CSE152

Introduction to Computer Vision

Tuesday, Thursday 11:00-12:30

Warren Lecture Halls, Rm. 2207

http://www-cse.ucsd.edu/classes/sp04/cse152

CSE152 Discussion Board

 

ANNOUNCEMENTS

            6/07/04:    Instructor office hour now move to 12:45-1:00 on Tuesday, June 8

            5/06/04:   Assignment 2 deadline is extended to 5.00pm, May 06.

            4/22/04:   'Turn in' has been setup for the class. You need to submit all your source code for the assignment using this service (besides the hard copy version handin in class). For instruction, go here. Deadline for electronic turnin is 23:59:59, Apr 27.

            4/06/04:   Sorry for miscommunication about the dicussion session last week. This week (Apr 9), the discussion will be in WLH 2207, 12:00 - 12:50. The main topic is Matlab (again!!).

            3/31/04:  Class will be held where originally scheduled in WLH, Rm. 2207.  Apparently, there was an error in the sign that was posted, and the class shouldn’t be moved. (In fact, the sign should have been posted on Solis 104 telling students to come to WLH, 2207…

Instructor: David Kriegman

Office: AP&M 3101

Phone: (858) 822-2424

Email: kriegman@cs.ucsd.edu

Office Hours:    Wednesday1:30-3:00
 

TA: Diem Vu

Location: AP&M 2444 

Email: d1vu@cs.ucsd.edu

Office Hours:  Wed 2:30-4.:30

 

Class Description: The goal of computer vision is to compute properties of the three-dimensional world from images and video.  Problems in this field include identifying the 3D shape of a scene, determining how things are moving, and recognizing familiar people and objects.  This course provides an introduction to computer vision, including such topics as feature detection, image segmentation, motion estimation, object recognition, and 3D shape reconstruction through stereo, photometric stereo, and structure from motion.4 units.

 

Required Text: Introductory Techniques for 3-D Computer Vision, E. Trucco and A. Verri, Prentice Hall, 1998.

 

Prerequisites: Linear algebra and Multivariable calculus (e.g., Math 20A & 20F), data structure/algorithms (e.g., CSE100), a good working knowledge of C,C++, or Matlab programming.
 

Grading:

            Assignments:     45%

            Midterm:          20%

            Final Exam:       35%

 

Late Policy: Written homework will be due in class and accepted thereafter with a penalty of 10% per day starting from the due date. Programming assignments will have a hand-in procedure described with the assignment, and also has a 10% per day late penalty. No assignments will be accepted after the graded assignments have been returned or the solutions have been released.

Assignments

Syllabus

[ Note that this Syllabus is tentative & subject to change]

Week

Date/

Link to lecture notes

Topic/Readings

1

Intro to Computer Vision / T&V Chapter 1

 

Human Visual System

2

Image Formation, T&V pp.15-19

 

Color, Color is well-treated in many image-processing texts. Some reasonable on-line sources include: Basics of Color, A FAQ on Color

3

Segmentation & Binary images, Horn Chapters 3&4, available at e-reserves See on-line resource

 

Binary Images &/Filtering

4

Filtering, T&V pp. 55-63

 

Canny Edge detection, T&V 67-81

5

Curves, Hough Transforms, T&V, pp. 97-100

 

Intro to Shape-from-x, Midterm Review

6

Midterm

 

Stereo I,  T&V pp. 140-171

7

Stereo II

 

Photometric Stereo ,T&V pp. 140-171

8

Discrete structure from Motion , T&V pp.195-202, 208-211

 

Continuous motion T&V pp. 178-194

9

Optical Flow

 

Statistical Pattern Recognition, T&V pp. 248, 262-269

10

 

Model-based recognition/Final Exam Review, T&V 249-261

Notes and links

1. Programming languages: For this course, all programming will be in Matlab. A short tutorial assignment will provide the introduction   Click here for Serge Belongie’s Matlab resource links.

 

2. Another excellent textbook: Computer Vision -- A Modern Approach, Forsyth and Ponce

CV-online: A useful on-line compendium of Computer Vision sources
 

3. Handy Math reference: MathWorld

  

4. Some other useful vision links