Main Article Content
Abstract
There are two major class of characteristics for wood species identification. The first class is general
characteristics such as colour, odour, wood grain, texture etc, those can be observed directly by common
sense,namely the eyes without using aditional tools ecxept a loupe with at least ten times magnification. The
second class is anatomical characteristics which provide wood structure including morphology and type of cell
wood components as well as their distribution, which can be observed by using microscope. By the artificial
intelligence system, wood species which one commonly used for industries become easier to identify. It takes
shorter time compared with the conventional activity. Aim of research was to create the neuro-fuzzy system
model, which able to identify wood species for construction and furniture utilizations based on their wood
anatomical characteristics, namely vessel element (their distribution, frequency and size) and ray parenchyma
(frequency, wide, high). Objects in this research were rubber wood, keruing, kamper, acacia, meranti and
jelutong.
The Neuro-Fuzzy System developed could identify with 0% errorness if the system using the same data in
the training process or using training sintetic data more than 1000 data.
Kata Kunci: anatomic characteristic, neuro-fuzzy, wood class identification.