Mapping Red Clover Ground Cover Using Unmanned Aerial Vehicles

Date

2015-05-14

Authors

Abuleil, Ammar

Journal Title

Journal ISSN

Volume Title

Publisher

University of Guelph

Abstract

In the field of precision agriculture (PA), Unmanned Aerial Vehicles (UAVs) are creat- ing new opportunities for remotely assessing various characteristics of crops. This thesis presents two main contributions: an integrated system for collecting, preprocessing and analyzing aerial data for the novel application of mapping RCGC at a patch-level, and a collected, ground-truthed, and preprocessed RCGC dataset that will be made public for further analysis. Several different machine learning classifiers were evaluated for map- ping image patches to discrete clover coverage levels, reaching an accuracy of 91%. The best performing classifier was then tested for its robustness and adaptiveness and obtained satisfactory performance considering the size of the dataset being used.

Description

Keywords

precision agriculture, red clover, remote sensing, classification, machine learning, ground cover, vegetation index

Citation