Tea Leaf Disease Classification Using AI

Abstract Tea cultivation represents a critical agricultural activity across several regions of Asia and Africa, contributing significantly to rural employment, national economies, and global beverage consumption. Despite its economic importance, tea production remains highly vulnerable to foliar diseases that directly affect leaf quality, yield, and long-term plantation sustainability. Diseases such as Red Leaf Spot, Brown Blight, Gray Blight, White Spot, and Anthracnose frequently occur in tea-growing regions and can spread rapidly if not identified at an early stage. Conventional disease diagnosis methods depend largely on visual inspection conducted by experienced farmers or agricultural experts. While effective in small-scale settings, this approach becomes increasingly inefficient, subjective, and error-prone when applied to large plantations or regions with limited expert availability. Recent advancements in artificial intelligence, particularly in the domain of computer vision, have introduced automated solutions capable of addressing these challenges. Convolutional Neural Networks (CNNs) have emerged as a powerful class of deep learning models due to their ability to automatically learn hierarchical visual features from raw image data. In agricultural applications, CNNs have demonstrated strong potential in identifying disease symptoms from leaf images, eliminating the need for handcrafted feature extraction and reducing human bias. However, the application of deep learning techniques to tea leaf disease classification remains comparatively underexplored, particularly using publicly available datasets and custom-built architectures designed for low-resource condition. Overall, this study highlights the feasibility of deploying deep learning-based diagnostic tools for tea leaf disease detection under limited data conditions. The proposed system offers a scalable, objective, and time efficient alternative to traditional inspection methods and serves as a foundation for future improvements through dataset expansion, transfer learning, and deployment on mobile or edge-computing platform.