CIS 595

1
CIS 595 Content Based Image and Video Retrieval
Dr. Rolf Lakaemper
2
About myself

• Rolf Lakaemper
• PhD (Doctorate Degree) 2000
• Hamburg University, Germany
• Since 1/2003 Assist. Professor at Department of
  Computer and Information Sciences, Temple
  University
• Main Research Area Computer Vision

3
Today

• Introduction Computer Vision Image Databases
• The Course (Grading etc.)
• Introduction to MATLAB
• Overview
• Some Details
• Computer Vision Examples

4
Computer Vision ? Multimedia Databases ?
5
Computer Vision ? Computer visions great trick
is extracting descriptions of the world from
pictures or sequences of pictures (Forsyth/Ponce
Computer Vision)
6
Computer Vision and Databases

• How to
• Represent
• Process / Prepare
• Handle
• Recognize
• Retrieve
• images / image objects

Comp. Vision Databases
7
Representation

• Digital Images
• Color Spaces
• Gray Images
• Binary Images
• Geometrical Properties

8
Representation

• Digital Images
• Color Spaces
• Gray Images
• Binary Images
• Geometrical Properties

9

• Processing
• Filters
• Edges
• Geometric Primitives
• Lines, Circles

10
Processing Segmentation


11

• Handling
• Image Data Representation
• Image / Video Formats
• JPEG
• GIF
• MPEG
• (this lecture will NOT cover this !)

12

• Low Level Object Handling
• Object representation

13

• Object Recognition
• Color, Texture, Shape

14

• Object Recognition
• Applications
• Character recognition
• Face Recognition
• Shape Recognition
• Motion, Movement Detection
• Behaviour Analysis

15
Retrieval Image-Databases
16
Example The ISS Database (Interactive Shape
Selection)
17
The Interface (JAVA Applet)
18
The Sketchpad Query by Shape
19
The First Guess Different Shape - Classes
20
Selected shape defines query by shape class
21
Result
22
Specification of different shape in shape class
23
Result
24
Let's go for another shape...
25
...first guess...
26
...and final result
27
Query by Shape, Texture and Keyword
28
Result
29
CIS 595
30
CIS 595

• Schedule
• MATLAB Introduction
• Basic Topics in Computer Vision
• Feature Extraction
• Database Techniques for Image Databases
• Presentations
• Final Exam
• Total 14 lectures Exam

31
CIS 595
TEXTBOOKR. Veltkamp, H. Burkhardt, and H.-P.
Kriegel (eds.) State-of-the-Art in
Content-Based Image and Video Retrieval Kluwer
Academic Publishers, Boston, 2001. 150 at
Amazon - Different Research Articles
32
CIS 595
Grading 20 Quizzes Class Activity 40
Project Presentation 40 Final Exam
(negotiable)
33
CIS 595
The Project will be a task of reading,
understanding and programming has to be
implemented in MATLAB, JAVA or C must be
presented in class using Power Point Slides
34
CIS 595
Team Work is allowed (except in the final exam),
giving you the opportunity to emphasize your
strength in either reading ? , programming or
presenting inside a group of 2 or 3.
35
CIS 595

• The website
• http//knight.cis.temple.edu/lakaemper/courses/ci
  s595_2004
• Will contain
• ppt-slides
• MATLAB files
• Research Articles
• (The Temple - Blackboard Tool will not be used)

36
CIS 595
My office Wachman Hall, 534 Hours Mo 5pm
715pm by App. Email lakamper_at_temple.edu Phon
e 215 204 7996
37
CIS 595
now lets start with MATLAB