ar_dlda

  在ar人脸库上实现Direct Fisher discriminant analysis,该方法首先对类内散布矩阵对角化，然后对类间散布矩阵对角化，使类间散布矩阵对角化的矩阵即使鉴别向量集(In the ar face database to achieve Direct Fisher discriminant analysis, this method first within-class scatter matrix diagonalization, and then between-class scatter matrix diagonalization, so that between-class scatter matrix diagonalization of the matrix even if the identification of a vector set)