研究人脸遮挡识别技术在智能备件柜身份安全认证中的应用,以提高智能备件柜的安全性。依据人脸肤色以及形状特征定位人脸位置并对图像进行灰度归一化处理,采用稀疏编码方法将遮挡人脸图像的局部特征转变为特征向量,运用支持向量机构建遮挡人脸图像分类器,根据输出结果判断实时采集的人脸与数据库人脸图像是否匹配,实现智能备件柜的身份安全认证。实验结果表明,该方法能够准确识别遮挡人脸完成身份安全认证,人脸识别的误识率低,识别速度快,可保障智能备件柜安全。

In the paper,the application of occlusion face recognition technology in the identity security authentication of the intelligent spare parts cabinet will be studied to improve the security of the intelligent spare parts cabinet.Locate the face position according to the skin color and shape features of the face and normalize the gray level of the image.Use the sparse coding method to convert the local features of the occluded face image into feature vectors.Use the support vector mechanism to build the occluded face image classifier.Judge whether the real-time collected face matches the face image in the database according to the output results,so as to achieve the identity security authentication of the intelligent spare parts cabinet.The experiment results show that this method can accurately identify the occluded face to complete the identity security authentication.The face recognition has low error rate and fast recognition speed,which can ensure the security of the intelligent spare parts cabinet.