Research on the Improved Kernel Discriminant Analysis Based on Face Recognition Algorithm
|更新时间:2026-07-10
|
Research on the Improved Kernel Discriminant Analysis Based on Face Recognition Algorithm
Vol. 15, Issue 3, Pages: 19-22(2010)
作者机构:
黑龙江省政法管理干部学院
作者简介:
基金信息:
DOI:
CLC:TP391.41
Published:2010
稿件说明:
移动端阅览
ZHAO Ying. Research on the Improved Kernel Discriminant Analysis Based on Face Recognition Algorithm[J]. 2010, 15(3): 19-22.
DOI:
ZHAO Ying. Research on the Improved Kernel Discriminant Analysis Based on Face Recognition Algorithm[J]. 2010, 15(3): 19-22.DOI:
Research on the Improved Kernel Discriminant Analysis Based on Face Recognition Algorithm
摘要
由于目前面临光照、表情、姿态等影响
人脸识别是计算机视觉领域中的一个难题.由于人脸图像蕴含着丰富的纹理信息
充分利用纹理特征也是提高人脸识别算法的有效方法.利用图像处理中信号处理和学习两种方法的优势
提出了基于Gabor小波和KDCV相结合的图像特征提取算法.首先利用小波分析提取人脸图像的纹理特征
然后用改进的核判别分析方法对提取的纹理特征进行降维
用降维后的数据作为人脸识别的特征.通过仿真实验表明
改进的核判别分析算法可有效提高人脸识别正确率
提高算法的实际应用性.
Abstract
Since face recognition faces the influences of illumination
facial expressions and poses
it has become a difficult problem in the computer vision area.Since the face images have the great texture information
it is an effective method to enhance the recognition performance through applying the texture features.Using the advantages of signal processing and learning in image processing
this paper proposes the algorithm based on Gabor wavelet and KDCV combined image feature extraction.Firstly