An Optimized Precision in Tea Leaf Disease Identification using Faster R-CNN Model

Tea cultivation is often affected by various diseases that impact crop yield and quality. Early detection of these diseases is crucial for minimizing economic losses and ensuring sustainable agricultural practices. This study aims to develop an automated tea bud disease prediction system using the Faster R-CNN algorithm, which enhances detection accuracy and efficiency. The primary objective is to provide farmers with a highly sensitive and reliable system capable of identifying tea bud diseases at an early stage, thereby improving crop health management. A comparative analysis is conducted between two detection models: the existing YOLOv7-based system, known for its real-time object detection capabilities, and the proposed Faster R-CNN-based approach, which incorporates a Region Proposal Network (RPN) to improve accuracy. While YOLOv7 demonstrates efficiency in rapid detection, its precision is insufficient for detailed disease classification. The proposed Faster R-CNN model, trained on annotated images of tea buds, effectively distinguishes between healthy and diseased leaves, achieving a recognition accuracy of 85%. This advancement allows for real-time monitoring of tea plantations, aiding farmers in taking timely preventive measures. By integrating artificial intelligence into agricultural disease detection, this research not only enhances precision and reliability but also supports sustainable farming practices. The study demonstrates the potential of deep learning-based solutions in optimizing crop disease management, reducing dependency on manual inspections, and improving overall tea production efficiency.