Effective feature selection plays a vital role in anterior segment imaging for determining the mechanism involved in angle–closure glaucoma (ACG) diagnosis. The aim of this project is to use anterior segment (AS– OCT) optical coherence tomography images to diagnose complicated diseases like Angle Closure Glaucoma. The collected input images are subjected to pre–processing. In the Pre–processing step, Sobel filter is implemented. In the feature extraction process, it implements the canny edge and Sobel is applied to extract the feature. In the feature selection process, we can apply the Watershed algorithm. In this step to implement it, the Naive Bayes Classifier is used. Naive Bayes classifiers are highly scalable, requiring some parameters, linear in the number of variables (features/predictors). Finally analysis of the performance of the process to find the TP, TN, FP, FN, Accuracy, Precision and recall is done. It has been shown that the accuracy in diagnosis of Angle Closure Glaucoma was better using L–Score method as compared to MRMR method