Recognizing and Classifying Sound Propagation in Oil Palm Stem for Tomography Construction

Abstract

This study is a part of research works concerning with the early detection of Ganoderma infection of oil palm tree. Sonic tomography equipment is used as a non invasive technique to determine the modification of the internal oil palm stem caused by the Ganoderma fungus. The oil palm stem will be exposed to the sound wave from many directions and the sound velocity on a sound line can be calculated. This research is articulated in three phases. Phase one, consists of selecting the value of the velocity at the intersection points between the different sound line. The selection is based on the use of rules develop from the expert’s knowledge. Phase two, consists of classifying the velocity of the intersection points into several classes dealing with the wood condition, by using an unsupervised classification method, Fuzzy C-Means. Fuzzy classification allows gradual memberships, therefore offers the opportunity to deal with data that belong to more than one class at the same time. Phase three, the interval of velocities from one intersection to another will be also classified. Finally, the results of the classification obtained are used to develop a simple tomographic image of the oil palm cross section, by extrapolation.