移动设备网络安全下人脸终端身份识别仿真

摘要

An identity recognition method of facial terminal puts forward based on facial detection recognition.The Adaboost is used to improve multi-formation facial detection method and local feature is searched based on skin color to determine facial direction.The segmentation of candidate regions is facilitated.Then,allocation guide of memory resources for stream data of facial image in facial terminal equipment is realized based on quality of identification information difference and the nearest neighbor classification method,and facial recognition ability is improved.Moreover,the facial optimal solution is marked and sample set is selected to put it in memory of mobile device.Thus,the identity recognition of facial terminal under network security of mobile device is realized.Experimental result shows that the proposed authentication method can realize the identity recognition of mobile device.%移动设备网络安全下人脸终端身份识别,能够高效保证用户身份信息安全.人脸终端身份的认证,需要对人脸终端设备中内存资源进行分配引导,对人脸最优解集进行标注,完成身份识别.传统方法采集人脸特征图像并进行检测和跟踪,再匹配后台数据库人脸信息获得识别结果,但忽略了对人脸最优解集进行标注,导致识别精度偏低.提出基于人脸检测识别的人脸终端身份识别方法.利用Adaboost改进多形态人脸检测方法,根据肤色特征搜索出局部特征确定人脸方向,便于候选区域分割;基于识别信息差异性质量及最近邻分类法,对人脸终端设备中的人脸图像流数据实现内存资源分配引导,以提高脸部识别能力,对人脸最优解集进行标注并且选择出样本集置于移动设备内存中,实现移动设备网络安全的人脸终端身份识别.实验结果表明,利用提出认证方法能够良好实现移动设备身份识别.