基于Gabor小波和支持向量机的人脸识别

Face recognition based on Gabor wavelet and support vector machine

提出一种将Gabor小波和支持向量机相结合的人脸识别算法。运用AdaBoost算法在复杂背景图像中快速准确地检测出人脸部分,进而用Gabor小波提取归一化人脸图像的特征。最后采用支持向量机进行人脸的分类识别。在ORL人脸库和CAS-PEAL-R1人脸库中对算法进行了测试,结果表明该算法识别率较高。

A face recognition algorithm is proposed based on Gabor wavelet and support vector machines. Using the Ada- Boost algorithm, the portion where human face located is detected in an image under complicated background, which is then normalized. The features are extracted from the normalized human face image with Gabor wavelet, and the human face is classified and identified by SVM. Tests on the ORL and CAS-PEAL-RI databases validate the efficiency of the algorithm.