Previous Work in Image Enhancement
Dr. Eugene Ageenko
The quality of document images may have faded during the document life cycle
and digitization process, while noise introduces unnecessary details in the
images. It degrades the image quality and weakens image compression. Several
filtering methods have been considered in the literature for image
pre-processing [TP80, Ber87, AKS90, ZD96]. These filters include logical
smoothing, variations of median filtering, isolated pixel removal, and
"crisp" and soft morphological filters [Ser82, Hei94]. All these
analyze the local pixel neighborhood defined by a filtering template. To accept
or reject the pixel, they use a set of rules, such as predefined masks or
quantitative description of the local neighboring area. Recent research in
mathematical morphology has shown that morphological filtering can be used as an
efficient tool for pattern restoration in an environment of heavy additive
noise, but it is not necessarily suitable for filtering the content-dependent
noise introduced by the image digitization process [SG91, Hei94, KA94, DA97].
References.
Research