主分量分析是一种线性特征抽取方法,被广泛地应用在人脸等图像识别领域。但传统的PCA都以总体散布矩阵作为产生矩阵,并且要将作为图像的矩阵转换为列向量进行计算。该文给出了一种利用图像矩阵直接计算的二维PCA,以类间散布矩阵的本征向量作为投影方向,取得了比利用总体散布矩阵更好的识别效果,并且特征抽取速度更快。在ORL和NUSTFDHⅡ标准人脸库上的实验验证了该方法的有效性。
Principal component analysis (PCA) is an important method widely used in images data compression and feature extraction. But conventional PCA usually uses total scatter matrix as a generation matrix, and two-dimension (2D) image matrices must be transformed into vectors. This paper gives a 2D-PCA, which uses original image matrices to compute between-class covariance matrix and its eigenvectors are derived for images feature extraction. The experiments on ORL and NUSTFDB Ⅱ face-databases indicate that the recognition rates are higher than PCA and 2D-PCA using total scatter matrix, and the speed of feature extraction is faster.