z-logo
open-access-imgOpen Access
Anatomical Properties and Near Infrared Spectra Characteristics of Four Shorea Species from Indonesia
Author(s) -
Danang Sudarwoko Adi,
Sung-Wook Hwang,
Dwi Ajias Pramasari,
Yusup Amin,
Hairi Cipta,
Ratih Damayanti,
Wahyu Dwianto,
Junji Sugiyama
Publication year - 2020
Publication title -
hayati journal of biosciences
Language(s) - English
Resource type - Journals
eISSN - 2086-4094
pISSN - 1978-3019
DOI - 10.4308/hjb.27.3.247
Subject(s) - principal component analysis , absorbance , spectral line , range (aeronautics) , mathematics , chemistry , analytical chemistry (journal) , botany , materials science , biology , statistics , chromatography , physics , astronomy , composite material
This study investigated the anatomical properties and absorbance characteristics of NIR spectra of four Shorea species from Indonesia. Macroscopic section revealed that Balau has similarity with Heavy Red Meranti, whereas White Meranti was almost identical with Light Red Meranti. All of the woods have diffuse porous and axial resin canals in tangential lines at the microscopic level. Original NIR spectra of Shorea species showed different absorbance characteristic. Wood density was assumed to be one of the factors that affected to the absorbances. Principal component analysis (PCA) of second derivative NIR spectra at the wavenumber 8,000-4,000 cm-1 (full) and 6,200-5,600 cm-1 (specific) showed different orientation among the Principal Component (PC) number. PC1, which contained highest spectral variation, had two closed clusters (1) Balau and Heavy Red Meranti and (2) White and Light Red Meranti at full spectral range. In contrast, the results at specific range were (1) Balau and White Meranti and (2) Heavy and Light Red Meranti. Hierarchical clustering dendrogram using PCA data from two spectral regions resulted in two types of clustering, the 8,000-4,000 cm-1 was somehow related to ‘density’, while the 6,200-5,600 cm-1 was grouped in ‘color’ information from visual inspection.  From both spectral regions, k-nearest neighbour (k-NN) classification models revealed 100% accuracy in identification four Shorea species using NIR spectra.

The content you want is available to Zendy users.

Already have an account? Click here to sign in.
Having issues? You can contact us here