Human face living body detection method, system, device and readable storage medium

Abstract

The invention discloses a human face living body detection method, which comprises the steps of obtaining coordinates of a human face frame and human face key points through human face detection, andperforming human face alignment through the coordinates of the human face key points; associating the same face IDs in the continuous video frames by using a face tracking technology; intercepting a human face from the original image, converting an RGB channel into an HSV and YCbCr space, and superposing the converted HSV and YCbCr images to obtain a superposed image; extracting Sobel features from the face region through a Sobel operator, and mapping the obtained Sobel features to an obtained Sobel feature map; respectively inputting the Sobel feature map and the overlay map of a preset framenumber of a human face ID from two input channels of a double-current neural network to obtain a living body judgment result of each frame of picture; and voting all living body judgment results of the human face ID, determining that the object is a living body when the number of frames of the living body is judged to be large, and determining that the object is a non-living body when the numberof frames of the attack is judged to be large. The method has the advantages of being high in safety coefficient, good in user experience and high in robustness.

Description

technical field [0001] The present invention relates to the technical field of image recognition, in particular to a face detection method, system, device and readable storage medium. Background technique [0002] Face recognition is a biometric technology for identification based on human facial feature information. This technology is being more and more widely used in various industries such as security, finance, and education. The current mainstream face recognition technology can only distinguish the difference between different faces, but cannot tell whether the user himself is using face recognition. When the user's face information is leaked, criminals can use the user's face information to make photos, videos, and 3D face models to deceive the face recognition system, resulting in the loss of the user's property and information. Determine whether the carrier of the face to be recognized is a real person, or a non-living attack method such as a photo, video, or mask....

AI Technical Summary

Problems solved by technology

Among them, the texture feature is related to the ambient light and camera type, and the moiré is related to the resolution of the camera, which leads to poor robustness of this type of feature and cannot cope with the live detection of multiple types of cameras in multiple environments; single frame based on deep learning The face depth estimation algorithm for face images only considers the features of the face area, and is also strongly related to the texture of the image, and the robustness of the features is still not high; the face depth estimation of multi-frame face images based on deep learning The algorithm combines the information in multiple frames of images, which improves the robustness of depth estimation to a certain extent, but still only considers the features of the face area, and does not introduce more robust features.

Images