摘要
为快速准确进行人脸识别,设计一种结合FFT和SVD的人脸识别方法.该方法先对预处理过的人脸图像进行FFT,用振幅谱来表示复杂的二维图像,再用SVD对振幅谱进行降维,最后进行模板的匹配识别.仿真实验表明该方法对光照和表情具有一定鲁棒性,识别效果比较好,但对人脸面部图像的完整性要求较高.
A face recognition method combining FFT and SVD is designed for quick and accurate face recog.nition.The method is FFT for preprocessed face images,complex two-dimensional images are rep.resented by amplitude spectrum,then the amplitude spectrum is reduced by SVD,and then the template matching is identified.The simulation results show that the method is robust to light and expression,and the recognition effect is good.However,the integrity of facial images is required.
作者
鲍庭瑞
BAO Ting-rui(Department of Electrical Engineering,Anhui Industry Polytechnic,Tongling,Anhui 244000,China)
出处
《通化师范学院学报》
2019年第8期71-73,共3页
Journal of Tonghua Normal University
