Matlab implementation of jpeg compression

Michael Christensen
ECE 533 Project proposal
12.8.2005

	Abstract:

The jpeg image compression algorithm is one of the most common of
compression standards in general use.  The purpose of this project is to
create a matlab-only implementation of the standard which is well
documented to explain it's processes.  The project will result in a
number of implementations depending on the detail of understanding
wanted.  The proposed implementations follow:

1.  Basic compression of a set-size grayscale image, detailing the basic
functionality of the algorithm.  The compression ratio and other
variables will be set to default values and will not be mentioned. 

2.  Compression of variable-size color images, again not detailing
compression ratios or other complex subjects. 

3.  Compression of variable-size color images.  Details of the basic
compression will be removed and higher level concepts will be detailed. 
Default values will be provided but subject to change by the user.  In
this way, users will be able to experiment with different values, and
observe the results directly. 

In each implementation, a file will be created that details certain
values of the source and resulting file.  Values to be included are the
sizes of the original and resulting image, the compression ratio,
information loss, etc.  These values will hopefully be instructive to
the user in his/her understanding of image compression.