USE OF FOURIER TRANSFORM INFRARED SPECTROSCOPY AND ARTIFICIAL NEURAL NETWORKS TO PREDICT THE WOOD DENSITY OF CEDRUS LIBANI A. RICH.

Abstract

The aim of this study was to measure the wood density of Cedrus libani A. Rich. samples from its Fourier Transform Infrared Spectroscopy spectrum. 40 density values were obtained by using 3600 properties belonging to C. libani tree in laboratory environment. Since 1045 properties between 832-1876 from 3600 properties were found to be sufficient to determine the density, 1045 data between 832 and 1876 were used for training and testing of the network. Data used as attribute were normalized between 0.1 and 0.9. 20% of the data were used as the test set and the remaining 80% of the data are used as the training set. This analysis indicated that Fourier Transform Infrared Spectroscopy combined with Artificial Neural Network can be used to measure the density of wood in less effort and in less time than other laboratory methods.