基于改进的LLE和FSVM方法在人脸识别中的应用

Based on Improved LLE and FSVM Method in Face Recognition of Application

针对人脸识别问题,提出了一种新的算法。该算法首先用gabor小波对人脸图像进行特征提取。然后采用LLE算法进行降维。最后用FSVM和三叉决策树相结合设计识别分类器进行人脸识别。在降维的过程中,针对高维空间相似性度量函数和自适应参数选取方法上,对LLE算法进行了改进。在ORL人脸数据库的仿真结果表明,该算法能有效提高人脸识别性能,具有较高识别率。

In order to improve the accuracy of face recognition, a novel algorithm is presented. First, the Ga- bor wavelet is used to extract the feature. Secondly, the low -dimensional features from the face character image da- ta was extracted by LLE algorithm with adaptive parameter estimation. In the LLE algorithm, a new function is presented to measure the proximity of objects in high dimensional spaces. At last, trained a classifier system for classification, which was designed combine the fuzzy support vector machine and the triple decision tree. The ex- perimental results on ORL face database show that the proposed algorithm performance effectively.