摘要
提出了一种2D-Gabor滤波和主成分分析相结合的新的人脸识别算法。首先将人脸库的所有样本进行图像预处理,并将由2D-Gabor滤波后产生的图像作为独立的样本加入到样本库中,从而减少了小样本问题对人脸识别效率的影响,再结合经典的主成分分析方法进行人脸识别。试验证明,与单独的主成分分析法相比,该方法可以有效的提高识别率。
A new face recognition algorithm was proposed based on two dimension Gabor filter and principal component analysis.All of the image data were preprocessed to decrease the influence of environment,to solve the small sample problem which affected the rate of face recognition,all the images filtered by 2D-Gabor algorithm were put into the face data as independent images.Face recognition was achieved by using the standard principal component analysis.Results show that compared with the single principal component analysis method,the algorithm proposed in this paper can effectively enhance the recognition rate.
出处
《长江大学学报(自科版)(上旬)》
CAS
2009年第01X期216-218,393,共3页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI & ENG
基金
广东省自然科学基金项目(07010869)
国家重点实验室开放课题基金项目(0505)
国家重点实验室开放课题(A0703).