Abstract :
Diabetic retinopathy (DR) is one of the serious
eye diseases and it originates from diabetes mellitus and is
the most common cause of blindness in diabetic patients.
Early treatment can prevent patients from being affected by
this condition or at least the progression of DR can be
slowed down. A key feature to recognize DR is to detect
micro aneurysms (MAs) in the fundus of the eye. Micro
aneurysms can be detected by excluding spurious
candidates that are effectively detected using MA detector
based on the combination of preprocessing methods and
candidate extractors. In this work, an integrated approach is
proposed for automated micro aneurysm detection with
high accuracy by candidate objects which are first located
by applying a dark object filtering process. Naïve Bayes
Classification and examining cross-sectional images are
prominently used. Newer and efficient methods are
implemented to counter the drawbacks present in earlier
ones. The proposed project determines the image-level
classification rate of the ensemble and hence promises to be
effective.
Keyword :
Computer-aided diagnosis, Image classification, Microaneurysm detection, retinal image, singular spectrum analysis, Diabetic retinopathy