Advertisement

Wood Science and Technology

, Volume 53, Issue 1, pp 275–288 | Cite as

Classification of thermally treated wood using machine learning techniques

  • Vahid Nasir
  • Sepideh Nourian
  • Stavros Avramidis
  • Julie Cool
Original
  • 77 Downloads

Abstract

Classification of thermally modified wood is a critical assessment and control task that assures the quality of thermally treated wood. Machine learning methods can be used for identifying the optimal feature(s) for wood classification. In this study, the performance of artificial neural networks (ANN), support vector machines (SVM), and naïve Bayes (NB) classifiers for thermowood classification was evaluated and compared. The moisture content, water absorption, swelling coefficient, color, hardness, and dynamic modulus of elasticity of untreated and thermally treated western hemlock wood were measured and analyzed to identify the optimal set(s) of feature(s) for wood classification. The results showed that mechanical attributes such as dynamic modulus of elasticity obtained from the stress wave timer test and wood hardness account for the least suitable features, whereas color measurement provided an accurate classification. Both SVM and naïve Bayes model showed significantly higher performance than ANN because the latter requires a higher number of tuned and optimized parameters. Having only one feature, the accuracy of SVM and naïve Bayes model obtained from the color lightness parameter (L*) was 0.960 and 0.949, respectively. By increasing the dimension of the features, naïve Bayes model outperformed SVM and resulted in a robust classifier with an accuracy of 0.990. A trade-off between increasing the model accuracy and minimizing the number of selected features was observed. The SVM and NB models showed promising performance for the classification of thermally modified wood, which could be implemented for in-line quality control.

Notes

Acknowledgements

This work was partially funded by the Natural Science and Engineering Research Council of Canada (NSERC) (Grant No. RGPIN-2015-03653).

Compliance with ethical standards

Conflict of interest

There is no conflict of interest associated with this research.

References

  1. Al-Aidaroos KM, Bakar AA, Othman Z (2012) Medical data classification with naive Bayes approach. Inf Technol J 11(9):1166–1174CrossRefGoogle Scholar
  2. ASTM D1037-12 (2012) Standard test methods for evaluating properties of wood-base fiber and particle panel materials. ASTM International, West Conshohocken, PAGoogle Scholar
  3. ASTM D143-14 (2014) Standard test methods for small clear specimens of timber. ASTM International, West Conshohocken, PAGoogle Scholar
  4. ASTM D2244-16 (2016) Standard practice for calculation of color tolerances and color differences from instrumentally measured color coordinates. ASTM International, West Conshohocken, PAGoogle Scholar
  5. ASTM D2395-17 (2017) Standard test methods for density and specific gravity (relative density) of wood and wood-based material. ASTM International, West Conshohocken, PAGoogle Scholar
  6. ASTM D4442-16 (2016) Standard test methods for direct moisture content measurement of wood and wood-based materials. ASTM International, West Conshohocken, PAGoogle Scholar
  7. Avramidis S, Iliadis L, Mansfield SD (2006) Wood dielectric loss factor prediction with artificial neural networks. Wood Sci Technol 40(7):563–574CrossRefGoogle Scholar
  8. Bächle H, Zimmer B, Wegener G (2012) Classification of thermally modified wood by FT-NIR spectroscopy and SIMCA. Wood Sci Technol 46(6):1181–1192CrossRefGoogle Scholar
  9. Bedelean B, Lazarescu C, Avramidis S (2015) Predicting RF heating rate during pasteurization of green softwoods using artificial neural networks and Monte Carlo method. Wood Res 60(1):83–94Google Scholar
  10. Brischke C, Welzbacher CR, Brandt K, Rapp AO (2007) Quality control of thermally modified timber: interrelationship between heat treatment intensities and CIE L* a* b* color data on homogenized wood samples. Holzforschung 61(1):19–22CrossRefGoogle Scholar
  11. Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297Google Scholar
  12. Dunn D (1992) A preliminary assessment of the Metriguard 239A stress wave timer. Dissertation, University of CanterburyGoogle Scholar
  13. Esteban LG, de Palacios P, Conde M, Fernández FG, García-Iruela A, González-Alonso M (2017) Application of artificial neural networks as a predictive method to differentiate the wood of Pinus sylvestris L. and Pinus nigra Arn subsp. salzmannii (Dunal) Franco. Wood Sci Technol 51(5):1249–1258CrossRefGoogle Scholar
  14. Fan RE, Chen PH, Lin CJ (2005) Working set selection using second order information for training support vector machines. J Mach Learn Res 6:1889–1918Google Scholar
  15. Fini SH, Farzaneh M, Erchiqui F (2015) Study of the elastic behaviour of wood–plastic composites at cold temperatures using artificial neural networks. Wood Sci Technol 49(4):695–705CrossRefGoogle Scholar
  16. Fletcher R (2013) Practical methods of optimization. Wiley, Hoboken.  https://doi.org/10.1002/9781118723203 Google Scholar
  17. Forman G, Cohen I (2004) Learning from little: comparison of classifiers given little training. In: European conference on principles of data mining and knowledge discovery. Springer, Berlin, pp 161–172Google Scholar
  18. Friedman J, Hastie T, Tibshirani R (2001) The elements of statistical learning. Springer series in statistics, vol 1. Springer, New York, pp 241–249Google Scholar
  19. Fu Z, Avramidis S, Zhao J, Cai Y (2017) Artificial neural network modeling for predicting elastic strain of white birch disks during drying. Eur J Wood Prod 75(6):949–955CrossRefGoogle Scholar
  20. García-Iruela A, Fernández FG, Esteban LG, de Palacios P, Simón C, Arriaga F (2016) Comparison of modelling using regression techniques and an artificial neural network for obtaining the static modulus of elasticity of Pinus radiata D. Don. timber by ultrasound. Compos B Eng 96:112–118CrossRefGoogle Scholar
  21. González-Peña MM, Hale MD (2009a) Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 1: colour evolution and colour changes. Holzforschung 63(4):385–393Google Scholar
  22. González-Peña MM, Hale MD (2009b) Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 2: property predictions from colour changes. Holzforschung 63(4):394–401Google Scholar
  23. Hinterstoisser B, Schwanninger M, Stefke B, Stingl R, Patzelt M (2003) Surface analyses of chemically and thermally modified wood by FT-NIR. In: Acker VJ, Hill C (eds) The 1st European conference on wood modification. Proceeding of the first international conference of the European society for wood mechanics, pp. 15–20Google Scholar
  24. Nasir V, Nourian S, Avramidis S, Cool J (2018a) Stress wave evaluation by accelerometer and acoustic emission sensor for thermally modified wood classification using three types of neural networks. Eur J Wood Prod.  https://doi.org/10.1007/s00107-018-1373-1 Google Scholar
  25. Nasir V, Nourian S, Avramidis S, Cool J (2018b) Prediction of physical and mechanical properties of thermally modified wood based on color change evaluated by means of ‘group method of data handling’ (GMDH) neural network. Holzforschung.  https://doi.org/10.1515/hf-2018-0146 Google Scholar
  26. Nisgoski S, de Oliveira AA, de Muñiz GIB (2017) Artificial neural network and SIMCA classification in some wood discrimination based on near-infrared spectra. Wood Sci Technol 51(4):929–942CrossRefGoogle Scholar
  27. Ozsahin S, Murat M (2018) Prediction of equilibrium moisture content and specific gravity of heat treated wood by artificial neural networks. Eur J Wood Prod 76(2):563–572CrossRefGoogle Scholar
  28. Palatucci M, Mitchell TM (2007) Classification in very high dimensional problems with handfuls of examples. In: European conference on principles of data mining and knowledge discovery. Springer, Berlin, pp 212–223Google Scholar
  29. Pérez A, Larrañaga P, Inza I (2009) Bayesian classifiers based on kernel density estimation: flexible classifiers. Int J Approximate Reasoning 50(2):341–362CrossRefGoogle Scholar
  30. Platt JC (1999) Fast training of support vector machines using sequential minimal optimization. In: Scholkopf B, Burges CJC, Smola AJ, Press MIT (eds) Advances in kernel methods—support vector learning. MA, USA, Cambridge, pp 185–208Google Scholar
  31. Schnabel T, Zimmer B, Petutschnigg AJ, Schönberger S (2007) An approach to classify thermally modified hardwoods by color. For Prod J 57(9):105–110Google Scholar
  32. Schwanninger M, Hinterstoisser B, Gierlinger N, Wimmer R, Hanger J (2004) Application of Fourier transform near infrared spectroscopy (FT-NIR) to thermally modified wood. Holz Roh- Werkst 62(6):483–485CrossRefGoogle Scholar
  33. Willems W, Lykidis C, Altgen M, Clauder L (2015) Quality control methods for thermally modified wood. Holzforschung 69(7):875–884CrossRefGoogle Scholar
  34. Wu H, Avramidis S (2006) Prediction of timber kiln drying rates by neural networks. Drying Technol 24(12):1541–1545CrossRefGoogle Scholar
  35. Yang H, Cheng W, Han G (2015) Wood modification at high temperature and pressurized steam: a relational model of mechanical properties based on a neural network. BioResources 10(3):5758–5776Google Scholar
  36. Zhang H (2004) The optimality of naive Bayes. AA 1(2):3Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Wood ScienceThe University of British ColumbiaVancouverCanada

Personalised recommendations