Welcome to the official website of Shenzhen Chongwen Technology Co., Ltd!
描述
描述
0755-29812996 

Analysis of face recognition technology

time :2020-07-15 author : from: scanning : classify :Technical articles
Face image acquisition: different face images can be collected by camera lens, such as static image, dynamic image, different position, different expression and so on.

Face recognition technology is based on the face features of the input face image or video stream. First, whether there is a face is judged. If there is a face, the position and size of each face and the location information of each main facial organ are given. Based on these information, the identity features of each face are extracted and compared with the known faces to recognize the identity of each face.



The broad sense of face recognition actually includes a series of related technologies of constructing face recognition system, including face image acquisition, face location, face recognition preprocessing, identity confirmation and identity search, etc.; while the narrow sense of face recognition refers to the technology or system of identity confirmation or body search through face.


The biometrics studied by biometric technology include face, fingerprint, palm print, iris, retina, voice (voice), body shape, personal habits (such as the intensity and frequency of keystroke, signature) and so on. The corresponding recognition technologies include face recognition, fingerprint recognition, palmprint recognition, iris recognition, retinal recognition, speech recognition (voice recognition can be carried out Only the former belongs to biometric recognition technology), body shape recognition, keyboard tapping recognition, signature recognition, etc.


Face recognition technology consists of three parts


(1) Face detection


Face detection is to judge whether there is a face image in the dynamic scene and complex background, and separate the face image. Generally, there are the following methods:


① Reference template method


Firstly, one or several standard face templates are designed, and then the matching degree between the sample and the standard template is calculated, and the existence of face is judged by threshold value;


② Face rule method


Since the face has certain structural distribution features, the so-called face rule method is to extract these features and generate corresponding rules to determine whether the test sample contains a face;


③ Sample learning method


This method adopts the method of artificial neural network in pattern recognition, that is, the classifier is generated by learning the face sample set and the non face sample set;


④ Skin color model


This method is based on the relatively concentrated distribution of skin color in the color space to detect.


⑤ Feature sub face method


In this method, all the face sets are regarded as a face image subspace, and the existence of an image is judged based on the distance between the sample and its projection in the subspace.


It is worth mentioning that the above five methods can also be used in the actual detection system.


(2) Face tracking


Face tracking refers to the dynamic target tracking of detected faces. The model-based method or the combination of motion and model is adopted. In addition, skin color model tracking is a simple and effective method.


(3) Face comparison


Face comparison is to identify the detected face image or search the object in the face image database. In fact, this means that the sampled face images are compared with the inventory images in order to find the best matching object. Therefore, the description of face image determines the specific method and performance of face recognition. Two description methods are mainly used: feature vector and mask template


① Eigenvector method


In this method, the size, position and distance of the facial contour of the eye iris, nose wing and mouth corner are determined, and then their geometric features are calculated, which form a feature vector describing the facial image.


② Mask method


In this method, some standard face image templates or facial organ templates are stored in the library. During the comparison, all pixels of the sample image are matched with all templates in the library by using the normalized correlation metric. In addition, there are also methods of using pattern recognition autocorrelation network or combining features with templates.


The core of face recognition technology is "local human feature analysis" and "graph / neural recognition algorithm" This algorithm is based on the human facial organs and feature parts. For example, the identification parameters are formed by multiple data corresponding to geometric relations, and compared, judged and confirmed with all the original parameters in the database. Generally, the judgment time is less than 1 second.


Identification process


Generally, there are three steps:


(1) First, the face profile is established. That is to use the camera to collect the face image files of the unit personnel or take their photos to form face image files, and then these face image files are generated into face print code and stored.


(2) Get the current human face. That is to use the camera to capture the face image of the current entry and exit personnel, or take the photo input, and generate the face texture code from the current face image file.


(3) Compare the current mask code with the file inventory. That is to search and compare the face texture code of the current face image with that in the archives inventory. The above "facial pattern coding" method works according to the essential features and the beginning of the face. This facial pattern code can resist changes in light, skin tone, facial hair, hair style, glasses, expression and posture, and has strong reliability, so that it can accurately identify someone from millions of people. The face recognition process can be completed automatically, continuously and in real time by using common image processing equipment.


Technical process


Face recognition system mainly includes four parts: face image acquisition and detection, face image preprocessing, face image feature extraction and matching and recognition.


Face image acquisition and detection


Face image acquisition: different face images can be collected by camera lens, such as static image, dynamic image, different position, different expression and so on. When the user is in the shooting range of the acquisition device, the acquisition device will automatically search and capture the user's face image.