Characterization of Defects in Lumber Using Color, Shape, and Density Information

To help guide the development of multi-sensor machine vision systems for defect detection in lumber, a fundamental understanding of wood defects is needed. The purpose of this research was to advance the basic understanding of defects in lumber by describing them in terms of parameters that can be derived from color and x-ray scanning technologies and to demonstrate how these parameters can be used to differentiate defects in lumber. Color and x-ray images of intergrown knots, bark pockets, stain/ mineral streak, and clearwood were collected for red oak (Quercus rubra), Eastern white pine, (Pinus strobus), and sugar maple, (Acer saccharum) Parameters were measured for each defect class from the images and class differences were tested using analysis of variance techniques (ANOVA) and Tukey’s pair-wise comparisons with = 0.05. Discriminant classifiers were then developed to demonstrate that an in-depth knowledge of how defect parameters relate between defect types could be used to develop the best possible classification methods. Classifiers developed using the knowledge of defect parameter relationships were found to provide higher classification accuracies for all defects and species than those which used all parameters and where variable selection procedures had been used.