Efficient and Accurate Hemorrhages Detection in Retinal Fundus Images Using Smart Window Features

Diabetic retinopathy (DR) is one of the diseases that cause blindness globally. Untreated accumulation fat and cholesterol may trigger atherosclerosis in diabetic patient, which obstruct blood vessels. Retinal fundus images are used as diagnostic tools to screen abnormalities linked affect eye. Blurriness low contrast major problems when segmenting retinal images. This article proposes an algorithm segment detect hemorrhages The proposed method first performs preprocessing on Then a novel smart windowing-based adaptive threshold utilized hemorrhages. Finally, conventional hand-crafted features extracted from each candidate classified by support vector machine. Two datasets evaluate algorithms. Precision rate (P), recall (R), F1 score for quantitative evaluation segmentation methods. Mean square error, peak signal noise ratio, information entropy, also method. achieves high with 83.85% DIARETDB1 image dataset 72.25% DIARETDB0 dataset. adequately adapts compared algorithms, hence will act tool segmentation.