Front Inner Page - Volume 2 No.3 June 2015

  • » Back to Index

  • Title

    :

    Segmentation and Classification of Diabetic Retinopathy Images Using Multiple Features

    Authors

    :

    M.Shankar1, Dr.K.Batri2, C.Vincent Raj3

    Keywords

    :

    Retinal Image; Diabetic Retinopathy; Micro aneurysm; Classifiers; Segmentation

    Issue Date

    :

    June – 2015

    Abstract

    :

    Diabetic retinopathy is becoming as one of the common disease it also can lead to blindness. The longer a patient with diabetes more likely developing diabetic retinopathy. DR retinal blood vessel collapse is. Micro aneurysm is one of the earliest symptoms of Diabetic Retinopathy. Due to the swelling of capillaries and weak blood vessels isolated dark red spots are created which are called as micro aneurysms. Based on the number of micro aneurysms the severity of the DR disease. Earlier micro aneurysms detection can reduce the incidence of blindness. Micro aneurysms are reddish in colour with a diameter less than 125 μ m. Detection of micro aneurysms in automated screening of diabetic retinopathy will be highly helpful in diagnosis and treatment. Generally MAs will appear as small red dots on retinal fundus images. In this paper we purpose a modified pre-processing method to remove background region and noisy pixels from retinal image, and to improve the efficiency of the detection median filter used along with Histogram. Feature extraction is done by GMM, KNN, and LMSE methods and Classification is done by RBF, Bays network. Based on the analysis carried out, the proposed system exhibits remarkable outputs.  

    Page(s)

    :

    1-6

    ISSN

    :

    2347- 4734

    Source

    :

    Vol. 2, No.3, June 2015

    Download

    :


  • » Back index