Databases Reference
In-Depth Information
download from Web servers to browser-equipped clients all over the world.
The problem is exacerbated by the inherent delay of these networks, which
is caused by such things as resource contention, congestion avoidance,
and node failures. This means individual packets almost always end up tak-
ing different routes to their destination.
Bit-intensive image files and video streams must therefore be com-
pressed and otherwise manipulated to minimize the download time, or risk
having users avoid the Web site entirely. Although some compression tools
will compress an entire database of images and decompress them dynam-
ically for access over the Web, this does not solve the fundamental prob-
lem of transferring large images over the Internet. The best way to com-
press a large image database on a Web server is to ensure that the right file
type is used in the first place.
Several compression standards have been issued to meet varying applica-
tions needs. For image files, a compression standard has been issued by the
Joint Photographic Experts Group (JPEG). For video streams, a compression
standard has been issued by the Motion Picture Experts Group (MPEG). For
other image databases used in the advertising and publishing businesses,
for example, there is fractal compression. There are many standards for
compressing audio files, including GSM, which is the same compression
algorithm used in conjunction with digital cellular telephone service.
JPEG
Most Web browsers support two image formats: GIF and JPEG. In general,
line drawings and flat-color illustrations are better stored in GIF format,
whereas photographs and complex images are more suitable for JPEG for-
mat. Of the two types of image files, only the latter offers user-selectable
compression. There is no compression selection with GIF files, but most
graphics programs allow users to specify high, medium, or low compres-
sion for JPEG files. Other image formats can be used on the Web, but they
require the use of an external viewer, which the browser must call up and
open. The time it takes for this process to work can disrupt the smooth flow
of information presentation. Most consumers have only low-speed dial-up
connections to the Internet and have little patience for long file downloads.
JPEG was designed for compressing continuous-tone (photographic-
quality), monochrome and color still images. It is based on a discrete
cosine transform (DCT)-based algorithm, which eliminates redundant
image information by dividing the image into small blocks. The higher the
compression ratio, the more information that is deleted and, therefore, the
more image detail that is lost.
JPEG can be used to compress images at ratios as high as 25 to 1 without
a noticeable loss of image quality. This is because the human eye has the
Search WWH ::
Custom Search