基于PCA算法的人脸检测研究

Face Detection Research Based On The PCA

本文主要利用主成分分析(PCA)方法提取人脸特征,将原来的自变量变换到另外的一个空间中,即特征子空间,然后选择其中一部分重要成分作为自变量(此时丢弃了一部分不重要的自变量),最后利用最小二乘方法对选取主成分后的模型参数进行估计。通过低维子空间表示高维数据,有效的对数据进行了压缩,识别起来简单有效。

In this paper, using principal component analysis （PCA） method to extract face feature, the original independent variables transform to another space, namely feature subspace, and then choose some of the important components as the independent variable （the rejected part not important independent variable）, finally using the least squares method of selecting principal components estimate model parameters.Through the low-dimensional subspace said high-dimensional data, effective for data compression, simple and effective identification.