摘要
针对人脸结构过于复杂,导致识别精度低的问题,提出基于支持向量机(Support Vector Machine,SVM)的人脸识别方法。首先,选取Gabor小波变换算法,利用高斯函数表示短时傅里叶变换的窗函数,通过对人脸图像的卷积运算,提取人脸图像的幅值特征以及相位特征。其次,选取主成分分析方法,对所提取的人脸图像特征进行特征降维处理。最后,设置完成降维处理的人脸特征值,作为SVM的输入,利用SVM输出人脸识别结果。实验结果表明,该方法在强光等复杂环境下,仍然可以精准识别人脸,人脸识别精度高于97%。
Aiming at the problem that the face structure is too complex,resulting in low recognition accuracy,a face recognition method based on Support Vector Machine(SVM)is proposed.Firstly,the Gabor wavelet transform algorithm is selected,and the Gaussian function is used to represent the window function of the short-time Fourier transform.Through the convolution operation of the face image,the amplitude feature and phase feature of the face image are extracted.Secondly,the principal component analysis method is selected to reduce the dimension of the extracted face image features.Finally,set the face feature value after dimension reduction as the input of SVM,and use SVM to output the face recognition results.The experimental results show that the method can still accurately recognize faces in strong light and other complex environments,and the accuracy of face recognition is higher than 97%.
作者
张家彬
孟建军
ZHANG Jiabin;MENG Jianjun(Lanzhou Jiaotong University,Lanzhou Gansu 730000,China)
出处
《信息与电脑》
2023年第4期205-207,共3页
Information & Computer
关键词
支持向量机(SVM)
主成分分析
人脸识别
Support Vector Machine(SVM)
principal component analysis
face recognition
