Crop Yield Prediction Using Artificial Intelligence and Remote Sensing Methods

Abstract

Agriculture is a major part of the economy in the majority of developing and third world nations. An accurate and well-timed crop yield prediction will not only help in crop management, crop insurance but will also facilitate policy and decision-makers to frame apt strategies and policies regarding food security to combat hunger and eventually achieve zero hunger, one of the most important Sustainable Development Goals. Over the past decades, crop yield has been predicted through mathematical, statistical, and survey-based models. Artificial intelligence (AI)-based methods such as Artificial Neural Network (ANN), Support Vector Machine (SVM), Random Forest, and Deep learning methods could be a potential replacement to statistical modelling, because it produces precise results and can handle complexity and nonlinearity in data much more effectively. There are several fields in agriculture in which remote sensing can be advantageous, viz. crop yield forecasting, soil property detection, crop type classification, and meteorological data assessment. This chapter provides a framework of the existing methodologies of crop yield prediction and aims to describe the recent crop yield prediction techniques based on artificial intelligence and remote sensing approaches. It also focuses on the potential advantages of artificial intelligence methods on crop yield prediction at field and regional levels.