The Fractal Evaluation of Wood Texture by the Triangular Prism Surface Area Method

Textures of the surfaces of fifteen wood species were characterized by fractal dimension of the triangular prism surface area method. Fractal dimension ranged from 2 to 3, and sharp lightness variation caused high fractal dimension, whereas low values related to smooth variation. Based on this index, wood specimens were generally divided into a hardwood group with value greater than 2.50 and a softwood group with value less than 2.50. Six types of fractal dimension distribution were explored in our experiments, including plane, inclined plane, concave, convex, zigzag distribution, and hilly distribution. From these both the features of local textures and the general variation tendency of the whole surface could be illustrated. It was strongly proposed that fractal dimension should be adopted to quantitatively evaluate wood texture with coarseness and evenness, because such variation was related to the number of grains, surface orientation, and location. For wood color matching, fractal dimension is of great importance in ensuring texture matching to achieve a constructed surface texture close to the features of natural variation. These distribution patterns were considered as a good reference previous to matching. Little variation of fractal dimension along the grain was observed, and this could be used to simplify texture matching by a very limited number of the indices. No significant relationship was found between fractal dimension and lightness, further implying that fractal dimension was independent of lightness