http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
( Te Ma ),( Tetsuya Inagaki ),( Satoru Tsuchikawa ) 한국목재공학회 2021 한국목재공학회 학술발표논문집 Vol.2021 No.1
This work was aimed to provide a rapid and nondestructive imaging method for visualizing the dynamic state of free, hydrogen-bonded water with lignocellulosic material. Near-infrared (NIR) spectral images in the wavelength range 1002-1847 nm was firstly used to visualize the distributions of moisture content (MC) over the surface of Japanese cedar by partial least square regression. Then, principal component analysis and curve fitting methods were utilized to explore the changes in water-wood structure characteristics based on peak shifts to longer wavelength in spectral signals caused by increasing MC. The experimental results were clear showing that the earlywood regions have higher MCs in the initial stage of drying, but their free water evaporates more rapidly than that in other regions. Furthermore, the edge of the samples dried most rapidly into strongly bonded water. It is concluded that NIR hyperspectral imaging has the potential to be a complementary methodology for studying the transient changes of wood-water interactions.
Deep Learning Approach for Visible and NIR Imaging for Wood Classification
( Tetsuya Inagaki ),( Te Ma ),( Satoru Tsuchikawa ) 한국목재공학회 2021 한국목재공학회 학술발표논문집 Vol.2021 No.1
From the viewpoint of combating illegal logging and examining wood properties, there is a contemporary demand for a wood species identification system. Several nondestructive automatic identification systems have been developed, but there is room for improvement to construct a highly reliable model. We tried to identify the Japanese hardwood species from the microscopic image from database of Forestry and Forest Products Research Institute, Japan, with aid of deep learning. We also propose cognitive spectroscopy that combines near infrared hyperspectral imaging (NIR-HSI) with a deep convolutional neural network approach. We defined “cognitive spectroscopy” as a protocol that extracts features from complex spectroscopic data and presents the best results without human intervention. Overall, 120 samples representing 38 hardwood species were scanned using an NIR-HSI camera. A deep learning prediction model was built based on the principal component (PC) images obtained from the PC scores of hyperspectral images. The results showed that the accuracy of wood species identification based on 6PC (PC1-PC6) images was 90.5%, which was considerably higher than the accuracy of 56.0% obtained with conventional visible images.