Automatic Retinal Lesions Detection of Diabetic Retinopathy Using Curvelet Based Enhancement Banuselvasaraswathy B1, Murugan C Arul2, Karthigaikumar P3 1Department of Electronics and Communication Engineering, Sri Krishna College of Technology, Coimbatore 2Department of Electronics and Telecommunication Engineering, Karpagam College of Engineering, Coimbatore 3Department of Electronics and Communication Engineering, Karpagam College of Engineering, Coimbatore Online published on 15 March, 2019. Abstract Diabetic Retinopathy is the eye disease caused in patient suffering from diabetes. Once the sugar level becomes high, it deteriorates retina leading to loss of vision. Therefore, lesion detection gained importance in treating diabetes. The main aim of this work is to propose an automatic detection of lesions using four steps such as pre-processing, optic disc removal and vessel extraction, post processing and candidate lesion detection. It starts with optic disc removal and vessel extraction followed by curvelet based enhancement to isolate dark lesions from retina. The optical disc is necessitous for this system. To segment microaneurysms from the retinal image we proposed morphological filtering techniques and transformation. Curvelet based enhancement is used to improve the parameters such as specificity, accuracy and sensitivity in the proposed system. Performance analysis is carried out for specificity, accuracy and sensitivity of the proposed method. Additionally red lesion defect and glaucoma defect is also analyzed in this work. Top Keywords Microaneurysms, Hemorrhages, Exudates, Lesions, Curvelet based enhancement, matched filter, Laplacian of Gaussian. Top |