基于空间灰度共生矩阵的木材纹理特征提取

The Extraction of Wood Texture Features Based on Spatial GLCM

讨论了生成步长、生成方向、图像灰度级对灰度共生矩阵特征参数的影响,在对国内50个树种计算分析的基础上,得出适于描述木材纹理的灰度共生矩阵构造方法。结合了灰度共生矩阵特征参数间的相关性及木材纹理自身的特征提取出了一套表征木材纹理的特征参数。

The influence of building step, building direction and image grey level on feature parameters of GLCM was discussed, and the building method that was appropriate to describe wood texture was acquired on the basis of calculating and analyzing fifty species of trees. Combining the correlation of feature parameters of GLCM with the features of wood texture, a set of feature parameters that can describe wood texture was extracted.