Empirical wavelet transform based pre-processing and entropy feature extraction from glaucomatous digital fundus images

The basic principle of feature extraction method in glaucomatous images is to represent the image in its compact and distinctive form. Glaucoma is a type of eye disease; it damages the optic nerve due to gradual increase in the fluid pressure and hence causes blindness. The available methods used in Heidelberg Retinal Tomography (HRT), Scanning Laser Polarimetry (SLP) and Optical Coherence Tomography (OCT) are costly and need experienced clinicians to use them. So, there is a need to analyze glaucoma accurately with low cost. Hence, in this paper, a new approach to extract the features is proposed which are further useful in automated diagnosis of glaucoma. Based on 2 Dimensional Empirical Wavelet Transform (2D EWT) entropy features are extracted. The 2D EWT is used to decompose the image and entropy features are obtained from decomposed EWT components. These extracted features are evaluated and used for the comparison of glaucoma images by 2 Dimensional Discrete Wavelet Transform (2D-DWT) and 2D EWT. The 2D EWT is employed for comparison with preprocessing and regularization. The results obtained put forward that the extracted features from 2D EWT are better for proposed method.