Processamento de Imagem e Visão

Image Processing and Vision, Fall Semester 2011/2012

João Paulo Costeira - (jpcosteira AT ist.utl.pt)

Announcements :
Insert Info about your project type a description and group members
SEE POSTERS HERE - poster session Friday 6th Jan 9.30AM
RESULTS OF PROJECT 1 HERE

PROJECT VIDEOS: See here videos of project demos !

FINAL SUBMISSION Project #2 dealine 12 Dec 24:00 -
Submission FORM here

POWERPOINT - TEMPLATE FOR POSTER

SUBMIT PROJECT 1 HERE | PROJECT 2 - One example

PIV is an introductory course to image processing and computer vision. In the end of the course students should be equipped with a "toolbox" of techniques which enable them to design algorithms to do, among other tasks, Image (view) quality improvement, remove image artifacts (noise, blur), detect and segment moving objects, select and match/track image patches/points along sequences, reconstruct 3D (and planar) objects from sequences and stereo pairs, compute pose and recognize image patterns and 2D shapes.

The course will follow closely course notes written by Prof. S. Marques though in a different sequence. The topic sequence will be close to that of R. Szelisky's book.

TOPICS

1-Introduction	2-Camera models	3-Image enhancement and restoration	4-Feature detection and matching
5-Segmentation	6-Motion Analysis	7-Recognition	8-Stereo

BIBLIOGRAPHY

Richard Szeliski, Computer Vision: Algorithms and Applications, draft, 2010 (http://szeliski.org/Book/)

J. Marques, course 2006-2010. (PDF)

Gonzalez and Woods – Digital Image Processing using Matlab

Slides : Presentation, Color, Camera, Filtering,Restoration,Features,Registration,Motion,Recognition Stereo all here

Tutorials you must read!

Least Squares tutorial

Singular Value Decomposition

GRADING

Final grade = project 1 (30%)+ project 2 (30%)+ written exam (40%)

Project 1	Calibrate the KINECT -- Develop a matlab program to compute kinect's parameters. Students must submit one matlab file and a 5 page report
	Groups: This project is done in groups of 2 students, randomly selected
	Deadline: Software - Nov 4th, Report - Nov 11th
Project 2	Take a KiNECT and bring one (cool) idea: Find an interesting application and present a concept/demo. you can use “anything” you wish. We encourage students to cooperate, brainstorm and use any software you may find. We will build a "community" software site.
	Groups: This project is done in groups of 2 or 3 (exceptionally 4) members without any constraint : You may form your group freelly . As a courtesy rule, ERASMUS students should join groups with Portuguese students.
	Project Proposal - You must submit a project proposal (1 or 2 pgs) by Nov 5th. please
	Deliverables - Could be a software demo, "a system", and one A0 Poster describing your idea. A poster session will be held on the last day of classes, where you have to present your idea to the course faculty and other people.
	Deadlines: Software/Demo - December 9th, A0 Poster December 12th.

ACKNOWLEDGEMENTS:.

To Prof. Manuela Veloso made us aware of the "leap forward" the kinect was, and offered us one kinect just two weeks after it was released (and nobody could get one!). Also, to the Carnegie-Mellon|Portugal program sponsored of the 10 kinects we are going to use in both projects. These contributions enabled this format to PIV

1º Exam Jan 13 (tentative, check official exam website)
2º Exam: Feb 1(tentative, check official exam website)

NEED INSPIRATION ? CHECK HERE