Make sure you hit refresh/reload to get the most recent update of this site

Back to Luke's homepage

Announcements

Last updated May 5th, 2003

• Andrew Zisserman's lecture & additional references for the interested are up  ( an in depth knowledge not required for exam).
• Assignment 2 at Engineering office filing cabinet.
• Exam: Tuesday 6th May 2003  9.00AM MELVILLE HALL.  
  Reading time: 9:00-9:15   Exam: 9:15-12:15
  No permitted materials

 

Lecturer

• Luke Fletcher
• luke.fletcher@anu.edu.au
• Phone X-58685
• rsise.anu.edu.au/~luke

 

Timetable

Time	Monday	Tuesday	Wednesday	Thursday	Friday
9	Lecture JD101	Lab Group A Chem. G51B
10	X
11	X		Lab Group B Chem. G51B
12	X	Lecture HA G052		Lecture JD101
1	X	X
2			X		Lecture JD101
3	X	X	X
4	X	X	X

Lecture Slides

Lecture slides will be available online as pdf documents. I will provide 6-up versions for printing and 1-up versions for viewing online.

2D Computer Vision

Images and Image Enhancement: 1-up 6-up

Binary Image Analysis: 1-up 6-up  Addendum: 1-up 6-up

Linear Filtering: 1-up 6-up

Feature Detection 1 Gradient and Edge Detection: 1-up 6-up

Colour Imaging: 1-up 6-up

Feature Detection 2 Hough Transform and Template Matching: 1-up 6-up

Tracking: 1-up 6-up

Images in the Frequency Domain: 1-up 6-up

3D Computer Vision

Homogeneous Coordinate Transformations: 1-up 6-up

Camera Model: 1-up 6-up

Linear Minimisation: 1-up 6-up

Epipolar Geometry and the Fundamental Matrix: 1-up 6-up

Calibrated Reconstruction: 1-up 6-up

Planar Rectification and Texture Mapping: 1-up 6-up

Reconstruction Ambiguities: 1-up 6-up

Assignment 2 hints: 1-up 6-up

 

Robot / Real-time Vision

Disparity Maps: 1-up 6-up

Optical Flow: 1-up 6-up

 

Guest Lectures

Andrew Zisserman: 1-up

Additional information from Andrew for interested people.  
For automatic computation of the fundamental matrix and reconstruction for sequences, probably the best papers to start with are:

            3D Model Acquisition from Extended Image Sequence, Beardsley, Torr and Zisserman, ECCV 96

            Automatic Camera Recovery for Closed or Open Image Sequences, Fitzgibbon and Zisserman, ECCV 98

            Both can be downloaded from: http://www.robots.ox.ac.uk/~vgg/publications/html/index.html

 

Review Lecture  

Review lecture : 1-up 6-up (mainly just  Ass1 points & topic headings)

 

Labs

• Lab allocations are just to ensure a relatively even distribution of students among classes.  No role is taken.  The only requirement is for me to see that you have completed the Lab exercise successfully.  
• I encourage students to start the exercises before the Lab time and just make an appearance to show me the result / help in debugging.  
• Lab Allocations have moved down in page.

Lab Sheets and Assignments

• Lab1: Matlab familiarisation, thresholding and image moments: ENGN4528  ENGN6528 
• test image: text.tiff

 

• Lab2: Colour Spaces and Faces:  ENGN4528  ENGN6528
• code skeleton: Lab2.m 
• colourspace conversion function: RGB2Lab.m (courtesy of Mark Ruzon)
• test image: face.jpg 

 

• Assignment 1: Face Detection and Tracking: ENGN4528  ENGN6528
• code skeleton: A1_XXXXXXX.m
• image sequence and supporting files: A1files.zip
• For an additional reference on particle filters in computer vision try the Forsyth and Ponce (Computer Vision - a modern approach) website (chapters that didn't make it into the book)  particularly the chapter on "Tracking with non-linear dynamic models".

 

• Lab3: Epipolar Geometry and the Fundamental Matrix:  ENGN4528  ENGN6528

·         test images: shed1.png, shed2.png

·         helper function: make_polygon_model.m

·         helper function: epipolar_viewer.m (courtesy of David Leibowitz)

 

• Assignment 2: Calibrated 3D Reconstruction from 2 Views: ENGN4528  ENGN6528
• test images: shed1.png, shed2.png
• camera calibration matrices: K1_K2.mat
• helper function: make_polygon_model.m
• helper function: plot_polygon_model.m
• helper function: format_figure.m
• helper function: show_planar_image.m
• helper function: view3d.m
• helper function: make_vrml_model.m
• helper function: epipolar_viewer.m (courtesy of David Leibowitz)
• helper function: vgg_imgwarp.m (courtesy of David Leibowitz and the Oxford Visual Geometry Group)
• helper html file: model.htm  to view your model in Cosmo player download and open this page in internet explorer

 

Revision Exercises

• Exercises 2D
• Exercises 3D

 

Images

• Images for Lab1: text.tiff
• Images for Lab2: face.jpg
• USC-SIPI image database

 

Recommended Reading

Computer Vision: A Modern Approach
by David A. Forsyth, Jean Ponce
Prentice Hall, 2001

Multiple View Geometry in Computer Vision
by Richard Hartley, Andrew Zisserman
Cambridge University Press, June 2000

Computational Vision: Information Processing in Perception and Visual Behavior (Computational Neuroscience)
by Hanspeter A. Mallot, John S. Allen

Additional Resourses

An excellent reference for 3D computer vision is Richard Hartley and Andrew Zisserman's recent book Multiple View Geometry in Computer Vision, Cambridge University Press, June 2000. Of particular interest is the sample chapter available online:

Epipolar Geometry and the Fundamental Matrix (pdf document)

The internet is full of material on computer vision, some excellent sources are:

• CVonline maintained by Edinburgh University.
• The Computer Vision Homepage maintained by Carnegie Mellon University.
• Multiple View Geometry in Computer Vision, includes sample chapter on Epipolar Geometry and the Fundamentral Matrix.

 

Online help for Matlab

• Matlab help desk

 

If you are unfamiliar with Matlab have a look at Gareth's

• Matlab Image Processing Tutorial (2000).

Course Objectives

This course will present an introductory overview of computer vision. Students will be introduced to 2D image processing techniques (such as image enhancement, feature detection and segmentation) and 3D projective geometry, and will apply these skills to real-world problems such as face detection and determining 3D structure. The course will be very practically oriented, with a strong emphasis being placed on applying the tools taught in lectures to actual images. All labs and assignments will be done using Matlab and the Matlab Image Processing Toolbox which will provide a user-friendly platform for experimenting with computer vision techniques.

Feedback

Any feedback on the course is welcome, be it regarding lectures, labs, this website or any other aspect. Email me luke@syseng.anu.edu.au or come and see me after a lecture.

Lab Allocations:

Tuesday 9-12:

1. Karl Pietsch
2. Chris Madden
3. David Tychsen-Smith
4. Alex Talberg
5. Jason Saragih
6. Aaron Watters
7. Daniel Frampton
8. Joanna Nicholls
9. Andrew Walter
10. Dave Ferrari
11. Clive Rossiter
12. Mark Euston
13. Joel McDonald
14. Yee Harn Teh
15. Omar Al-Kadi
16. Michael Sexton
17. Grant Baldwin
18. Robert Ewin
19. Jim Baker
20. Michael O'Connor
21. Adam Wood
22. Tod Sirawattananon
23. Nopparat Phongsajjanukul

Wednesday 11-2:

1. Harsha Seneviratne
2. Jonathon Kocz
3. David Biddle
4. Ian McRobert
5. Simon Green
6. Arved von Brasch
7. Eric Oberlin
8. Max Guyatt
9. Joel Mortimer
10. Nicolaus Strauch
11. Supakit Charnvanichborikarn
12. Daniel Biglia
13. Emily Butlin
14. Nathan Cattle
15. .
16. .
17. .
18. .
19. .
20. .