关键词: 木材表面纹理, 定量分析, 空间灰度共生矩阵, 纹理特征, 因子分析

Abstract:

This paper introduces spatial gray level cooccurrence matrix (GLCM) into wood surface texture quantitative analysis. On the basis of fifty species datas, it concludes that d equals to 3, and θ equals to 0° in radial texture, equals to the average of 0°, 45° and 135° in tangential texture are suitable for wood texture analysis. It induces four principal components from eleven GLCM features, and explains their meanings; then it analyzes distribution law of wood textures in principal component space, and probes wood texture characteristics both in radial and tangential directions, as well as their changing law; at last, it advances a calculation formula for computating wood texture value comprehensively, and how to use it to judge the similarity of two textures.

Key words: Wood surface texture, Quantitative analysis, Spatial gray level cooccurrence matrix, Texture features, Factor analysis