CS 545I - Advanced Image Databases

Winter quarter, 1996
Fridays, 3:15 - 4:30
Gates building, room 100

Dragutin Petkovic (petkovic@almaden.ibm.com) and 
Oscar Firschein (oscar@infolab.stanford.edu) use for queries

sponsored by Gio Wiederhold, (gio@cs.stanford.edu).
Some References.

Planned Schedule for CS545I, (click for class summaries)

  1. Friday January 12: (Initial, organizational meeting) Oscar Firschein, Dragutin Petkovic, Gio Wiederhold
  2. Friday January 19: Overview of Issues Oscar Firschein
  3. Friday January 26: Basic Issues in Content-based Retrieval Dr. Dragutin Petkovic
  4. Friday February 2: Retrieving Images on the Web Dr. Dragutin Petkovic
  5. Friday February 9: Site Model Oscar Firschein
  6. Friday February 16: Content-based image retrieval, where queries are image-like objects Prof. Carlo Tomasi
  7. Friday February 23: Compression and indexing

  8. Gio Wiederhold
  9. Friday March 1: Medical Imaging Prof. Parvati Dev
  10. Friday March 8: Content Management for Multimedia Project Development - User' View Christine M. Okon
  11. Friday March 15: Wrap Up, MPEG Student presentation
  12. Friday March 22: Student presentations

Seminar abstract

CS 545I is a reading and projects seminar. The topic is advanced image databases. The emphasis will be on combining image-derived descriptors and text databases to retrieve local and online images. Applications will be surveyed to illustrate the strengths or weaknesses of specific techniques. This seminar focuses on the following topics:
  1. The problem of describing image content -- manually, semi-automatically, or automatically -- and retrieving images based on these descriptions. Topics also include image capture, annotation, data population and indexing.
  2. Experiments using an interactive image retrieval system. The technical basis for the Ultimedia Manager will be discussed in detail and class projects using the system will be carried out.
  3. On-line image databases based on WWW. We will review unique WWW issues, and look into several available applications, including Query by Image Content (QBIC) on WWW.
  4. The RADIUS system for interactive image analysis of aerial images uses 3-D site models linked to text databases. The site model provides a powerful indexing mechanism to both image and text databases.

Seminar format

A typical seminar meeting will consist of an instructor or student presentation of papers from the literature and a discussion of the material. Credit for the seminar only and a presentation will be 1 unit. Additional units (3 normally) under CS-395 will be available for students who wish to execute a project of their own choosing. We will use the IBM Ultimedia Manager for doing research in image databases.


Students should have a background in database concencpts of at leat CS145.


Dragutin Petkovic is the Manager, Advanced Algorithms, Architectures and Applications at IBM Resesarch in the Alamaden Laboratory. He has led the development of the QBIC project. QBIC received the Seybold award for the most innovative software product in 1995.

Oscar Firschein recently retired from SRI International and an assignment to ARPA, where he was Program Manager for Image Understanding. He has performed research and supervised many advanced projects in image processing. He is, with Marty Fischler, co-author of a survey book in the topic.

Tentative course outline:

Introduction (2 lectures):
The users/customers of retrieval from image databases
Problems in describing image content.
Image indexing for content-based retrieval.
Automated and semi-automated approaches.
Image capture, annotation and database population
Integrating image-derived descriptors and conventional SQL-like databases.
Overview of current work in image content based retrieval and current applications

WWW - on-line image databases
Overview of some specific issues like low bandwidth and how to overcome them, browsers, copyright issues etc. We will have WWW access in the class and look into several systems on the WWW, including QBIC.

Introduction to QBIC with demo
Details on the technology behind QBIC (texture, shape characterization) Review of the work of others like MIT Media Lab PhotoBook.Medical images and their use: radiography, tomography, magnetic resonance imagery

Coordinated 3-D site models, imagery, and conventional databases in RADIUS

Results of QBIC experiments

Summary and discussion
Future work and opportunities
- Image indexing
- Architectures of an extensible system where new features can be added dynamically
- Efficient integration within traditional databases
- Image databases on the WWW
- User interface issues

Equipment available for student projects:

We will use the IBM Ultimedia Manager, a set of tools that helps the user organize a visual database and find its cataloged images. The Ultimedia Manager uses both QBIC technology and traditional SQL. The equipment will be available to registered students on the 4th floor of the Gates building.