中文摘要：

提出尺度不变特征转换（scale invariant feature transform,SIFT）和局部相位量化（local phase quantization,LPQ）相结合的人脸图像识别算法.利用SIFT算法检测两幅人脸图像特征点,得到两幅图像的特征点集合;计算特征点邻域的LPQ特征,并将特征点邻域的集合表述为LPQ直方图序列,作为人脸图像的描述算法;采用特征点邻域距离比值进行匹配判定.实验表明在ORL标准人脸数据库上,该方法具有较高的识别率.

英文摘要：

A face recognition method is presented based on scale invariant feature transform（SIFT） and local phase quantization（LPQ） in this paper.Two sets of feature points are detected from the two face images for matching by applying the SIFT algorithm.The neighbor feature point of LPQ is calculated,and its set is expressed by the LPQ Histogram Sequence which is used as the descriptor of face image.The matching pairs between the two sets of feature points are determined by using the neighbor feature points distance ratio based matching judgment.The experimental results show that the proposed method obtains better recognition rate on ORL face database comparing to other classical methods.