Human face living detection method based on light field camera

Abstract

The invention discloses a human face living detection method based on a light field camera. The method comprises steps: (1) original data of the light field camera are extracted and decoded; (2) a refocus image for different depths of the face is extracted from the light field data; (3) focus features and power spectrum features are extracted in the refocus image, and the two kinds of features are connected as integrated features for the human face image; (4) the above integrated features are extracted on all acquired light field images, the integrated features of a training set are inputted to a support vector machine for training to obtain a model for human face living detection, and the integrated features of the to-be-detected image are inputted to the well-trained model to obtain a detection result. The characteristic that the refocus image for different depths can be acquired after one-time photographing by the light field camera is used, human face refocus change features are extracted for human face living detection, and compared with the traditional human face living detection method, the detection result is accurate, the detection efficiency is higher, the detection is simpler and more convenient, and requirements of practical applications are met.

Description

technical field [0001] The invention relates to the research field of face detection, in particular to a face detection method based on a light field camera. Background technique [0002] The purpose of the live face detection technology is to distinguish the authenticity of the collected faces, to ensure that the face information used for authentication is obtained from live users, so as to ensure the safe operation of the face recognition system. With the application and popularization of face recognition systems, face liveness detection technology is becoming more and more important. [0003] There are still some performance and accuracy deficiencies in the traditional face liveness detection technology in practical applications. The traditional face liveness detection technology based on image texture frequency and other information is easily affected by light and photo quality. Face liveness detection technology based on motion information such as blinking, head movem...

AI Technical Summary

Problems solved by technology

The traditional face detection technology based on image texture frequency and other information is easily affected by light and photo quality

Face liveness detection technology based on motion information such as blinking, head movement, mouth movement, etc. This type of technology requires real-time action response cooperation of living users, which is not user-friendly, or needs to be based on video, and has strong data dependence. , the verification time required is relatively long

However, the face detection technology based on living body feature information, such as using thermal infrared imaging, will be affected by outdoor temperature, occluded objects, etc., and the equipment is expensive and expensive.

Images