Abstract
This study was carried out to test methods for separating knots from clearwood in a digital image stack when scanning for internal defects with a medical CT-scanner. Scots pine knots, represented by its tangential surface density image extracted from a CT-image stack, have been classified by two different methods showing equal results. The knots are classified in four knot types by an Artificial Back-propagation Neural Network (ANN) and a Partial Least Squares Modelling with Latent Variables (PLS) model. The classification precision of aim of four different knot types, is between 85% and 97%. The results indicate that both methods may be useful tools in order to describe and classify knots in concentric surfaces around the pith in CT-images and thereby extract parametrical models from CT raw data image stacks. A simplified classification model has been obtained, by analysing the learning patterns in both the ANN and PLS model, that classify knots and transform density related data to segmented and classified parametrical descriptions.
Zusammenfassung
Kiefern-Rundholz wurde zum Erkennen innerer Holzfehler mit Hilfe der Computer-Tomographie untersucht. Aus dem gespeicherten Bild material sollten Astbereiche von fehlerfreiem Holz differenziert werden. Die Äste im Innern des Kiefernholzes konnten mit zwei verschiedenen Auswertesystemen aufgrund ihrer Dichteprofile erkannt und in vier verschiedene Klassen unterteilt werden. Die eine Methode benutzt ein künstliches neuronales Netzwerk (ANN = Artificial Back-propagation Neural Network), die andere basiert auf einem Algorithmus, der geeignet ist, sehr verrauschte oder unvollständige Datensätze auszuwerten (PLS = Partial Least Squares Modelling with Latent Variables). Die Treffsicherheit der Einordnung in vier Güteklassen lag zwischen 85% und 97%. Beide Methoden sind somit geeignet, Astfehler in konzentrischen Schichten innerhalb von Rundholz zu klassifizieren. Aus den gespeicherten Bildern nach der Computer-Tomographie können Qualitätsparameter des untersuchten Holzes abgeleitet werden. Ein vereinfachtes Schema ergab sich nach Ausnutzung der Lernfähigkeit des ANN- und des PLS-Systems. Nach entsprechendem Training (anhand visueller Klassifizierung der aus dem Holz gesägten Bretter) werden die Dichteunterschiede der Rohdaten in Klassifizierungs-Parameter umgesetzt.
Similar content being viewed by others
7 References
Birkeland, R.; Holöyen, S. 1987: Industrial methods for scanning of log defects. A progress report on ongoing project in Norway. Proceedings 2nd International Conference on Scanning Technology in Sawmilling. Oct. 1–2, Oakland, CA.
Brunner, C. C.; Maristany, A. G.; Butler, D. A.; Funck, J. W. 1993: Enhancing Color-image Data for Wood-Surface Feature Identification. Proc. 5th Int. Conf. on Scanning Technology & Process Control for the Wood Products Industry. 25.27 Oct. Atlanta, USA.
Chang, S. J.; Wang, P. C.; Olson, J. R. 1987: Nuclear magnetic resonance imaging of hardwood logs. Proceedings 2nd International Conference on Scanning Technology in Sawmilling. Oct. 1–2, Oakland, CA.
Duda, R. O.; Hart, P. E. 1973: Pattern classification and scene analysis. John Wiley & Sons, Inc pp. 10–39.
Eberhard, R. C.; Dobbins, R. W. 1990: Neural Network PC Tools. A practical Guide. Academic Press, Inc.
Grundberg, S.; Grönlund, A. 1991: Methods for reducing Data when Scanning for Internal Log Defects. Proceedings 4th International Conference on Scanning Technology in the Wood Industry. Oct. 28–29. San Francisco, CA.
Hagman, O.; Grundberg, S. 1993: Multivariate Image Analysis Methods to classify Features on Scots Pine: Evaluation of a Multisensor Approach. Proceedings of the 5th International Conference on Scanning Technology & Process Control for the Wood Products Industry. Atlanta, USA.
Hagman, O. 1993: Automatic Quality Sorting ofPicea abies Logs with a Gamma Ray Log Scanner. Scand. J, For. Res. pp. 583–590.
Lindgren, L. O. 1991: Medical CAT-scanning: X-ray absorption coefficients, CT-numbers and their relation to wood density. Wood Science and Technology. 25: 341–349.
Richard, M. D.;Lippmann, R. P. 1991: Neural Network Classifiers Estimate Bayesiana posteriori Probabilities. Neural Computation 3, pp. 461–483.
Roder, F.; Scheinman, L. E.; Magnuson, P. 1989: High speed CT scanning of logs. Proceedings 3rd International Conference on Scanning Technology in Sawmilling. Oct. 5–6, Oakland, CA.
Rummelhart, D. E.;McClelland, J. L. 1986: Parallel Distributed Processing-Explorations in the Microstructure of Cognition, Volume 2: Psychological and Biological Models, MIT Press, Cambridge. Ch. 22.
Ruck, D. W.; Rogers, S. K.; Kabrisky, M.; Oxley, M. E.; Suter, B. W. 1990: Multilayer Perceptron as an Approximation to a Bayes Optimal Discriminant function. IEEE TRANSACTIONS ON NEURAL NETWORKS. Vol. 1. No 4 Dec.
Spreeuwers, L. 1992: Image filtering with neural networks. Applications and performance evaluation. Thesis. University of Twente. Netherlands. pp. 92–93.
Wold, H. 1966: Nonlinear Estimation by Iterative Least Squares Procedures. Research Papers in Statistics, Festscrift for J. Neyman, David, F. N. Ed.; Wiley, New York.
Wold, S. 1989: Multivariate Data Analysis: Converting Chemical Data Tables to Plots. Computer Applications in Chemical Research and Education. Dr Alfred Hüthig Verlag, Heidelberg.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Hagman, P.O.G., Grundberg, S.A. Classification of scots pine (Pinus sylvestris) knots in density images from CT scanned logs. Holz als Roh-und Werkstoff 53, 75–81 (1995). https://doi.org/10.1007/BF02716393
Issue Date:
DOI: https://doi.org/10.1007/BF02716393