主分量分析(PCA ) 是统计学中分析数据的一种有效的方法, 可以将数据从高维数据空间变换到低维特征空间, 因而

可以用于数据的特征提取及压缩等方面。在该文的形状识别系统中, 用PCA 法提取图像的形状特征, 能够较好地满足识别

层的输入要求。在识别层研究了3 种识别方法: 最近邻法则、BP 网络及协同神经网络方法, 均取得了满意的实验效果。-Principal component analysis (PCA) is a statistical analysis of data in an effective method to data from high dimensional data space transformation to the low-dimensional feature space, which can be used for data feature extraction and compression and so on. In this paper, the shape recognition system using PCA extraction of the shape of image features, can be used to satisfy the identification requirements of the input layer. In the recognition layer studied three kinds of identification methods: nearest neighbor rule, BP network and the synergetic neural network methods, have achieved a satisfactory experiment results.