【摘要】 两个方向的费希尔主成分分析方法(2D)2FPCA结合了二维主成分分析方法(2DPCA)和二维费希尔线性方法(2DFLD)的特点,很好地解决2DPCA特征提取时比传统PCA需要更多系数来表达图像信息的问题。根据木材体视图受光照影响及同一树种的样本图片之间差别较大等特点,适当增加了识别的类内散布矩阵从而提高了木材的识别率。(2D)2FPCA为木材的智能识别提供了一条新途径。更多还原

【Abstract】 Two-oriental Fisher Principal Component Analysis (2D)2FPCA combines the features of both Two-dimensional Principal Component Analysis (2DPCA)and Two-dimensional Fisher Linear Discriminant (2DFLD). It can resolve the problem which needs more coefficients to express image information using 2DPCA,compared with traditional PCA. According to the features that timber body views are affected by light and the large differences between the sample pictures of the same species of timber,within-class scatter matrix intended for recognition is added appropriately in order to increase timber recognition rate.(2D)2FPCA provides a newmethod for intelligent recognition of timber.更多还原