基于模糊的特征提取算法在人脸识别中的应用
Based on fuzzy characteristic extraction algorithm in human face recognition application
Fisherface是人脸特征提取中常用的方法,KPCA+LDA能更好地解决非线性问题。本文把模糊技术与KPCA+LDA相结合提出了一种新的特征抽取方法。首先用KPCA进行初次特征提取,然后利用FKNN计算图像对各类别的隶属程度,再在此基础上用LDA进行二次特征提取。在ORL人脸库上的实验结果表明了该方法的有效性。
Fisherface is in the human face characteristic extraction the commonly used method, KPCA+LDA can solve the non-linear problem well.This article unified the fuzzy technology and KPCA+LDA proposed one new characteristic extraction method.First uses KPCA to carry on the primary characteristic extraction, then uses the FKNN computation image to various categories subordination degree, again carries on two characteristic extractions in this foundation with LDA.Has indicated this method validity in the ORL person face storehouse experimental result.
李铁
计算技术、计算机技术
人脸识别fisherface特征提取模糊最近邻分类
Human face recognitionfisherfaceCharacteristic extractionThe fuzzy most close neighbor classifies
李铁.基于模糊的特征提取算法在人脸识别中的应用[EB/OL].(2011-12-19)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201112-489.点此复制
评论