A Quick Defect Detection Algorithm for Magnetic Resonance Images of Hardwood Logs
Advances in computer image interpretation are allowing computers to identify defects on scan data and images of various materials to improve product quality. Identification of defects such as knots in logs before the cutting operation would allow the lumber mill to maximize the value of usable lumber from each log. This paper presents images obtained from scanning an oak log with magnetic resonance imaging (MRI). The unique characteristics of MRI images of hardwood logs are noted and are used to derive a quick algorithm to isolate defects. Defect scans have at least some areas that vary considerably in intensity from their background. These areas provide a seed for defining the defect region. Because of growth rings, there is overlap between the image intensity of the defects and clear wood. Therefore, traditional thresholding techniques will not cleanly separate these regions. Instead, region-growing methods are used to extract the defects from the surrounding normal wood. The algorithm grows the defect region seed until the region-border pixel gray levels approach the median level of the neighborhood. The results show that region-growing methods obtain accurate borders of defects.
Forest Products Journal
Coates, E. R.,
Chang, S. J.,
Liao, T. W.
(1998). A Quick Defect Detection Algorithm for Magnetic Resonance Images of Hardwood Logs. Forest Products Journal, 48(10), 68-74.
Available at: https://aquila.usm.edu/fac_pubs/5093