer; Office: EV.5.227; Office hours: by appointment; Email: ; Website: http://users.encs.concordia.ca/~amer/teach/elec6631/

Rational: Video is becoming an integral part in various applications such as entertainment, education, medicine, security, and even wireless devices. In recent years, video processing technologies have started emerging from universities. They are being increasingly used in social, commercial, industry and military applications. The effective integration of video signals in modern devices, requires understanding of video processing techniques such as video sampling, enhancement, and video analysis.

Course description: This course introduces basic theoretical concepts required to understand and design video processing systems. Students will be able to understand many algorithms described in current video processing literature. They will gain hands-on experience through a comprehensive final project. The course will cover: Video signals and systems, Fourier analysis of video signals, video scanning and transmission, spatio-temporal sampling, selected material on the Human Visual System, modeling of video components, motion representation and estimation, video filtering and conversion, introduction to video compression.

Course prerequisite: "Digital signal processing"

References: Required textbook: Yao Wang, Jörn Ostermann, and Ya-Qin Zhang, Video Processing and Communications, Prentice Hall, 2001; ISBN-10: 0130175471; ISBN-13: 9780130175472

Grading: The contribution of course components to the final grade are as follows:

  1. Lab Assignments -------------- 20%
  2. Midterm paper presentation - 10%
  3. Final project presentation --- 15%
  4. Final project report ------------ 20%
  5. Final Exam --------------------- 35%

Assignments: Assignments will include theoretical and computer lab problems. The objective is to implement and experiment with learnt theories in their context. See the course web site for a list of assignments, submission deadlines, and submission instructions.

Midterm paper presentation: You are required to present, in class, a recent technical paper in the field of video processing or compression, or their applications; planned for Week 6.

Final Exam: The final exam will cover theoretical problems. The exam will be pen-only, closed book, primitive calculators, and 2 pages (1 sheet) of formulas. No make-up exam will be set.

Final project: In the final project each student will work on a problem in digital video processing or compression. To complete your project you need to follow these steps:

Learning outcomes:

  1. Illustrate the principles of video models, capture, and display:
    Is a video an array?
  2. Apply multidimensional Fourier to analyze video signals:
    What are sources of temporal frequency? What are spatial frequencies?
  3. Apply elementary spatial and temporal video operations:
    What is the space-time convolution theorem?
  4. Demonstrate visual features extraction:
    Why we extract features of interest from a video?
  5. Interpret the multidimensional sampling theorem on video signals:
    Can we obtain a high-quality discrete video?
  6. Shows the use of Motion Estimation and Motion Compensation:
    How do we estimate motion vectors for different applications?
  7. Demonstrate the foundations of video filtering:
    How do we separate video signal from noise? Is video filtering not ill-posed problem?
  8. What makes high-resolution video possible?
  9. Apply fundamentals of video coding:
    What makes video compression possible? How do we compress video?

Course notes: Course notes will be posted on the course web site so you can view them at your convenience. You can find notes based on the textbook here.