Automatic Disease Detection and Monitoring System in Plants Using K-Means Algorithm

Abstract

Agriculture is the main sources of all countries throughout the world. Many plants are easily affected to diseases due to changes in various factors such as soil moisture, varying climatic conditions, nutrients, temperature etc. Automatic plant disease detection and monitoring system using k-means algorithm helps to monitor the plant diseases through moisture values of the soil. It helps to reduce water use. The system is developed with threshold values of soil moisture sensor that is programmed into a microcontroller based gateway to control water quantity. It involves two processes such as disease detection and monitoring the soil. At first stage, wireless camera is used to capture the plant image and processed using image processing by Matlab software. Here the leaf images of paddy crops are already trained by artificial neural network by k-means algorithm. At second stage, soil moisture sensor values are compared with threshold values which are given to microcontroller. Depending upon the results of microcontroller, motor is pumped ON/OFF which is driven by driver circuit and relay. If the plant is infected to diseases, the information about the diseases is sent to the authorised person through GSM module. This system helps to increases the crop yield by identifying and classifying the diseases before it affects the entire field. Development of this system leads to high productivity.