Glaucoma is a vision loss disorder caused by a rise in intraocular pressure that destroys the visual nerve fibers. Since there are usually no symptoms until significant vision loss develops, it is challenging to diagnose in the early stages. However, more precise and dependable glaucoma detection and diagnostic systems have been made possible by recent advances in deep learning and machine learning techniques. The objective of this proposed effort is to improve diagnosis accuracy and identify those who are at risk for glaucoma by examining the use of deep learning and the DenseNet algorithm to classify retinal fundus photos as normal or abnormal. This proposed method used a modified version of the DenseNet 121 algorithm, trained on a preprocessed dataset of optic disc pictures classified as either normal or glaucoma. This may make it possible to diagnose and treat glaucoma early on, lowering the chance of blindness. Although these results are promising, more research is required to determine the therapeutic effectiveness of this method in an actual setting. In order to improve the model's generalizability and explore the viability of applying this strategy in a clinical context, it is advised that future study enlarge the dataset. To effectively categorize images, this suggested model is optimized by converting the final output layer to a binary classifier and adding further optimizing layers.A separate set of test photos was used to assess the model's performance. Our findings showed that, with an overall accuracy of 96.28% and a loss of 1%, the modified DenseNet algorithm was highly accurate in classifying retinal fundus images as normal or glaucoma. This model emphasizes how deep learning techniques increase the accuracy of glaucoma diagnosis and detection. The modelclassified pictures as normal or glaucoma with remarkable accuracy using a modified version of the DenseNet algorithm. Lastly, a possible tool for the diagnosis and detection of glaucoma is the DenseNet algorithm, which has been improved by deep learning. Because of its great accuracy and potential for early detection, it is a valuable supplement to the diagnostic techniques already used by doctors,With more research and improvement, this method could improve patient outcomes and alter the diagnosis of glaucoma.