CSC I6716  Computer Vision- Fall 2010


Instructor: Professor Zhigang Zhu
http://www-cs.engr.ccny.cuny.edu/~zhu

Teaching Assistant: Mr. Wai L. Khoo
WKhoo@gc.cuny.edu

Credits:     3.0
Class Meet Time:     Monday and Wednesday 12:30-01:45PM,  Room: NAC-7225
Office Hours:   Monday and Wednesday 02:00-03:00 PMRoom: NAC  8/210

    City College of New York


Course Update Information

 August 30 (Monday), 2010. First class meet of our course.
 August 31, 2010. Assignment 1 online, due on Sept 13 before class.  You may submit your homework to Mr. Wai L. Khoo at  WKhoo@gc.cuny.edu
 September 02, 2010. For those who cannot read pptx files, slides are also posted in PDF formats.
 September 13, 2010. Assignment 2 online, due on September 27th before class
 September 22, 2010. Grading for Assignment 1. We will discuss the answers in class today.
 September 27, 2010.  Assignment 3 online, due on Oct 25, 2010 before class
 October 03, 2010. Grading for Assignments 1-2
 October 20, 2010.  Assignment 4 online, due on Nov 29 , 2010 before class
 November 01, 2010. No class meet and office hours on Wed (Nov 03). If you have questions, please come to Prof. Zhu's office at 2:00 - 3:00 pm on Tuesday (Nov 2).
 November 08, 2010. Grading for Assignments 1-3
 November 24, 2010. Exam Review (slides in pptx and [PDF])  Happy Thanksgiving!
 November 29, 2010. Student Presentation Schedule(Updated)
 December 04, 2010. Grading for Exam and Assignments 1-4
 December 15, 2010. Final Grading.  Happy Christmas to ALL!

Course Objectives

Computer vision has a rich history of fundamental work on stereo and visual motion, which has dealt with the problems of 3D reconstruction from multiple images, and structure from motion from video sequences. Recently, in addition to these traditional problems, the stereo and motion information presented in multiple images or a video sequence is also being used to solve several other interesting problems, for example, large-scale scene modeling, video mosaicing, video segmentation, video compression, video manipulation and video surveillance. This is sometimes summarized as video computing. Computer vision is playing an important and somewhat different role in solving these problems in video computing than the original image analysis approach in the early days of vision research.The course "Computer Vision" will include advanced topics in video computing as well as fundamentals in stereo and motion.



Course Syllabus and Tentative Schedule (mm/dd)

(Academic calendar, Fall 2010)

Part I. Computer Vision Basics 

I-1. Introduction: What, Why and How (pptx slides) [PDF] - 08/30
I-2. Image Formation: Digital Image Basics (pptx slides) [PDF] - 09/01 (Assignment 1)
I-3. Image Enhancement (1): Histogram and How to Make a Picture Prettier (slides in pptx  and [PDF] for I-3 and I-4) - 09/08 (No class meet on 09/06)
I-4. Image Enhancement (2): Noise Removal and How to Make a Picture even Prettier (Assignment 2) - 09/13
I-5. Edge Detection (1): Edge Detectors and Sketch Generation (slides in pptx and [PDF] for I-5 & I-6) - 09/15
I-6. Edge Detection (2): Hough Transform for Obtaining Shapes - 09/20

Part II.  3D Computer Vision

II-1.  Camera Models (slides in pptx and [PDF])-
    09/22 (Geometric Projections:  the Rules Governing 2D Imaging),
    09/27 (Camera Parameters: the Secrets Inside Your Camera ) (Assignment 3),
    09/29 (Linear Algebra Models: Make the Math Simpler)
II-2.  Camera Calibration (slides: [pptx], [pdf], [handouts-for-print]) -
    10/04 (Problem Definition:  the Tools You Must Know),
    10/06,10/13 (Direct Approach: Divide and Conquer), (No class meet on 10/11)
    10/18 (Projective Matrix Approach: All in One ) 

II-2a
Project Topics and Assignment Discussions  - 10/20

II-3.  Stereo Vision  (slides: [pptx], [pdf], [handouts-for-print])
    10/25 (Problem Definition: Two is Better Than One) ( Assignment 4),
    10/27 (Epipolar Geometry: the Trick for Simplifying Your Task),
    11/01 (Correspondence Problem: The Key to Success) - (No class meet on 11/03, Office hours moved to the same time on Tuesday)
    11/08 (Reconstruction Problem: Getting 3D from 2D Data) 

II-4.  Visual Motion - (slides: [pptx], [pdf], [handouts-for-print])
    11/10 (The Motion Field of Rigid Motion: See the Motion in Images) ,
    11/15 (Optical Flow Approach: Spatio-Temporal Gradients in Work) ,
    11/17 (Feature-based Approach: Tracking Individual Points) ,
    11/22 (Advanced Topics: Video Mosaicing, Target Tracking and Video Coding)

Part III. Exam, Project and  Project Presentations

III-1. Exam Review (slides in pptx and [PDF])  -11/24 (before Thanksgiving)
III-2. Advanced Topics & Projection Discussions   - 11/29
III-3. Exam  -  12/01
III-4. Student Project Presentations (Schedule) - 12/6, 12/8, 12/13


Textbook and References

Main Textbook:
    “Introductory Techniques for 3-D Computer Vision”,  Emanuele Trucco and Alessandro Verri, Prentice Hall, Inc., 1998  (ISBN: 0132611082, 343 pages ).
    (The book is out of print, but you may find copies of the book at Amazon or Barnes & Noble, among other places)

Reference Textbook:

  1.     “Computer Vision – A Modern Approach” , David A. Forsyth, Jean Ponce, Prentice Hall, 2003 (ISBN: 0130851981 , 693 pages).
  2.     “Three Dimensional Computer Vision: A Geometric Viewpoint” , Olivier Faugeras, The MIT Press, November 19, 1993 (ISBN: 0262061589 , 695 pages)
Supplements: 
    Online References and additional readings when necessary.
 

Grading and Prerequisites

The course will accommodate both graduate and senior undergraduate students with background in computer science, electrical and computer engineering, or applied mathematics. Students who take the course for credits will be required to finish 4 assignments (40%), one midterm exam (40%), and  one programming project (20%, including submit a report and give a small presentation to the class at the end of the semester). The topics of the projects will be given in the middle of the semester and will be related to the material presented in the lectures.

This course is under I6700: Topics in Scientific and Statistical Computing, in the area of "Computing Methodologies and Mathematical Computing"  of our Computer Science Master Program.  Students are required to have a good preparation in both mathematics (linear algebra/numerical analysis) and advanced programming.


Copyright @ Zhigang Zhu , Fall 2010