Smart irrigation system for the reinforcement of Precision agriculture using prediction algorithm: SVR based smart irrigation

This paper explains the improvements in Precision agriculture interfaced with Iot applications. The dearth of water resources across the globe has provoked the need of optimum utilization. To scheduled irrigation process some constraints have to be measured so that irrigation happens only when it is required. To Drive with this application evapotranspiration is one such parameter which helps in detecting the soil moisture content and its retain ability with real time data from sensors and weather forecast. With the application of SVR + k, the accuracy of the system is increased with low minimum square error. This paper explains the approach of nonlinear characteristics with respect to kernel canonical correlation analysis(KCCA).With this exertion the input vectors of the prediction model is shortened by using the kernel functions. To estimate the soil moisture content the prediction algorithm is programmed with the support vector machine based kernel functions and to detect the evapotranspiration rate the canonical correlation analysis is is used. Finally the results experimented show the actual value for scheduling the irrigation process with respect to threshold. With SVR the accuracy is developed.