Sawn timber defect identification system

Hau, Tze Min (2017) Sawn timber defect identification system. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Sawn timber defect identification system.pdf - Submitted Version Download (648kB)
	Text (Full Text) Sawn timber defect identification system.pdf - Submitted Version Restricted to Registered users only Download (1MB)

Abstract

Automated inspection of timber defect has shown to be of great importance in the wood industry. Due to the decreasing forest resources and increasing cost of timber, the application of automated vision inspection is seen as a solution to optimize resources and save production cost, while maintaining the output of products with reliable quality. This project aims to investigate suitable algorithm and develop a tool for identification of timber defects on timber images. The prototype will be tested against 8 types of natural defect commonly found on tropical timber. This project is a standalone system that will be develop using MATLAB. There are 3 phases in the methodology of this project. The first phase will focus on study and investigate the suitable algorithm of defect identification while in phase 2 we will develop a system based on Matlab. In phase 3, we will test our system using the data obtained from timber defect image database developed by Computation Intelligence and Technology Lab, UTeM (Hashim, 2015). The objective of this project is to investigate the suitable algorithm for defect detection as well as develop the system. Then, we will test the system againts types of defect and multiple species. The goal of this project is to develop a timber surface defect identification system which is capable of identifiying different types of timber defect on multiple species. This project has promising impact in quality control process of wood based industry with contribution to the automated visual inspection (AVI) domain.

Actions (login required)

View Item