AN IMPROVED FACE RECOGNITION METHOD BASED ON GABOR WAVELET TRANSFORM AND SVM

摘要：

This paper proposed an improved face recognition method based on Gabor wavelet transform and SVM. Firstly, the approach utilizes discrete Gabor wavelet transform to face images. Then the dimensions of Gabor feature vectors are reduced and transform feature base can be got by using the improved Fisherface method proposed in this paper. By projecting the face image onto the base, the transform coefficient used to be the input of classifier can be got. Finally face image can be classified using a SVM with error correction. Experiment results show that the method of this paper is robust and effective.

关键词： face recognition Gabor wavelet transform SVM

作者: Zaiying Liu Lixiao Zhang Linling Zhu

作者单位: Institute of Information Science Shanghai Sanda University Shanghai, China School of Electronics and Information Tongji University Shanghai, China

会议类型: 国际会议

会议名称: 2012 Fifth International Symposium on Computational Intelligence and Design 第五届计算智能与设计国际会议 ISCID 2012

会议地点: 杭州

会议语种:英文

页码: 378-381

在线出版日期: 2012-10-28（万方平台首次上网日期，不代表论文的发表时间）