Computer Vision 535/635 Wiki


This is the course wiki for 535 Computer Vision 2007 Fall semester


CPSC 535/635, Dr. J. E. Boyd

Lab Time: TR at 13:00, EDC 289

email: cpsc535@gmail.com

Additional resources by topic are listed below, to be updated regularly.

Student-editable Q&A wiki, where you can post q's and resources.

EXAM PREP- while i give no guarantee of full coverage, I have made efforts to make my slides/worksheets reflect what I would expect on an exam.

  • Go through all my slides, and make sure you get the concepts. Whenever I do a numerical or practical example, do it yourself! change a number, or get your own image, and do it.
  • do my worksheets!
  • Later today, I will post some things that have shown up on previous exams. Unfortunately I cannot post sample questions.
  • I will still be available via email over the weekend.

Dec 6

Review of image pyramids, partial derivatives, and more optical flow!! (Lukas Kanade)

Dec 4

Plane/Rotation Induced Homographies, Optical Flow Basics

Dec 1.

  • pdf versions of pptx slides uploaded. Thanks to helen for the help

Lab-Nov 22nd

Vicon / Camera Calibration demonstration

  • Lab will be held in MS 680. Wait outside 680 for your TA to arrive

Lab-Nov 20th

Bhattacharya coefficient, and basic camera calibration

Lab-Nov 16th

Midterm Solutions review

  • I will go over the midterm solutions

Lab-Nov 8th

Image Pyramids and Haar wavelets example

Lab-Nov 6th

Normalised cross correlation

REMINDER - Nov 5th

Second assignment due a week from today (nov 12th). After that, each assignment is due a week after!

Lab-Nov 1st

Template matching, correlation, cross correlation

Nov 2

I posted below the due-dates for your assignments.

Lab-Oct 25

Midterm review -- I will do 2 sample questions from previous midterms

  • Dr. Boyd has posted them here

XKCD: (thanks to YuNing)

Lab-Oct 23

More on 2D Kernels. explain the different edge detectors

Assignment 3 Notes

There was some question in the lab as to which deconvolution methods you have to implement. As specified in the assignment, you need to do both.

Jim

Lab-Oct 18th

Histogram Equalisation, Basic Convolution Kernels, Assignment 3 question

Lab-Oct 16th

Finish Binary Morphology, Hit or Miss Operators, and do Histograms/Normalisation

Note on Assignment 3

The supplementary material mentioned later in the assignment is a typo (mistake?) from a previous version. There is no supplementary material. All required readings are now included in the course, and, the second paragraph the assignment points to Section 5.4 of Sonka, Hlavac and Boyle.

Lab-Oct 11th

Assignment 1 overview, basic binary morphology (erosion / dilation)

Assignment 1 marks in

Will bring commented assignments to thursday's lab. If you do not come, you can arrange for pick up. Please check your marks below.

Changed Assignment Submission Requirements.

After fighting with the printer for Assignment 1, I have decided to require a paper submission as well as an electronic one. For those of you who have already submitted assignments before this change, I will accept the electronic-only submission....

Annoying printers...

Lab-Oct 9th

Region Identification, remote octave demo, assignment 3

Lab-Oct 4th

2D convolution example, sampling

Lab-oct 2nd

Convolution introduction

Lab-sep 27th

more fourier and assignment 2

So, I have decided to use the lab time over the next weeks to introduce each assignment, one by one, starting tomorrow. This will hopefully help those interested to get started and will allow you to ask initial questions. The assignment will be discussed at the end of the lab so those not interested can leave early.

Lab-sep 25th

introduction and fourier transform

Assignment Submission Instructions

DUE DATES: Last days to submit your assignments:

2nd assignmentNov 12
3rd assignmentNov 19
4th assignmentNov 26
5th assignmentDec 3

Please email your assignments as follows:

  • send to cpsc535@gmail.com
  • subject line: [user id] Assignment #. For example, [308442] Assignment 1
  • attach the assignment to the email. if it is a single file, attach that. Feel free to use zip/tgz/rar for larger submissions
    • please attach all relevant files. This includes source, generated images, write ups (see below), etc.
  • if i have trouble executing your code, I will require a live demo!
  • late submissions will not be accepted

Also, submit a hard copy which includes

  • code print outs
  • plots
  • your write ups

For your write ups:

  • word (.doc/.docx), .pdf, OpenOffice, postscript, are acceptable.
  • put your name and ID in the document
  • explain methods and approach, pitfalls, difficulties
  • explain algorithms not taught in class
  • give result images/diagrams/numerical data, and an analytical discussion of what your results mean.

Basically, show me that you understand the methods you use, why you use them, and why the results you gain are significant and useful.

Marks

IDAS1 (/30)AS2 (/30)AS3AS4AS5AS6AS7AS8Remaining AllowedFinal
0759242728.229.1  29.4  92%
37712023.14.814.7   25.5 59%
37721824.65.120.1   27.9 64%
52692526.421.319.5  27.6  80%
670625         
40222428.525.518.3   30 84%
5493          
74542719.8 25.8   19.2 61%
74612929.72728.5   21 90%
00352222.52727.3   25.2 83%
57461623.719.521     53%
85232023.1        
518019         
9495          
55132923.425.521.6 21.6   81%
09122726.723.727.4   24.9 87%
82132929.727   2427.6 92%
7225272422.527   30 87%
29541626.422.2 26.1 22.5  76%
6799232727.629.7  29.7  91%
Avearage 23.3325.3521.9223.5526.115.626.6425.7 79%