Face snapshot method, computer storage medium and electronic equipment

Abstract

The invention provides a face snapshot method, a computer storage medium and electronic equipment. The method comprises the following steps: S1, acquiring an input video image; s2, judging whether aninput video image is an initial frame or not, if the input video image is the initial frame, performing face detection on the video image to realize face frame positioning, and if the input video image is not the initial frame, performing face tracking on the video image to realize face frame positioning; and S3, key point detection is performed on the face obtained through face frame positioning,and the face with the positive key point position is selected to output face frame information. According to the face snapshot method provided by the embodiment of the invention, the average time consumption of single-frame face positioning can be greatly reduced; according to the invention, the key point model with the face confidence coefficient is adopted to carry out face optimization duplicate removal, repeated comparison of the same person in a short time is avoided, and through optimization, the face with the positive key point is selected and sent to the background for comparison, sothat the efficiency of the face recognition system can be well improved.

Description

technical field [0001] The present invention relates to the field of face recognition, and more specifically, to a face capture method, a computer storage medium and electronic equipment. Background technique [0002] The face capture algorithm is a technology based on images obtained by image acquisition / storage devices such as cameras, to determine whether the image contains a human face, and to accurately locate the position of the human face in the image. [0003] This technology is widely used in public security, transportation, schools, criminal investigation, entertainment and other fields. Its technical route can be summarized as follows: traverse the area in the image where there may be a human face, extract the feature information of the human face, and judge whether it belongs to the human face according to the confidence. The early schemes mostly used the method of manually designing features, such as CN201610953612.1 using the MWR AdaBoost algorithm to detect f...

AI Technical Summary

Problems solved by technology

[0005] Methods based on deep learning have gradually appeared in some technical solutions. For example, in patent CN201811109879, MTCNN is used as a face detector. MTCNN is a multi-level cascading method that can achieve real-time face detection in some occasions. However, due to the limitation of feature extraction ability, this method does not have good generalization performance, and in some application scenarios, the effect is not as expected

In addition, MTCNN is not an end-to-end solution, and the detection speed fluctuates greatly

In addition, some people have proposed some end-to-end face capture methods. For example, in the patent CN201810506447, the inventor uses a face detection algorithm based on Single ShotDetector (SSD). The disadvantage is that the network volume is large, and it is difficult to achieve real-time detection effect on the mobile terminal

Images