A kind of living body detection method and device

Abstract

The present invention provides a living body detection method and device, the method comprising: respectively acquiring the left view image and the right view image of the object to be detected through different cameras; processing the left view image and the right view image to obtain Fusion of grayscale images; extracting texture features from the fusion grayscale images to obtain plane texture features; based on the camera parameters of the different cameras, the left view image and the right view image, using a binocular reconstruction algorithm to perform Three-dimensional reconstruction to obtain 3D face features; according to the plane texture features and the 3D face features, determine whether the object to be detected is a living object. The invention also discloses a corresponding living body detection device.

Description

technical field [0001] The present invention relates to the field of biological feature identification, in particular to a living body detection method and device. Background technique [0002] As the face detection system is widely used in various scenarios, the liveness detection function in face recognition has received more and more attention and attention. Common forms of attack include photos, video playback, and 3D models. Current liveness detection algorithms are mainly divided into four categories: motion-based methods, texture-based methods, 3D structure-based methods, and other feature-based methods. [0003] Motion-based methods mainly detect human responses to commands, or human physiological activities such as eye blinks, head bobbing, and mouth movements. These methods mainly use the information between frames and can usually achieve better results than other methods. However, this method often takes a while to collect pictures, and may require the cooperati...

AI Technical Summary

Problems solved by technology

For example, using a binocular system to analyze whether it is an attacking face by calculating the percentage of coplanar points, this method is ineffective for attacks on folded photos

Or based on a traditional monocular camera, the 3D depth information of the face is obtained through a video sequence for liveness discrimination. This method has good results on different devices, but it takes a long time for the depth recovery step.

At the same time, many methods using 3D structures often ignore other features of the input image, such as texture information.

Images