Welcome to RelliableJPEG homepage

This research was supported in part by the NSF through grant CCR-0104851
 
Project Team
Robert G. Redinbo
3110 Engineering II
Phone: (530) 752-3087
FAX: (530) 752-8428
E-mail:
Cung Nguyen
2211 Engineering II
Phone: (530) 756-3243
FAX: (530) 752-8428
E-mail:

Abstract

Data compression techniques are the central to many types of modern transmission systems. The compressed data formats are sensitive to channel or storage errors. The integrated error control schemes, therefore, are being used to protect against those errors. However, at a more fundamental level, compression is even more susceptible to the different errors that are due to failures in the computing hardware and software facilities implementing the compression algorithms. The research project is developing new techniques to detect errors in both hardware and software integrated throughout the compression and transmission parts of the systems; enhancing international compression standards by including protection against hardware and software failures. As electronic structures become smaller and employ low-power methods, computer-induced errors will become important, particularly in medical and remote sensing applications. There is currently little active research on this subject because there is no immediate commercial interest and it is very difficult to address the broad impact of systems failures and errors in algorithms designed on the data.

Table of Contents

  1. Introduction
  2. Work Done and Roadmap
  3. Overview the JPEG Image Compression Systems
  4. Fault Tolerance Designs for JPEG Still Image Compression Systems
    1. Discrete Cosine Transform (DCT)
    2. Huffman Coding
    3. Quantization
    4. Error Checking for JPEG Compression Formats
  5. Technical Reports and Implementation Programs
    1. Research Proposal on Fault Tolerance Designs for Image Compression Systems (1.4Mb)
    2. Presentation on Fault Tolerance Issues in JPEG (4Mb)
    3. Fault Tolerance Design for 8x8 2D-DCT ( 600k)
    4. Fault Tolerance in JPEG Image Header Using Backus-Naur Form ( 1.2Mb)
    5. Concurrent Error Detection in Wavelet Lifting Transforms ( 997k)
    6. Protecting EBCOT Encoding System for JPEG2000 - Fast Abstract ( 47k)
    7. Protecting Wavelet Lifting Transforms ( 328k )
  6. JPEG Organization Links
  7. Acknowledgements