特征提取是模式识别研究领域的一个热点.本文提出了一种基于独立成分分析和线性鉴别分析的特征提取方法.谊方法中引入了零空间的概念,指出了前人算法中的不足之,并且给出了一个完整的独立成分分析和线性鉴别分析的组合算法.在ORL和Yale人脸数据库上的实验表明了该方法的有效性.-Feature extraction is a hot field of pattern recognition research. In this paper, which is based on independent component analysis and linear discriminant analysis feature extraction method. Yi method introduced the concept of zero-space, pointing out that the algorithm predecessors shortcomings, and gives a complete independent component analysis and linear discriminant analysis of the combined algorithm. In the ORL and Yale face database, the experiment shows the effectiveness of this method.