Low-cost and Transferable Precision Agriculture Technique to Monitor Crop Anomalies in Rice Fields

Authors Organisations
Type

Student thesis: Doctoral ThesisDoctor of Philosophy

Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award date2022
Links
Show download statistics
View graph of relations

Abstract

Global agricultural land is expected to keep expanding in the coming years, especially in Sub-Sahara Africa and Latin American countries. However, many existing Precision agriculture (PA) techniques are challenging to transfer to agricultural systems in these regions as they rely on prohibitively expensive crop monitoring systems. Satellite remote sensing has the potential to support the delimitation of Site-specific Management Zones as it offers an inexpensive and non-destructive way of providing frequent information systematically at different spatial scales. This research presents the Earth Observation-based Anomaly Detection (EOAD) approach: a system that detects underperforming areas in croplands using medium and high-resolution satellite EO imagery. The EOAD is a simple anomaly detection technique, based on the deviation of image statistics from a normal distribution using dynamic thresholding, without the need for manual calibration or prior expertise in spectral analysis of crops. The EOAD approach demonstrated a strong agreement, 80% overall accuracy, with field observations of crop anomalies within rice plots in the Ibague Plateau, Colombia, using vegetation indices derived from optical Sentinel-2 and PlanetScope imagery. Areas identified as anomalous during the booting stage were significantly (p