Abstract: Modular 2DPCA method based on within-class weight average face recognition is presented.Firstly， the within-class weight average of each sub-image of all training samples in each class is calculated，and it is used to normalize corresponding sub-images of the within-class samples.After that，the optimal projecting matrix from general matrix that is made up of all normalized sub-images can be obtained accordingly.Secondly，when all sub-images of training samples and testing samples are projected to the optimal projecting matrix that has been gotten above，the recognition features are produced.Finally，the nearest distance classification is used to distinguish each face.The experiment results on ORL face database indicate that the recognition performance of the proposed method is obviously superior to that of 2DPCA and general modular 2DPCA.

Key words: face recognition, two-dimensional principal component analysis, within-class weight average

摘要： 提出了一种基于类内自适应加权平均值的模块2DPCA人脸识别方法。该算法对每一类训练样本中每个训练样本的每一子块求类内自适应加权平均值，并用类内自适应加权平均值对训练样本类内的相应子块进行规范化处理，然后由所有规范化后的子块构成总体散布矩阵，从而得到最优投影矩阵；由训练集的全体子块的加权平均值对训练样本的子块和测试样本的子块进行规范化后投影到最优投影矩阵，得到识别特征；最后用最近距离分类器分类。在ORL人脸库上的实验结果表明，提出的方法在识别性能上明显优于2DPCA方法和普通模块2DPCA方法。

关键词: 人脸识别, 二维主成分分析, 类内自适应加权平均值