A Unified Methodology for Early recognition of Diabetic Retinopathy

Diabetic retinopathy is the occlusion of the retinal blood vessel because of diabetic mellitus issues, which can result in visual diminution. The best way to solve this issue is to utilize a retinal screening device that can detect retinal impairment at a premature stage. This work proposes the use of morphological operations and segmentation techniques to locate blood ducts, exudates, and microaneurysms. The extracted features are then subjected to Haar wavelet transformation. After that, the principal-component analysis technique tends to improve Characteristic selection. The photos are categorized as diabetic or non-diabetic using back propagation neural network and one rule classifier approaches. Experiments were carried out on a publicly available dataset on diabetic retinopathy. Unfortunately, conventional methods are incapable of making the best use of global data, and model complexity is significant. This project proposes a method for detecting diabetic retinopathy that extracts the precise area and quantity of microaneurysms from color fundus images. Two aspects of this research have been determined, namely the number and area of MA. Green channel extraction, histogram equalization, and the morphological process will be employed as pre-processing approaches at first. Principal-component analysis (PCA), contrast limited adaptive histogram equalization (CLAHE), structural procedure, and averaging filtering are all employed to discover microaneurysms.