Classification of wood surfaces according to visual appearance by multivariate analysis of wood feature data

Abstract

Its natural aesthetics make wood an attractive material for construction and design. However, there is no detailed understanding of the relationships between human perception of the appearance and measurable features of wood surfaces that could be used for controlling sawn timber production. This study investigated whether wood surfaces can be classified according to their visual appearance on the basis of wood feature measurements. Cluster analysis was used to discover a classification based on a set of feature pattern variables in a sample of 300 softwood floorboards. A finely graded visual appearance sorting provided a reference. Discriminant analysis was applied to identify the relevant variables from the tested set and to assess predictability of the classification. The results indicated that visual appearance sorting could be approximated quite well by the variable-based classification after pregrouping according to board position in the log. Ambivalent results were obtained for group prediction within the validation sample. While for boards from some groups prediction was mostly or entirely correct, boards from other groups were largely misclassified. An effect of the available sample was one of the surmised causes, making repetition of the analysis based on a larger sample a desirable focus of further research.

How do consumers express their appreciation of wood surfaces? Norway spruce floors in Germany as an example

“…Only pith as a unique feature is found separated into Cl4 and with a relative low appreciation. However, a conversion of quantitatively measurable technical surface parameters into 'appearance classes' based on cluster analysis for the same material is presented in Breinig et al (2015).…”

“…However, the visual surface characteristics of the entire floor resulting from a specific composition of the single boards were not systematically assessed and quantitatively measured, as e.g. Breinig et al (2015) did when classifying Fig. 3 Average cluster ratings of the tested items retrieved after cluster prediction.…”

“…The technically accepted floor boards were sorted aiming homogeneous groups of visually similar floor boards which led to 15 distinctively different visual classes (VC). The same material and VCs were tested in Breinig et al (2015) to establish a technical board classification based on measured variables, mainly quantitative descriptors of different knot attributes (e.g. size, state) and distribution (e.g.…”

Optimization potential for perception-oriented appearance classification by simulated sawing of computed tomography-scanned logs of Norway spruce

“…All boards were sorted by iteratively grouping them based on distinct visual appearance, aiming to maximize within-group homogeneity (see Breinig et al 2014). The visual impression was mostly governed by the appearance of knots, orientation of the annual rings on the board face and colour variations.…”

“…Apart from details on the board sorting procedure, a description of data processing, feature measurement and the explorative classification approach described can be found in the study by Breinig et al (2014). An overview of the elements of the analysis is provided in Figure 3.…”