Sistem Identifikasi Kayu Ramin Berbasis Citra Menggunakan Local Binary ------- Pattern dan Probabilistic Neural Network

Abstract

The purpose of this research is to create dekstop application system in effective and efficient way to identify the ramin wood and similar ramin wood based on imagery by using LBPV and PNN methods. LBP variance (LBPV) is used to characterize the local contrast information into the one-dimensional LBP histogram. This research uses the Probabilistic Neural Network (PNN) technique to classify the LBPV. This research was carried out using 21 different types of similar ramin wood and one species of ramin wood. For each type of wood, 20 image were collected. The results of this research conclude that of the 3 operators used one operator was obtained that produce the highest accuracy, which is LPBV (24,3) with 79,77% accuracy. The result of this research indicate that LBPV (24,3) is more accurate to distinguish ramin wood texture and similar ramin wood. This research used only a small-size database, so for further research is needed to use more feature extract methods and types of ramin wood and similar ramin wood