IMAGE PROCESSING ALGORITHM FOR AUTOMATED DETECTION OF ANNUAL RING ORIENTATION ON WOOD SHINGLES CROSS SECTION

摘要

In this paper we introduce an automated detection algorithm to determine the annual ring orientation on wood shingles cross section. We use a window moving, frequency analysis and filtering procedure to evaluate local orientation coupled to a global integration approach to determine the main annual ring orientation on the cross section of wood shingles. Our algorithm is simple, fast and easy to implement. We compare our results with human decision on a large sample of wood shingles and find a match for almost all shingles. In most cases without ambiguity our automated detection algorithm was able to determine on which side of the shingle was the core of the log before the shingle was sawn. This information is lost in the wood industry when the shingles are sawn of the log and it needs to be recovered at a further step because the wood shingles will bow according to the annual ring orientation when exposed to outdoor climate conditions.