Taxonomy for lossless image compression

Nasir D. Memon, Khalid Sayood

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We give a classification scheme for both decorrelation and coding. The two classifications put together give us a framework for discussing, comparing and evaluating lossless image compression schemes. The classifications also enable us to clearly identify areas where not enough work has been done, thus identifying avenues for future research. We identify three component functions in an image decorrelation scheme, which are induced by a given image. These are: 1) An Ordering function, 2) A Neighborhood function and 3) A Replacement function. These functions can be static, backward adaptive or forward adaptive Static functions are fixed for all images. The above definitions provide us with a classification scheme for image decorrelation techniques. We also give a simple classification of residual image encoding techniques, similar to the one for decorrelation schemes. While this is not exhaustive survey, it does point out a few interesting things. One is the relatively small amount of work in error modeling. Another is the paucity of adaptive ordering functions. Finally, the classification reinforces the view that the development of decorrelation techniques is far from being a mature area.

Original languageEnglish (US)
Title of host publicationProceedings of the Data Compression Conference
EditorsJames A. Storer, Martin Cohn
PublisherPubl by IEEE
Pages526
Number of pages1
ISBN (Print)0818656379
StatePublished - 1994
EventProceedings of the Data Compression Conference - Snowbird, UT, USA
Duration: Mar 29 1994Mar 31 1994

Publication series

NameProceedings of the Data Compression Conference

Other

OtherProceedings of the Data Compression Conference
CitySnowbird, UT, USA
Period3/29/943/31/94

ASJC Scopus subject areas

  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Taxonomy for lossless image compression'. Together they form a unique fingerprint.

Cite this