Multispectral Registration, Undistortion and Tree Detection for Precision Agriculture
No Thumbnail Available
Date
2019
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
Multi-lens multispectral cameras allow us to record multispectral information for a whole area of terrain, even though we may only need the vegetation data. Based on the intensity of each multispectral image we can retrieve the contours of the trees that appear on the recorded terrain. However, multispectral cameras use a physically different lens for each range of wavelengths and misregistration effects could appear due to the different viewing positions. As these types of lenses are dedicated to capture larger areas of terrain, their focal distance is lower and because of this we get what is called a fisheye distortion. Therefore if we want to retrieve the shape of each tree and its multispectral data we need to process the channels so them all are representated as undistorted images under a same reference system.
Description
@inproceedings{10.2312:ceig.20191209,
booktitle = {Spanish Computer Graphics Conference (CEIG)},
editor = {Casas, Dan and Jarabo, Adrián},
title = {{Multispectral Registration, Undistortion and Tree Detection for Precision Agriculture}},
author = {Ruiz, Alfonso López and Rodríguez, Juan Manuel Jurado and Anguita, Carlos Javier Ogayar and Higueruela, Francisco Ramón Feito},
year = {2019},
publisher = {The Eurographics Association},
ISBN = {978-3-03868-093-2},
DOI = {10.2312/ceig.20191209}
}