Automatic Diabetic Retinopathy Diagnosis using Sobel Edge Detection from Fundus Imaging

Abstract

In the era of image processing various diseases can be diagnosed automatically without any human intervention. A symptomatic disease can be recognized either of the imaging techniques such as X-Ray, CT Scan, Fundus imaging and MRI using various approaches present in the field of digital image processing. Diabetic Retinopathy is a symptomatic disease which is very dangerous because it leads to permanent blindness that cannot be cured. But routine checkup can save their vision that can be done under proper scanning and observation. It may cause due to diabetes and most of the cases pertained due to long diabetic duration. Here the proposed system is able to detect Diabetic Retinopathy automatically using Sobel edge detection technique and color mapping. Sobel is the best way to analyze the sensitive data by extracting the impairments in eyes or fundus imaging. Fundus imaging is a scanning technique for obtaining retinal image. System achieved 93% of accuracy which is bit higher than the previous works.