Parallel/Distributed Computing for Digital on Demand

Advances over the past decade in many aspects of digital technology -
especially devices for image acquisition, data storage, and bitmapped printing
and display - have brought about many applications of digital imaging. However,
these applications tend to be specialized due to their relatively high cost.
One of the key obstacles for these applications is the vast amount of data
required to represent a digital image directly. Use of digital images often
is not viable due to high storage or transmission costs, even when image
capture and display devices are quite affordable. Therefore, image compression
is required to reduce the amount of data and consequently reduce high
storage and transmission costs.

Modern image compression technology offers a possible solution. 
State-of-the-art 
techniques can compress typical images from 1/10 to 1/50 their uncompressed
size without visibly affecting image quality. Digital image applications 
have become widespread in today's market place, so that there are different 
image compression technologies are exist.
A standard image compression method is needed to enable interoperability of
equipment from different manufactures. A standardization effort known by
the acronym JPEG (Joint Photographic Experts Group) has been working toward
establishing the first international digital image compression standard for
continuous-tone (multilevel) still images, both gray-scale and color. The
"Joint" in JPEG refers to a collaboration between CCITT and ISO. The
JPEG standard is based on the Adaptive Discrete Cosine Transform (ADCT)
algorithm. 

The three main elements of the JPEG standard are an encoder, 
a decoder, and an interchange format. An encoder is an embodiment of 
an encoding process. An encoder takes as input digital source image data 
and table specifications, and by means of a specified set of procedures 
generate as output compressed image data.
A decoder is an embodiment of an decoding process. A decoder takes as input
compressed image data and table specifications, and by means of a specified
set of procedures generates as output digital reconstructed image data.

In this demonstration, we apply parallel and distributed computing to perform
the tasks required to compress, transfer, and uncompress a digital image in
software using parallel as well as distributed computer architectures.
In this demo, we evaluate the performance of 4/8 nodes of CM-5 parallel
computer against the performance of several distributed system 
architectures: 4/8 workstations of IBM SP-1, 4/8 workstations of DEC
Alpha cluster interconnected by an FDDI network, and 2/4 workstations
connected by an ATM network. We use AVS to visualize the computations and
the application performance on these different computing system 
architectures.