The Real-Time Mobile Application for Identification of Diseases in Coffee Leaves using the CNN Model
P Divyashri., Lishma Anisha Pinto, Ligina Mary, P. Manasa, Sandhya Dass
Abstract
Coffee plants are extremely susceptible to a wide range of pests and diseases. Usage of reckless pesticides can lead to increased pathogen resistance in the long run, severely limiting coffee plants' ability to fight. Plant health monitoring and disease detection in plants are very critical for sustainable agriculture. As there is a rapid rise in the number of diseases and low awareness of these conditions, disease detection and prevention remain a major concern. Coffee leaves have certain textures and visually striking similarities that can be used to identify the disease type. The proposed solution is used to detect coffee leaf diseases and classify them into five categories: healthy, diseased leaves with Brown eye spots, Coffee Leaf Blight, Coffee Leaf Rust, and Coffee Leaf Miner. The paper focuses on building a CNN model which processes the leaf images that can be used for the detection of plant diseases with digital image processing techniques for disease diagnosis using visual image processing techniques. This solution serves its purpose by identifying and classifying the diseases based on the features derived from the Coffee leaf picture samples with an accuracy of 88.35%. With the aid of the above technique, coffee plant farmers may be able to identify diseases more quickly, increasing India's coffee production output. The proposed system was designed to benefit farmers and the agricultural sector, especially in the Karnataka region.