Determining the average annual ring width on the front side of lumber

摘要

Visual features of lumber can be used to assure its quality in stiffness and strength. Specifically, the average annual ring distance of the planks and the position of the center of the annual rings of the front side supply a close relation to some quality parameters of planks. Unfortunately, it turns out to be difficult to detect the average annual ring width by simple image vision methods due to distortions in the front side image of a plank caused by the cutting process. In this paper we propose two integrating methods which are capable of being used in an industrial application. One is based on quantizations of color images, the other on local Fourier transformations to detect the main wave in an image.