CarDen: A software for fast measurement of wood density on increment cores by CT scanning

Autor: Jean-Michel Leban, Baptiste Kerfriden, Antoine Billard, Philippe Jacquin, Fleur Longuetaud, Frédéric Mothe
Přispěvatelé: SILVA (SILVA), Institut National de la Recherche Agronomique (INRA)-AgroParisTech-Université de Lorraine (UL), Unité de recherche Biogéochimie des Ecosystèmes Forestiers (BEF), Institut National de la Recherche Agronomique (INRA), ANR-11-LABX-0002-01, Institut National de la Recherche Agronomique (INRA)-Université de Lorraine (UL)-AgroParisTech
Jazyk: angličtina
Rok vydání: 2019
Předmět:
0106 biological sciences
Mean squared error
Reference data (financial markets)
Root mean square difference
Soil science
analyse non destructive
[INFO.INFO-SE]Computer Science [cs]/Software Engineering [cs.SE]
Horticulture
01 natural sciences
Cross-validation
Wood quality
Software
Non destructive measurement
cerne d'accroissement
rayons x
forest inventory
inventaire forestier
Computed tomography
Fast measurement
business.industry
scanner tomographique
logiciel informatique
Forestry
04 agricultural and veterinary sciences
Tree ring analysis
15. Life on land
Mass measurement
Computer Science Applications
040103 agronomy & agriculture
computer software
0401 agriculture
forestry
and fisheries
business
Agronomy and Crop Science
Geology
010606 plant biology & botany
Volume (compression)
densité du bois
Zdroj: Computers and Electronics in Agriculture
Computers and Electronics in Agriculture, Elsevier, 2019, 156, pp.606-617. ⟨10.1016/j.compag.2018.12.008⟩
ISSN: 0168-1699
DOI: 10.1016/j.compag.2018.12.008⟩
Popis: Increment cores are often used in wood science for measuring wood density of trees non destructively and at large scale, for instance with the objective to assess the available biomass in a forest resource. This paper presents a software allowing to measure by X-ray computed tomography (CT) the wood density of thousands increment cores. The software is able to process 3000 cores per hour semi-automatically. Manual intervention may be required to control and eventually adjust the positioning of the cores. The software was tested on 30 trees from 13 temperate species. Two increment cores were taken from each tree: one 5 mm diameter core and one 4 mm diameter core. The obtained CT density of the cores was compared to reference data obtained by volume and mass measurement on the 5 mm cores. The reference data were used for tuning the software settings by leave-one-out cross validation method. The obtained root mean square error was below 10 kg/m3 (1.7%) for the 5 mm cores. For the 4 mm cores, the root mean square difference with the reference density of the 5 mm cores was 25 kg/m3 (4.2%).
Databáze: OpenAIRE
Externí odkaz: