特征提取是人脸识别技术中一个基本而

又十分重要的环节，寻找有效的特征是解决识别问题的关键本文提出了一种新的人脸特征提取方法。该方法通过可调因子有效结合人脸局部流形结构信息和样本

的类别信息，充分提取样本的判别信息，将LPP 发展成为有监督方法。实验结果表明，该方法具有较好的

识别效果。-Feature Extraction in Face Recognition Technology is a fundamental and very important to find the characteristics of an effective solution to identify the crux of the problem is proposed in this paper a new facial feature extraction methods. This method is through the effective integration of tunable factor partial face manifold structure of information and samples of the types of information, the full extraction of the sample discriminant information, will be developed into a supervised LPP method. Experimental results show that the method has better recognition effect.