Glaucoma can cause permanent vision loss, and also early identification is challenging due to less symptoms in its early stages. We can slow the progress of glaucoma with appropriate treatment if it is diagnosed timely. The computer-based ophthalmologic disease detection technique helps eye professionals in the diagnosis of glaucoma by automatic testing of retinal fundus images. Convolutional neural network (CNN) models have shown exceptional performance in various medical image analysis tasks over conventional techniques. Therefore, incessant exploration of the related methods is required. In this research paper, the CNN variant Exception is utilized to diagnose glaucoma by analyzing retinal fundus images. The model classified fundus images by automatically extracting features. The EyePACS datasets were utilized to analyze the proposed model. To examine the performance of the model, the EyePACS dataset is utilized for training, validation, and testing of the proposed model. The effectiveness of the Xception model is tested by different factors like accuracy, sensitivity, precision, specificity, and F1 Score.