This research work explores the landscape of automated crop disease identification techniques, focusing on grape leaf disease detection. The study highlights the prevalent use of publicly available crop leaf image datasets in existing research, which often lack the complexity of real time datasets, leading to challenges in accurate disease identification and generalization. With a significant portion of research contributions relying on such datasets, there is a critical need for advanced techniques, particularly utilizing Convolutional Neural Networks (CNN), to process real-time images effectively. To address these limitations, this survey emphasizes the necessity for a more comprehensive and precise automated system for grape leaf disease recognition. The proposed approach involves leveraging CNN models on a newly curated grape leaf image dataset and implementing tailored hyperparameter tuning to enhance system performance and generalizability.