SMART FARMING: MEASURING SOIL CONDITION & SELECTING BEST CROPS USING IOT SENSORS

Abstract

The growth of agriculture industry is depending on the available resources and the way of using those resources. It is very important to use available resources effectively to meet the maximum productivity and fulfil the demand for agricultural products that increasing with the growth of population. Farm land conditions such as soil parameters and environmental conditions are the most important parameters for growth of the crops. Using of traditional agriculture practices are not the best way to make decisions on farm land conditions. It is not easy and effective to use fertilizes, Water and chemicals without knowing real conditions. This can be a reason to waste fertilizer, money, water, time and to reduce the harvest. It is better to select the most suitable crop according to the current farm land conditions or change current farm conditions to reach optimum conditions of selected crop. The main objective of this research was to design and implement a device to measure the farm land conditions in highest accuracy using IOT device and present those data to farmers through a mobile application. Further, propose the most accurate classification algorithm through Naive Bayes, Bayes Net, Random Forest and Random Trees algorithms to predict the most suitable crop for current farmland conditions. This device facilities farmers to stay touched with their farm land from anywhere in the world and this is a step to improve smart agriculture concepts using IOT technology. And also, this device facilitates to understand patterns of environmental conditions using past data through web interface. This system is able to send notifications to farmers whether any parameter is overdue or not its optimum conditions. As the result of this research it is found that the random forest classification algorithm is the most accurate algorithm from selected classification algorithms in crop prediction