使用部件信息改进弹性匹配人脸识别

Human Face Recognition Using Information of Face Components to Improve Elastic Matching

摘要:

提出了一种综合利用部件信息和整体信息的人脸识别算法。利用部件定位得出人脸关键点（例如暗孔和鼻尖）的位置，然后利用这些关键点信息来限制弹性匹配的搜索范围。实验结果表明，这种方法在很大地提高了识别速度的同时能使识别率保持不变甚至还有所提高。

Abstract:

We present a new face recognition algorithm using both component information and global information from human faces.We first obtained the positions of key points,such as eye iris and nose tip,in human face by locating face components,then used these positions to limit the search area of elastic matching.Experimental results show that higher recognition speed is obtained while the recognition rate remains undercreased even increased.