Vegetation mapping from vegetation indices using a uav-based sensor

Abstract

The current advances in technology in many fields have revolutionized conventional agricultural practices. However, the use of remotely sensed data for agricultural purposes has not been fully explored in Nigeria. This study explores this limitation to understand performance and usability. In this work, remotely sensed information in the form of UAV images were used to assess crop greenness and vegetation cover. A maize farmland of 6 hectares was captured in Ogun State using a DJI phantom 4 UAV (which operates in true colour, RGB);The 165 images acquired were mosaicked using Agisoft Metashape software. Vegetation cover and greenness were assessed through various RGB-based vegetation indices and the conclusion was that Red Green Blue Vegetation index (RGBVI) produced the best results in depicting both vegetation cover and greenness. Excessive Green Index (ExG) and Modified Green Red Vegetation Index (MGRVI) produced above-average results but were not as informative as RGBVI. From the maps of each vegetation index, information about crop greenness and vegetation cover was adequately derived. This study showed the adequacy of UAV-based sensors for vegetation mapping and is recommended.