Citation
Silva E. L., . and Vieira H. C., . and Muñiz G. I. B., . and Soffiatti P., . and Santos J. X., . and Nisgoski S., . Near infrared spectroscopy for separation of tauari wood in Brazilian amazon native forest. pp. 227-236. ISSN 0128-1283
Abstract
This work aimed to characterise different wood species known as tauari using the near infrared spectroscopy (NIR) as a support tool for identification. The spectra were obtained in a Tensor 37 spectrometer in the range of 4000“10000 cm-1 using 35 samples of solid wood from four different species with 21 different origins. Specimens measuring 2.5 centimeters long were extracted from each sample. The spectra were obtained on all faces of the samples and the analysis was performed in three different situations: i) sample mean ii) only cross-sectional spectra and iii) only longitudinal (tangential and radial) spectra. The classification methods tested were linear discriminant analysis combined with principal component analysis (PCA-LDA). The best pretreatment was the second derivative highlighting Couratari guianensis species with accuracy ranging from 55.6 to 61.1. The LDA classification based on information from the NIR spectra showed greater discriminatory potential for the species. However due to the complexity of the separation of the tauari wood the use of near-infrared (NIR) spectroscopy together with the anatomical identification is suggested for confirmation of the species.
Download File
Full text available from:
Official URL: https://info.frim.gov.my/infocenter_applications/j...
|
Abstract
This work aimed to characterise different wood species known as tauari using the near infrared spectroscopy (NIR) as a support tool for identification. The spectra were obtained in a Tensor 37 spectrometer in the range of 4000“10000 cm-1 using 35 samples of solid wood from four different species with 21 different origins. Specimens measuring 2.5 centimeters long were extracted from each sample. The spectra were obtained on all faces of the samples and the analysis was performed in three different situations: i) sample mean ii) only cross-sectional spectra and iii) only longitudinal (tangential and radial) spectra. The classification methods tested were linear discriminant analysis combined with principal component analysis (PCA-LDA). The best pretreatment was the second derivative highlighting Couratari guianensis species with accuracy ranging from 55.6 to 61.1. The LDA classification based on information from the NIR spectra showed greater discriminatory potential for the species. However due to the complexity of the separation of the tauari wood the use of near-infrared (NIR) spectroscopy together with the anatomical identification is suggested for confirmation of the species.
Additional Metadata
| Item Type: | Article |
|---|---|
| AGROVOC Term: | Near infrared spectroscopy |
| AGROVOC Term: | Lecythidaceae |
| AGROVOC Term: | Wood |
| AGROVOC Term: | Native forests |
| AGROVOC Term: | Species |
| AGROVOC Term: | Sampling |
| AGROVOC Term: | Spectroscopy |
| AGROVOC Term: | Moisture content |
| AGROVOC Term: | Cellulose |
| AGROVOC Term: | Hemicellulose |
| Depositing User: | Mr. AFANDI ABDUL MALEK |
| Last Modified: | 24 Apr 2025 00:54 |
| URI: | http://webagris.upm.edu.my/id/eprint/9093 |
Actions (login required)
 |
View Item |
