Face acquisition and recognition method and system and storage medium
A face collection and recognition method technology, applied in the field of face recognition, can solve problems such as fraud and achieve the effect of improving accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0034] Please refer to figure 1 , showing a method for face acquisition and recognition of the present invention, including:
[0035] Step S1: obtain a large amount of face images, extract the face features of the large amount of face images, and mark the corresponding face position points of the face features, and store them in the face database;
[0036] Step S2: trigger the intelligent terminal to carry out face recognition; extract the face features of the face to be recognized;
[0037] Optionally, said extracting the face features of the face to be recognized includes:
[0038] Based on the convolutional neural network, the shape-driven depth model of facial feature points is established;
[0039] Detect faces through the RetinaFace face detection model, and generate multi-layer detection frames of different sizes; the detection frames contain target frames of different sizes; in each layer of detection frames, the face feature point shape-driven depth model is trained...
Embodiment 2
[0077] Please refer to figure 2 , shows a schematic flow chart of a method according to another embodiment of the present invention. Unlike the previous embodiment, the method of the present invention further includes:
[0078] Step S5: compare the facial features of the face to be recognized with the facial features of a certain face image in the face library, and return the similarity between the facial features of the face to be recognized and the facial features of the face image.
[0079] Specifically, the similarity of facial features can be measured by Euclidean distance and cosine distance. When the direct angle θ between two vectors is closer to 0, the closer the two vectors are, the smaller the difference will be. At this time, cosθ=1, that is, the closer the value is to 1, the more similar the faces are.
[0080] Optionally, the method of the present invention also includes:
[0081] When the distance between the position point of the face feature of the face to ...
Embodiment 3
[0085] A kind of face acquisition characteristic training system of the present invention, comprises:
[0086] The human face sample acquisition module is used to obtain a large number of human face images, extract the facial features of the large number of human facial images, and mark the corresponding human face position points of the human face features, and store them in the human face bank;
[0087] The face recognition module is used for face recognition; extracting the face features of the face to be recognized;
[0088] The face matching module obtains the position points of the face features of the face to be recognized; compares the position points of the face features of the face to be recognized with the position points of the face features of n face images in the face database , n is a natural number; when the distance between the position point of the face feature of the face to be recognized and the position point of the face feature of a certain face image in ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


