2-D Empirical LP Wavelet Transform based Automated Framework for Glaucoma Screening

Abstract

Glaucoma is a severe condition that causes eyesight loss. The ability to recognize glaucoma in its early stages is critical in preventing long-term vision loss. This paper presents a two-dimensional empirical Littlewood—Paley (LP) wavelet transform (2D-EWT)-based method for glaucoma detection using retinal fundus pictures. For the decay of the preanalyzed photographs into different sub-bands, EWT is used in this investigation. High-frequency sub-band images are then used to compute the features. The ReliefF method chose the valuable descriptors from the extricated include set. Finally, selected descriptors are classified using the random forest (RF) classifier. We use the RIM-ONE public online glaucoma database for performance evaluation of the proposed framework.