PurposeGlaucoma is an eye disease that can result in perpetual eye vision loss if not perceived and treated in time. Therefore, an accurate and efficient automated method has been proposed for diagnosing Glaucoma. This method enables mass screening for such conditions and reduces the task of healthcare professionals. MethodologyThis proposed method involves splitting the fundus images into red, green, and blue channel components, where the green channel component is selected for further processing. The processed green channel of the images has been decomposed using Flexible analytical wavelet transform (FAWT) in sub-bands. Furthermore, the features are mined from decomposed green channel sub-band images using the modified Gauss-Kuzmin-distribution-based Gabor (GKDG) filter (eight scales and five orientations) and NCA by employing them to lessen the dimensionality of the silent features. Then, the selected features are graded using the student t-value, and the graded features are fed to the kernels like linear, gradient descent, Morlet, and Mexican-hat-based LS-SVM classifier to classify the Glaucoma. Results and ConclusionThe results of this research suggest that the newly proposed glaucoma classification model demonstrates superior performance compared to contemporary techniques. The model's evaluation was conducted using tenfold cross-validation, revealing an enhanced accuracy of 95.84 %, a specificity of 97.17 %, and a sensitivity of 94.55 %.