Disease Detection of Plant Leaves Using K-Means Clustering and CNN Algorithms

Detection of plant leaf disease is crucial for achieving good yields and maintaining crop health. This early identification aids farmers who may not be familiar with all crop diseases by notifying them so they can take immediate action. This paper presents a system for detecting plant leaf disease using K-Means Clustering and Convolutional Neural Networks (CNN) implemented in MATLAB based on a trained dataset. It includes an Arduino Mega 2560 microcontroller connected to a Camera module for capturing images of plants. Additionally, it monitors environmental factors such as temperature, humidity, and soil moisture rate using the DHT11 sensor and Soil Moisture sensor. These ecological factors can be used for special treatment of crops as per the conditions. After identifying an unhealthy leaf, the system activates the GSM module to send an SMS alert containing the disease name. It makes it easy to recommend pesticides and to take other actions to save crops.