CS 390 Special Topics in Computer Science: Image and Video Understanding, Fall 2009


Objectives

This course is to introduce principles and methods of obtaining information from images and videos. Topics include image features, image processing, shape analysis, image matching, 3D reconstruction, human pose and action, and modern image database and retrieval. The prerequisite of the course is CS1. No image processing and Matlab experiences are required.

Instructor

Hao Jiang, Room 554, Campanella Way 21, hjiang@cs.bc.edu, 617-552-8983.
Office hours: T TH 2:45-4:00pm in my office.

Lecture Schedule (Approximate)

Week Date Topic
1 09/08/2009 : 09/10/2009 Introduction. Image Basics  
2 09/15/2009 : 09/17/2009 Image Filtering  
3 09/22/2009 : 09/24/2009 Image Features  
4 09/29/2009 : 10/01/2009 ICCV09, Guest Lecture  
5 10/06/2009 : 10/08/2009 Shape  
6 10/13/2009 : 10/15/2009 Image Matching  
7 10/20/2009 : 10/22/2009 Camera  
8 10/27/2009 : 10/29/2009 3D Reconstruction from Multiple Views  
9 11/03/2009 : 11/05/2009 Motion  
10 11/10/2009 : 11/13/2009 Tracking  
11 11/17/2009 : 11/19/2009 Human Pose  
12 11/24/2009 : 11/26/2009 Thanks Giving (No Class)  
13 12/01/2009 : 12/03/2009 Action
14 12/08/2009 : 12/10/2009 Image Retrieval

Grading Assignments

Each week, there is something you need to practice. The homeworks are as follows. After week 10, you will focus on your final project. The assignments will guide you through a bunch of small projects. You will answer questions, do experiments, write programs, and analyze results.

Week 1 Image processing using ImageJ
Week 2 Using matlab in image processing
Week 3,4 Blob extraction
Week 5 Looking for similar shapes
Week 6 Object localization with Ransac
Week 7,8 3D shape from multiple calibrated images
Week 9 Motion Estimation
Week 10 Blob Tracking

Late Penalties


Textbook Reference books