Vegetation Index Estimation in Precision Farming Using Custom Multispectral Camera Mounted on Unmanned Aerial Vehicle

Abstract

In the last decades several vegetation indices were introduced in order to quantify the spectral properties of green plants. This paper presents an integrated solution (hardware and software) affordable for acquiring and processing spectral data in order to estimate the vegetation status. A brief description of what are the vegetation indices and why are they important is followed by the hardware description, software setup and experimental workflow. The entire hardware system consisting of a modular multispectral camera and the Small Board Computer that controls the acquisition process, was developed within a research project taking into account certain conditions as modularity, low cost, small size, low weight, remote data access. The multispectral camera is carried by an Unmanned Aerial Vehicle (UAV) in order to take images of the investigated area. Regarding the software, the developed applications are mainly based on open-source solutions: a web-based interface designed for the acquisition control and one special program developed for vegetation indices computation. Other applications are used for data processing. Georeferenced orthophotomaps are generated using acquired images, each one for one wavelength of the multispectral camera. Using these data, different types of vegetation indices are computed and corresponding maps are generated. The key outcomes of the study are related to the multispectral camera that has a modular design, offering flexibility in changing the desired wavelengths and also to the data processing solutions.