Aflatoxin is a carcinogenic toxin produced by a soilborne fungi, called Aspergillus flavus, causing a difficult struggle for the peanut industry in terms of produce quality, price and the range of selling market. This study aims to develop a successful U-Net CNN (Convolutional Neural Network) model, a reliable image segmentation method, that will help in distinguishing high probability zones of occurrence of Aflatoxin in peanut fields using remotely sensed hyperspectral imagery. The research was delegated to two objectives; 1.) Enabling research to understand and mitigate aflatoxin contamination in the field: Infield aflatoxin hotspot prediction and management and 2.) Deep learning-based postharvest aflatoxin detection using hyperspectral fluorescence imaging.