Enhanced on-line stress grading of sawn lumber using near infrared spectroscopy

Takaaki Fujimoto, Kazushige Matsumoto, Yohei Kurata, Satoru Tsuchikawa

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Near-infrared (NIR) spectroscopy, coupled with multivariate analysis, has been used to predict the wood properties of sawn lumber, whose diffuse reflection spectra were acquired under moving conditions (30 m/min). Models predicting for bending strength (Fb), modulus of elasticity in bending (Eb), dynamic modulus of elasticity (Efr), air-dried wood density (DEN), and moisture content (MC) were developed using the partial least squares (PLS) analysis. For all wood properties, calibration models showed relative strong relationships between laboratory-measured and NIR-predicted values with coefficients of determination ranging from 0.55 to 0.80. The vector of regression coefficient revealed the important explanatory variables included in each model. It was concluded that NIR spectroscopy could be applied to on-line grading of lumber quality.

Original languageEnglish
Title of host publication11th World Conference on Timber Engineering 2010, WCTE 2010
Pages1538-1542
Number of pages5
Publication statusPublished - 2010
Externally publishedYes
Event11th World Conference on Timber Engineering 2010, WCTE 2010 - Trentino, Italy
Duration: 20 Jun 201024 Jun 2010

Publication series

Name11th World Conference on Timber Engineering 2010, WCTE 2010
Volume2

Conference

Conference11th World Conference on Timber Engineering 2010, WCTE 2010
Country/TerritoryItaly
CityTrentino
Period20/06/1024/06/10

Keywords

  • Multivariate analysis
  • Near infrared spectroscopy
  • Nondestructive test
  • Stress grading

Fingerprint

Dive into the research topics of 'Enhanced on-line stress grading of sawn lumber using near infrared spectroscopy'. Together they form a unique fingerprint.

Cite this