Early detection of coffee plant disease using the k-nearest neighbour (KNN) algorithm

Coffee plants (Coffea spp. L) are highly susceptible to pest and disease infestations. Diseases in coffee plants often manifest as changes in leaf colour and physical appearance. Coffee farmers and the general public face challenges in managing and maintaining coffee due to the following reasons: visible disease symptoms are often similar, lack of knowledge and experience in identifying disease types, and the vast coffee plantation areas with numerous coffee plants make monitoring and caring for coffee plants a significant challenge. Therefore, it is crucial for farmers and the public to have access to technology that can identify coffee diseases as part of early detection efforts to combat pest and disease attacks. Early detection is essential for effective disease control and to prevent potential declines in coffee productivity. Hence, a technology is needed to detect diseases in coffee plants, which would aid in disease identification and guide farmers and the public in applying appropriate treatments for specific diseases. Recognizing diseases at an early stage makes it easier to implement plant care, thus avoiding losses from coffee plant deaths and diminishing coffee bean quality. This research focuses on implementing a model for early detection of coffee plant diseases using the K-Nearest Neighbour (KNN) algorithm. The research results can be utilized to provide solutions for early detection or initial diagnosis of the types of diseases affecting coffee plants. The KNN algorithm was chosen due to its high accuracy value. The research stages are: data collection, data preprocessing, KNN algorithm implementation, and model evaluation. The disease detection model evaluation yielded an accuracy rate of 98%, a recall rate of 98%, and a precision rate of 98%. From the evaluation results, it can be concluded that this model exhibits excellent performance in detecting coffee plant diseases.