Trends in Horticulture (2021) Volume 4 Issue 1
doi:10.24294/th.v4i1.1797
29
Original Research Article
Volumetric spatiality of wood in forests with artificial neural net-
works based on satellite imagery
Argemiro José Moreno Arteaga
*
, Monica Fabiana Bento Moreira Thiersch, Roberta Oliveira Averna Valente,
Lucas Kröhling Bernardi, Sophia Lourenço Vasconcelos, Cláudio Roberto Thiersch
Universidad Federal de São Carlos (UFSCar), Brazil. E-mail: argemor29@gmail.com
ABSTRACT
Sustainable agriculture of forest plantations requires the permanent estimation of the amount or volume of
wood being produced at any given time, which is difficult in large forest areas if only manual procedures are used in the
field. In the present research, multilayer perceptron artificial neural networks (ANNs) were modeled for the spatial es-
timation of wood volumes in a Eucalyptus sp. plantation located in the state of Mato Grosso do Sul, Central-West re-
gion of Brazil. For this purpose, spectral bands, band textures obtained with gray level cooccurrence matrix and vegeta-
tion index, which were derived from SPOT 6 digital satellite image, were used as prediction variables. The resulting
ANN with the best performance presented an accuracy of 93.32% and a coefficient of determination of 0.9761, with
respect to values obtained with field measurements; however, it presented a relative mean square error of 16.32%
(RMSE of 7.85 m
3
/hm
2
), but the distribution of residuals was not biased, therefore, the model was promising for map-
ping timber volumes in large areas without overestimating or underestimating the prediction. The constructed network
showed greater precision and accuracy when compared to other methods using similar estimation variables, including
when compared to neural models using only spectral bands and vegetation indices.
Keywords: Biomass; Dendrometry; Remote Sensing; Multilayer Perceptron; Texture
ARTICLE INFO
Received: 18 January 2021
Accepted: 8 March 2021
Available online: 27 March 2021
COPYRIGHT
Copyright © 2021 by author(s).
Trends in Horticulture is published by
EnPress Publisher LLC. This work is li-
censed under the Creative Commons At-
tribution-NonCommercial 4.0 International
License (CC BY-NC 4.0).
https://creativecommons.org/licenses/by-nc/
4.0/
1. Introduction
Wood obtained from Eucalyptus sp. plantations is a fundamental
resource for the series of socioeconomic services it provides in the
manufacture of furniture, paper and energy generation, which has stim-
ulated its massive production in an economically sustainable manner
with efficient forest management of plantations, which requires perma-
nent monitoring of the volumes of wood produced.
The estimation of the amount of timber available or produced in
forest plantations is typically done through allometric equations using
tree measurements obtained in the field with trained personnel and spe-
cialized materials and equipment, a difficult task to do in large areas,
having as a conditioning factor the monetary investment and time
[1]
.
The development of human knowledge has led to the integration of
technologies in the activities involved in production systems, increasing
their competitiveness in the market and reducing several activities that
were previously performed manually. In this sense, digital images ob-
tained by satellite with remote sensing are useful technologies for ob-
taining land cover information, reducing field actions that can lead to
procedural errors
[2,3]
. Image components such as spectral and de-
rived bands of textural features and vegetation indices can be used as