Smart Agriculture Implementation using IoT and Leaf Disease Detection using Logistic Regression

Abstract

The need of robotics, automation, IoT and its implementation based on precision agriculture is very much necessary to eliminate the usage of manpower in farming. In order to make this more efficient system, adding machine learning based leaf disease detection is very much needy. This can result in saving the time, energy of the farmer and thereby improve productivity and efficiency. Usage of various electronic sensors and use of cloud-based service can also facilitate accurate measurement of any parameters and fast processing in regard to farming. A prototype of an agricultural rover to perform ploughing and seed sowing, a separate automatic irrigation, and fertilizer sprinkling system are made to facilitate effective cultivation. The rover is controlled by a smartphone through Wi-Fi module. The irrigation, fertilizer sprinkling system is automated and the parameters are made to be displayed on a cloud based IoT analytics service called thing speak and OLED. For the leaf disease detection part, machine learning based logistic regression is used with some optimizations to improve the accuracy.