CAP 5015 - Multimedia Compression on the Internet

Fall 2009
MW 4:30pm to 5:45pm Room ENGR 0286
Keycode:93780

Prof. Amar Mukherjee
Office 304 HEC
Office and/or email hours: MW 3:15-4:15pm
Email Address: amar@cs.ucf.edu
Telephone: 407-823-2763

Academic Calendar

Methods of Evaluation (Subject to Change)
Homework Assignments 25%;Programmimg Assignment: 5% Class participation and attendance: 5%; First Midterm: 15% Second Midterm: 15%; Term Project: 20% Final: 15% Grading: >95% is A; 90-95% A-; >85% B+, 82-85% B,80-81% B-
C and D grades are broken down similarly like B in the 70's and 60's percentages. Any score below 60% is F.

---

Final Exam Schedule

Background:

In recent times, we have witnessed an explosive growth of multimedia technologies.The need to store and transmit large masses of information is growing very rapidly along with increasing number of applications in diverse fields covering the spectrum of human civilization in the new millenium. Engineers and computer scientists need a fundamental understanding of data compression technologies to work with large variety of data ( text, image, video, sound and any combination of these types) types and increasingly data-intensive applications. This course will provide the fundamental theories and techniques of data compression. The students will benefit from the balanced presentation of basic theory, algorithms and implementations.

Pre-requisite:

Senior standing with background on design and analysis of algorithms. Open to all graduate students. This course would be especially useful to students in image and text processing, computer vision, graphics, visualization and internet technology but that other students will also benefit from the course.

Course Outline

Why data compression? The bandwidth gap for multimedia data: text, true color images, binary images, video and sampled sound. Input source format.

Ad hoc compression methods. Lossles and lossy compression algorithms.The components of a data compression system. performance metrics.

Introduction to Information Theory.

Statistical compression methods: Huffman and Arithmetic entrpy coders.The JBIG standards. Dictionary based compression methods: LZ family. Universal Lossless source coding. Model based compression methods: PPM, DMC. Transform based text compression: BWT transform. Work at UCF.

Image compression methods. Mathematical preliminaries. Scalar and vector quantization. Predictive coding. DPCM. Hierarchical vector quantization.

Transform Coding: DCT(JPEG/MPEG) and wavelet transform(EZW/RMF). Subband Coding. Wavelet Compression.

Brief discussion on Video and Sonud Compression.

Recommended Reference Text:

1. David Salomon. "Data Compression: The Complete Reference". Springer-Verlag, Second Edition, 2000.

2. I. H. Witten, A. Moffat, and T. C. Bell. "Managing Gigabytes", Morgan Kaufmann Publishers, Inc., Second Edition,2000.

3. Khalid Sayood, "Introduction to Data Compression", Morgan Kaufman, 1999.

4. J.D.Gibson,T. Berger, T. Lookabaugh, D. Lindbergh and R.L. Baker. " Digital Compression for Multimedia", Morgan Kaufman,1998.

5. Alistair Moffat and Andrew Turpin, ``Compression and Coding Algorithms", Kluwer Academic Publishers, 2002.

6. R. M. Rao and A. S. Bopardikar, "Wavelet Transforms: Introduction to Theory and applications", Addison Wesley, 1998.

7. Khalid Sayood(Editor), "Lossless Compression handbook",Elsvier/Science Academic Press, 2003.

8. E.J. Stollnitz, T.D. DeRose and D. H. Salesin, Wavelets for Computer Graphics",, Morgan Kaufman, 1996.

Assignments: Weekly/bi-weekly homework assignments,reading assignments, a term project, midterm and final examinations.

Webpage: The webpage for this course is http://www.cs.ucf.edu/courses/cap5015/.This webpage will be used to post most of the course material including lectures, homework assignments, term projects etc. If you click 'Lectures ' below you will get slides and lecture notes. These material are not intended to be complete set of notes. The material included in these slides and lectures are based on material contained in some of the references. These notes and slides should be used as supplementary material along with live lecture presentation. Copying for distribution and/or sale is strictly prohibited by copy right laws. Another useful webpage is http://vlsi.cs.ucf.edu/.
VLSI and Data Compression Lab: M5 Research Group
This webpage gives information on current research on data compression supported by Natinal Science Foundation at UCF. Of special interest to you will be the compression utility where you can click to try most of the currently available algorithms in the literature on text compression.