基于局部相位量化与多通道颜色的鲁棒人脸识别  

Local phase quantization and multi-color channel based robust face recognition

作　　者：张妍琰[1] 段娜[1] 

机构地区：[1]河南城建学院计算机科学与工程学院,河南平顶山467036

出　　处：《河南城建学院学报》2014年第1期67-72,共6页Journal of Henan University of Urban Construction

基　　金：河南省教育厅科学技术研究重点项目(12A520006)

摘　　要：针对非可控条件下的人脸识别问题,提出了一种新的融合多颜色通道的局部相位量化方法。首先,在不同颜色空间的各颜色通道图像上计算局部相位量化响应图,然后对每幅响应图分块计算统计直方图,最后将特定颜色通道的相似度进行融合。该方法充分利用了图像不同颜色通道间的互补特性。实验结果表明,该方法实用性好,鲁棒性强,在大规模非可控人脸数据库上取得了较好的结果。In order to solve the problems that exist in uncontrolled condition face recognition, a variational method based on fusion of local phase quantization information with muhi-color channel was proposed. First, compute the local phase quantization image on every color channel of different color spaces. Then,extract the statistic histograms on the local phase quantization images. And finally, fuse the similarity scores from different color channels. The proposed method utilizes the complementarity of different color channels. The experimen- tal results show that the proposed method has effectiveness and better robustness, and it achieves state-of-the- art performance on large scale uncontrolled condition face database （FRGC）.

关 键 词：人脸识别 局部相位量化 颜色空间 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]