UWB and UW Seal
   
Home

Syllabus
Homework assignments
Lecture materials
Wikipedia readings
Links

Message board
Assignment dropbox
CSS 487 – Computer Vision – Spring 2011

CSS 487 – Computer Vision – Autumn 2017

University of Washington, Bothell - School of STEM - Computing & Software Systems

 

Professor:

Dr. Clark F. Olson

Class:

MW 1:15-3:15pm, DISC-162

Office: UW1 – 271B

Office Hours:

M 12:15-1:15pm, UW1-271B

Phone: (425) 352-5288

 

W 3:30-5:45pm, UW1-271B

E-mail: cfolson@uw.edu

 

or by appointment

 

Prerequisites

In order to succeed in this class you should have the following attributes:

  • Previously taken CSS 343 with a 2.0 or higher grade
  • Comfortable programming medium-sized programs in C++
  • Not scared of math (we’ll deal with a fair amount, but I’ll try to cover necessities)
  • Ready to learn and sufficient time to succeed. Please note this is a math and programming intensive course. Students have told me that taking this course concurrently with courses like 422/427/430/450/457 is difficult unless you are strong in these areas.

 

Course content

This course is meant to be a broad overview of the field of computer vision. This field is quite large and we won’t be able to cover every aspect. I expect to have some coverage of at least the following topics:

  • Image formation and camera geometry
  • Linear filtering, template matching, and edge detection
  • Segmentation, including the Hough transform and other methods for detecting structure
  • Extracting 3D data from 2D images using stereo and motion estimation
  • Object recognition using geometry, appearance-based methods, and SIFT
  • Applications: image databases, Photosynth, mobile robot navigation (if time permits)

 

Textbooks

Optional:

  • Computer Vision by L. G. Shapiro and G. C. Stockman, 1st ed., Prentice Hall, 2001.

https://courses.cs.washington.edu/courses/cse576/99sp/book.html

The following useful texts are also available at the library or online:

  • Computer Vision: Algorithms and Applications by R. Szeliski, 1st ed., Springer, 2011.

http://link.springer.com/book/10.1007%2F978-1-84882-935-0

http://szeliski.org/Book/

  • Introductory Techniques for 3-D Computer Vision by E. Trucco and A. Verri, 1st ed., Prentice Hall, 1998.
  • Learning OpenCV 3: Computer Vision in C++ with the OpenCV Library by A. Kaehler and G. Bradski, O’Reilly, 2017.

You can also find articles about computer vision on Wikipedia. Links to some are on the website.

 

Grading

Programs:

50%

Homework:

20%

Final exam:

30%

 

I use a linear grading scale such that: 75% = 2.0, 85% = 3.0, 95% = 4.0.

This class requires programming assignments, written homework, and reading at home. To be successful you should expect to spend 10-20 hours per week working outside of the classroom.

 

 

Message board

https://catalyst.uw.edu/gopost/board/cfolson/44154/

 

Class policies

All assignments are to be turned in using Catalyst prior to class on the due date at:

https://catalyst.uw.edu/collectit/dropbox/cfolson/40955/

Assignment will be marked as late promptly at the start of class.

 

There will be no individual extensions, but assignments can be turned in up to 5 days late with a 5% per day (including weekends) grade reduction. All assignments are to be completed independently (except for the project). While general discussion of the problem and clarification is acceptable, work that is turned in must be performed without collaboration, unless explicitly allowed in the specifications. It is not acceptable to use (or even view) code found on the web (for example, GitHub) and it is also not acceptable for your code to be publicly accessible on the web (for example, a public GitHub repository). Plagiarism will result in an assignment score of zero and a misconduct letter in your student record. Please be very careful to adhere to the student code of conduct:

http://www.washington.edu/cssc/student-conduct-overview/student-code-of-conduct/

 

Make-up/rescheduled exams will not be given except under extraordinary circumstances. It is not acceptable to schedule a trip or another appointment at the same time. Exceptional circumstances must be discussed with the instructor in advance, except in the case of an emergency, which should be well-documented.

 

Please turn off all cell phones prior to class. Talking and other noise during lecture should be kept to a minimum as a courtesy to other students who are trying to learn.

 

Tentative Schedule (subject to change)

Date

Topic

Reading

Assignments

Sep 27

Introduction, geometry

Chapter 1, 2, 12.5

 

Oct  2

Linear algebra, transformations

Course notes, Chapter 11.3

P1 assigned

Oct  4

Linear filters, convolution

Chapter 3.1-3.2, 5 (skip 5.9, 5.11)

 

Oct  9

Edge detection

Chapter 10.3.2-10.3.3

P1 due, P2

Oct 11

Hough transform, RANSAC

Chapter 10.3.4, 10.4

 

Oct 16

Color, texture

Chapter 6.1-6.4, skim Chapter 7

HW1 assigned

Oct 18

Segmentation

Chapter 3.4, 3.8, 6.5, 10

 

Oct 23

Class holiday – no lecture

 

 

Oct 25

OpenCV

See papers/wiki/web site

P2 due, P3

Oct 30

SIFT

See papers/wiki/web site

 

Nov  1

Application: Image databases

Chapter 3.6, 4.3-4.5, 8

HW1 due

Nov  6

Appearance-based recognition

Chapter 11

HW2 assigned

Nov  8

Geometric object recognition

Chapter 14.4

P3 due, P4

Nov 13

Feature matching and

Chapter 9.1-9.3.4

 

Nov 15

    Stereo vision

Chapter 12.6

 

Nov 20

Feature selection and

Chapter 10.5.2, 9.3.2

 

Nov 22

    Motion estimation

Chapter 13.11

HW2 due

Nov 27

Application: Photosynth

See papers/wiki/web site

 

Nov 29

Application: Mobile robots

See papers/wiki/web site

 

Dec  4

Project demonstrations

 

 

Dec  6

Project demonstrations

 

P4 due

Dec 13

Final exam

 

 

 

See also: School of STEM Course Policies