Human face living body detection method based on 3D point cloud geometrical characteristics

Abstract

The invention relates to a human face living body detection method based on 3D point cloud geometrical characteristics. The method includes: performing calculating to obtain average face point cloud according to the 3D point cloud data of the plurality of real persons, calculating FPFH features obtained by taking the left eye, the right eye, the nose tip, the left mouth corner and the right mouthcorner of the average face of the real persons as centers according to the average face point cloud, and connecting the FPFH features in series to obtain FPFH total features of the average face; calculating FPFH features of five key points of a left eye, a right eye, a nose tip, a left mouth corner and a right mouth corner of the test face, and connecting the FPFH features in series to obtain FPFHtotal features of the test face; and calculating the Euclidean distance between the FPFH total features of the test face and the total features of the average face, if the distance is greater than athreshold, determining that the face is a real person, and otherwise, determining that the face is an attack. According to the method, a user does not need to carry out complex cooperation instructions, the flexibility is good, paper printing attacks and video replay attacks can be easily defended, and even bent and wrinkled photos can be defended.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a human face living body detection method based on 3D point cloud geometric features. Background technique [0002] At present, with the advancement and development of image processing and computer vision technology, face recognition is more and more widely used in daily life. Face recognition security is particularly important. Most of the current face detection methods are based on 2D images, by extracting the features of 2D texture images, using machine learning or deep learning methods. These methods are greatly affected by the lighting scene, posture, expression, etc. Under different environments and scenes, the detection effect is not stable enough. Using 3D face point cloud can reduce the influence of lighting and posture factors and improve the detection accuracy. Some existing camera equipment, such as RealSense SR300, etc., can easily obtain 3D face point clo...

AI Technical Summary

Problems solved by technology

The existing face anti-counterfeiting method based on 3D point cloud directly uses 3D point cloud coordinates, does not fully mine the geometric information of the point cloud, only considers the point cloud coordinate point information, and does not deeply extract the overall characteristics of the 3D face point cloud , the feature description is very single, and it is difficult to defend against attacks on curved or wrinkled printed photos

Images