|
|
New techniques are evolving in bio-medical field which are helpful in diagnosis of diseases in early stages of onset as well as
from distanced locations to shorten the response time for treatment. One such technique involves comparison of image(s) of
diseased part of body of the patient with those stored in data base(s) for diagnostic purposes. The challenge lies in comparison
being as realistic as possible to make the diagnosis full proof. This paper reports an improved technique of retrieval of best-fit
image for a test image utilizing multi-level imaging based on texture features, LBP features, retrieval feature, which are
processed using hash coding classifiers andrelevance feedback to get improved final result.
|