Identification of Diabetic Retinopathy from Segmented Retinal Fundus Images by using Support Vector Machine
P.R. Thorat1, K.V. Patil2

1P.R. Thorat, Professor, Savitribai Phule Women’s Engineering College, Aurangabad, (Maharashtra). India.
2K.V. Patil, Student, Savitribai Phule Women’s Engineering College, Aurangabad, (Maharashtra). India.

Manuscript received on 15 April 2016 | Revised Manuscript received on 25 April 2016 | Manuscript Published on 30 April 2016 | PP: 93-98 | Volume-5 Issue-4, April 2016 | Retrieval Number: D4499045416/16©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Segmentation of images has become important and effective tool for many technological applications like vessel segmentation from fundus images, medical imaging and many other post-processing techniques. Diabetic Retinopathy is an eye disease which is caused by changes in blood vessels of retina of diabetes. It is the primary cause of blindness in the universe. To avoid blindness of diabetes detection of diabetic retinopathy as early as possible is the only option as number of persons are becoming blind because of this disease. Many studies have shown that early diagnosis is the most efficient way to cure this disease. This paper presents identification of diabetic retinopathy from segmented retinal images by using support vector machine. In pre-processing, first the input image is converted into green channel image and converted into binary image. After that we have segmented the vessels using thresholding. For tracing of vessels, graph tracer algorithm is used. Through the project we have developed an algorithm for identifying the diabetic retinopathy from fundus images. For identification we have used GLCM features and SVM classifier together. The results indicate a potential for developing an automatic algorithm to segment and trace vessels and diabetic retinopathy classification for planning of treating the disease. For this, we have collected the database of 24 retinal fundus images from Dongaonkar Eye hospital, Kranti chowk, Aurangabad. The proposed system is implemented in MATLAB software.
Keywords: Vessel Segmentation, Graph Tracer Algorithm, Feature Extraction, GLCM (Gray Level Co-occurrence Matrix), SVM (Support Vector Machine).

Scope of the Article: Algorithm Engineering