LinearDiscriminantAnalysis

代码说明：

  提出了一种新的基于图像分块重构和线性判别分析相融合的方法，主要用于人脸识别。该方法通过计算两幅图像之间图 像块的重构均值误差，运用线性判别分析求出两幅图像降维后的欧式距离，融合重构误差和欧式距离计算这两幅图像之间的差别 程度。(A new block-based image reconstruction and the integration of linear discriminant analysis method is mainly used for face recognition. This method between the two images by calculating the reconstructed image block average error, calculated using the linear discriminant analysis after dimension reduction of two images Euclidean distance, the integration calculation of reconstruction error and the Euclidean distance difference between these two images level.)