Novel Approach to Improve Crop Yield and Yield Prediction Using Remote Sensing the data for Precision Agriculture

Abstract

Precision agriculture, a transformative force in modern farming, has evolved significantly newlinewith the integration of remote sensing data. This innovative approach, utilizing advanced newlinetechnologies like satellite and drone-based imagery, alongside sophisticated data analytics, newlineempowers farmers to optimize practices and enhance productivity. This shift not only newlinereshapes agriculture but also addresses challenges posed by global food demand, resource newlineconstraints, and environmental sustainability. Some of the papers highlight the importance of newlinepredicting agricultural yield as a crucial task for nations. The escalating global food demand newlinedue to population growth impacts prices and availability. Crop yield forecasting emerges as a newlinesolution, enabling the prediction of market prices, planning imports and exports, minimizing newlinesocio-economic impacts, and facilitating humanitarian food assistance. While traditional newlinemethods like manual surveys are expensive and challenging to scale, crop simulation models newline(CSM) face limitations in data availability, especially in developing countries. Remote newlinesensing, defined as observing objects without physical contact, emerges as a preferred newlinetechnique due to its affordability and the availability of vast collections of free and opensource newlinedata. newlineThe work performed in the course of this research demonstrates the effectiveness of newlinecombining Sentinel-2 data with biophysical variables and vegetation indices to accurately newlinemap rice fields in the Davangere region. Utilizing supervised and unsupervised algorithms newlinelike K-Means and Random Forest, the study achieves accurate rice yield estimation. The newlinemethodology proves transferable to various satellite data sources. Dynamic variables such as newlineNDVI, SAVI, LAI, fAPAR, CI, and water content, along with the commencement date of rice newlinecultivation, play crucial roles. Disturbingly, the study reveals a significant conversion of rice newlinefields into plantations, leading to reduced yields. newlineIn this reasearch work, Artificial Neural Networks (ANN) enter the scene for paddy