Agriculture is the backbone of the Indian economy. Now that crop production is decreasing more quickly, farmers find it difficult to sell their goods when their output is decreased due to diseases that affect trees, particularly the leaves. To improve quality and productivity, it is imperative to treat any serious illnesses as soon as possible. This issue led to the creation of expert systems for disease prevention and innovative technologies for the detection and diagnosis of plant diseases. A detailed literature survey highlighted the applications of different architectures of deep learning models for identifying infected and healthy leaves. Most of the existing literature commonly emphasized on the use of machine learning techniques for activities like sorting, disease detection, fruit identification related to agriculture. This paper uses three deep learning models YOLOv8, Faster R-CNN, and Custom CNN to classify whether the leaf is healthy or diseased. A dataset containing two classes of photos of healthy and infected leaves has been used, and YOLOv8 achieves the highest accuracy of 96.56% and among all the parameters of precision, recall, and F1- score when compared to Faster R-CNN with an accuracy of 87.98% and Custom CNN which achieved an accuracy of 85.64% for the classification task. However, more optimization is required, taking into account variables like dataset variations, computational performance, and real- world applicability.