Face-mask recognition for fraud prevention using Gaussian mixture model

With the rapid development of biometric identification technology, face recognition has been one of the most widely used as its important component. It facilitates a series of applications such as security, military, transportation, education and other fields. The demand for face feature recognition is increasing. However the current techniques still exist some deficiencies. In this paper, we proposed a face-mask recognition method for fraud prevention based on Gaussian Mixture Model. We address the problem of identifying the face and mask in the area of financial security precaution. And we show how to combine opencv with dlib to recognition face and extract it. We use Gaussian Mixture Model (GMM) to construct the model of human faces. According to this, we calculate the similarity between the face sample and the model. By analyzing and learning the features of faces, we can predict whether the image of which we test is a human face or a mask. Compared with other traditional method of face recognition, our approach has been targeted to strengthen the ability to recognize abnormal faces such as sunglasses, masks and respirator, and reduce the potential danger of these unusual faces in the security field. It is simple to be calculated and has a higher accuracy. In addition, our method have enhanced the robustness of the algorithm about mask recognition.