融合SIFT的多尺度分析人脸识别算法研究

Research on Face Recognition Algorithm with Multi-Scale Analysis Fusing SIFT

由于人脸识别过程中会受到光照、尺度等因素的影响,采用传统的SIFT算法进行人脸识别时会出现匹配效果较差、无法匹配或是错误匹配的情况,因此提出了一种融合SIFT算法的多尺度分析的人脸识别算法。首先在粗尺度上,采用形态学边缘检测算法对原始的人脸库进行轮廓特征提取,形成新的人脸库;然后在细尺度上,采用SIFT算法对新的人脸库进行人脸识别与匹配。采用ORL人脸库对改进后的算法进行验证,实验结果表明改进后的SIFT算法较好地解决了传统SIFT算法的不足,计算量大大减少,人脸图像的识别效率与匹配效率得到了有效提升。

Due to the influence of illumination, scale and other factors in the process of face recognition, poor matching, mismatch or error matching would appear when using the traditional SIFT algorithm for face recognition,a face recognition algorithm based on multi-scale analysis fusing SIFT algorithm was proposed.Firstly,the morphological edge detection algorithm was used to extract the contour feature of the original face database,and new face database was formed; then,the SIFT algorithm was used to face recognition and matching in the new face database.The ORL face database was used to verify the improved algorithm, the experimental results showed that the improved SIFT algorithm could solve the shortcomings of the traditional SIFT algorithm, the calculated amount was greatly reduced, and the recognition efficiency and matching efficiency of face image were improved effectively.