Computational Intelligence in Agriculture

Abstract

Agriculture is the backbone of a country's economic system as it produces food, one of the fundamental needs for human beings and provides employment for the speedily growing population. The easy and convenient availability of low-cost, low-energy sensors made it possible to enhance food production, introduce automation and reduce human efforts. Intelligent sensing for agriculture incorporates remote sensing using satellite or aircraft, covering a large spatiotemporal and in- accessible agricultural area without disturbing the people and the environment. Smart sensing for agriculture facilitates quantitative and qualitative improvements in agricultural productions. The volume of the collected sensory data is huge and continuously increasing. It is beyond human capability to analyze all the data and make a correct decision. As existing computer-based agricultural systems are not good enough, new soft computing paradigm known as Computational Intelligence for Agriculture (CIA) is being pursued. CIA incorporates Machine learning (ML) and Deep Learning (DL) models to handle the increasing complexity for maximizing production. ML techniques construct the computer program to learn on its own using historical data and take proper decisions the way humans do. DL techniques mimic the human brain in processing actual data and identifying patterns from the data in decision making for agriculture using multiple layers. Moreover, the CIA automate multiple processes such as monitoring the soil and plant health, the growth rate of crops, detecting abnormalities in crop growth, improving productivity and quality, etc. This chapter discuss various uses of the CIA for agriculture, such as estimation and improvement of crop yield, water conservation, and environmental change, soil and plant health, fertilizers, and pesticides as well as plant disease detection. © 2022 John Wiley & Sons, Ltd.