CN112784808A - Remote face recognition method and system - Google Patents

Remote face recognition method and system Download PDF

Info

Publication number
CN112784808A
CN112784808A CN202110156623.8A CN202110156623A CN112784808A CN 112784808 A CN112784808 A CN 112784808A CN 202110156623 A CN202110156623 A CN 202110156623A CN 112784808 A CN112784808 A CN 112784808A
Authority
CN
China
Prior art keywords
face
recognized
haar
classifier
recognition
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110156623.8A
Other languages
Chinese (zh)
Inventor
龚朋朋
熊敏
陈立伟
姜筱华
单丰武
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi Jiangling Group New Energy Automobile Co Ltd
Original Assignee
Jiangxi Jiangling Group New Energy Automobile Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangxi Jiangling Group New Energy Automobile Co Ltd filed Critical Jiangxi Jiangling Group New Energy Automobile Co Ltd
Priority to CN202110156623.8A priority Critical patent/CN112784808A/en
Publication of CN112784808A publication Critical patent/CN112784808A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
    • G06V10/446—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering using Haar-like filters, e.g. using integral image techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Health & Medical Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Collating Specific Patterns (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a remote face recognition method, which comprises the following steps: carrying out Haar feature detection on a face to be recognized to obtain a face image to be recognized, and accelerating Haar feature evaluation by applying an integral graph algorithm in the detection process; classifying and identifying the Haar characteristics of the face image to be identified by using a weak classifier, wherein the weak classifier is obtained by training the facial characteristics of a sample through an Adaboost algorithm; if the face is identified by the weak classifiers, classifying and identifying the Haar features of the face to be identified by applying a strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers; and if the face is identified by all the strong classifiers, classifying and identifying the Haar features of the face to be identified by using a cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series. The weak classifier, the strong classifier and the cascade classifier are combined for use, so that the recognition accuracy and the recognition speed of the human face can be ensured, and the safety and the efficiency of identity authentication are enhanced.

Description

Remote face recognition method and system
Technical Field
The invention relates to the field of remote face recognition, in particular to a remote face recognition method and system.
Background
Automobiles are developing greatly in the direction of more comfort, intelligence and automation as main vehicles in modern society. The existing intelligent vehicles generally relate to personnel information safety when in use, and some intelligent vehicles need to authenticate the identity of a driver through face remote identification, like the currently widely used network appointment vehicles and sharing vehicles.
The face recognition is a typical problem of image pattern analysis, understanding and classification calculation, and in the face recognition technology, efficient face description features and corresponding high-precision recognition are the key of the technology. In order to realize high-precision face recognition, the existing face recognition system usually directly adopts a strong classifier to classify and recognize face features, but too many face features extracted by the strong classifier bring difficulty to a recognition algorithm, cause recognition errors, cause too low face recognition speed and cannot ensure face screening efficiency; and due to the adoption of the weak classifier, the identification precision is insufficient due to too few identification characteristics, and the safety of identity authentication cannot be ensured.
Disclosure of Invention
The invention aims to provide a remote face recognition method and system, which aim to solve the problems that in order to realize high-precision face recognition, a face recognition system in the prior art usually directly adopts a strong classifier to classify and recognize face features, but because too many face features are extracted by the strong classifier, the recognition algorithm is difficult, recognition errors are caused, the face recognition speed is too low, and the face screening efficiency cannot be ensured.
The invention provides a remote face recognition method, which comprises the following steps:
carrying out Haar feature detection on a face to be recognized to obtain a face image to be recognized, and accelerating Haar feature evaluation by applying an integral graph algorithm in the detection process;
classifying and recognizing the Haar features of the face image to be recognized by using a weak classifier, and judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier, wherein the weak classifier is obtained by training the facial features of a sample through an Adaboost algorithm;
if the Haar features of the face to be recognized pass the recognition of the weak classifiers, applying a strong classifier to classify and recognize the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers;
if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series;
and if the Haar features of the face to be recognized pass through the recognition of the cascade classifier, displaying the personal identity information corresponding to the face image to be recognized.
The remote face recognition method provided by the invention has the following beneficial effects:
according to the invention, the weak classifier, the strong classifier and the cascade classifier are combined for use, so that the identification accuracy of the matched face can be ensured, and the screening speed of the unmatched face can be accelerated, thereby enhancing the safety and efficiency of identity authentication; firstly, classifying and recognizing the Haar features of the face image to be recognized by using a weak classifier, judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier or not, and rapidly recognizing and screening the non-face and the face with low matching degree by using fewer face sample features in the weak classifier; the human face with higher matching degree can enter the next step through the screening of the weak classifiers, and the strong classifiers are used for carrying out feature classification and identification; after the characteristic recognition of the strong classifier is passed, the cascade classifier is applied to classify and recognize the Haar characteristic of the face to be recognized, so that the face characteristic matching precision is further improved, and the safety of identity authentication is well guaranteed.
In addition, the remote face recognition method provided by the invention can also have the following additional technical characteristics:
further, the step of performing Haar feature detection on the face to be recognized to obtain a face image to be recognized, and the step of applying an integral graph algorithm to accelerate Haar feature evaluation in the detection process further comprises:
preprocessing the face image to be recognized, wherein the preprocessing mode comprises the following steps: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
Further, the step of applying the weak classifier to perform classification recognition on the Haar features of the face image to be recognized and judging whether the Haar features of the face to be recognized pass the classification recognition of the weak classifier includes:
calculating the category of the Haar feature vector of the face image to be recognized by applying the weak classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
Further, the step of applying a strong classifier to perform classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier includes:
calculating the category of the Haar characteristic vector of the face image to be recognized by applying the strong classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
Further, the step of applying the cascade classifier to perform classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier includes:
calculating the category of the Haar feature vector of the face image to be recognized by applying the cascade classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the cascade classifiers.
Further, before the step of performing Haar feature detection on the face to be recognized to obtain the face image to be recognized, the method further comprises:
creating the classifier and a trained face recognition library;
projecting Haar features of training samples in the face recognition library to a PCA subspace to obtain Haar feature vectors of the training samples;
adding Haar feature vectors of the training samples to the classifier.
The invention provides a remote face recognition system, which is characterized by comprising the following components:
a detection module: the system comprises a face to be recognized, a face image acquisition unit, a face image processing unit and a face feature evaluation unit, wherein the face to be recognized is subjected to Haar feature detection to obtain a face image to be recognized, and an integral graph algorithm is applied to accelerate Haar feature evaluation in the detection process;
a first identification module: the system comprises a weak classifier, a face recognition module and a face recognition module, wherein the face recognition module is used for applying the weak classifier to carry out classification recognition on the Haar features of the face image to be recognized and judging whether the Haar features of the face to be recognized pass the classification recognition of the weak classifier, and the weak classifier is obtained by training the face features of a sample through an Adaboost algorithm;
a second identification module: if the Haar features of the face to be recognized pass the recognition of the weak classifiers, applying a strong classifier to classify and recognize the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers;
a third identification module: if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series;
an identity display module: and the personal identity information corresponding to the face image to be recognized is displayed if the Haar features of the face to be recognized are recognized by the cascade classifier.
Further, the detection module is further configured to pre-process the facial image to be recognized, where the pre-processing mode includes: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
Further, the first recognition module is also used for applying the weak classifier to calculate the category of the Haar feature vector of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
Further, the second recognition module is also used for applying the strong classifier to calculate the category of the Haar feature vector of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow chart of a remote face recognition method according to an embodiment of the present invention;
fig. 2 is a block diagram of a remote face recognition system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below. Several embodiments of the invention are presented in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
As shown in fig. 1, an embodiment of the present invention provides a remote face recognition method, which includes steps S101 to S104.
S101, performing Haar feature detection on a face to be recognized to obtain a face image to be recognized, and applying an integral graph algorithm to accelerate Haar feature evaluation in the detection process.
Wherein, the step of performing Haar feature detection on the face to be recognized to obtain the face image to be recognized further comprises:
creating the classifier and a trained face recognition library;
projecting Haar features of training samples in the face recognition library to a PCA subspace to obtain Haar feature vectors of the training samples;
adding Haar feature vectors of the training samples to the classifier.
The method comprises the following steps of performing Haar feature detection on a face to be recognized to obtain a face image to be recognized, and accelerating Haar feature evaluation by applying an integral graph algorithm in the detection process:
preprocessing the face image to be recognized, wherein the preprocessing mode comprises the following steps: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
And S102, classifying and recognizing the Haar features of the face image to be recognized by applying a weak classifier, and judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier, wherein the weak classifier is obtained by training the facial features of a sample through an Adaboost algorithm.
The step of applying the weak classifier to classify and recognize the Haar features of the face image to be recognized and judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier comprises the following steps:
calculating the category of the Haar feature vector of the face image to be recognized by applying the weak classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
S103, if the Haar features of the face to be recognized pass the recognition of the weak classifiers, a strong classifier is applied to carry out classification recognition on the Haar features of the face to be recognized, whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier or not is judged, and the strong classifier is obtained by combining a plurality of kinds of the weak classifiers.
And S104, if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series.
The step of applying the strong classifier to classify and recognize the Haar features of the face to be recognized and judging whether the Haar features of the face to be recognized pass the classification and recognition of the strong classifier comprises the following steps:
calculating the category of the Haar characteristic vector of the face image to be recognized by applying the strong classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
And S105, if the Haar features of the face to be recognized are recognized by the cascade classifier, displaying the personal identity information corresponding to the face image to be recognized.
The step of applying the cascade classifier to classify and recognize the Haar features of the face to be recognized and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier comprises the following steps:
calculating the category of the Haar feature vector of the face image to be recognized by applying the cascade classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the cascade classifiers.
In the embodiment, the face information of the user is acquired through the camera to obtain the picture in the RGB format, the picture is processed through the image conversion module to obtain the picture in the YUV format, the picture is subjected to image processing, the image is transmitted to the server through the 4G module, the server compares the transmitted image with local sample data, and the authentication of the user is completed.
The process of the camera acquiring images: the face data enters an imaging chip OV9284 through a LENS LENS, oscillator provides an oscillation circuit for the imaging chip, an image is converted into an image format through an ISP decoding chip, an RAW format is converted into a YUV image format, and the image is subjected to primary processing through a small-size deserializer MAX96705 chip and then enters a main chip for image recognition.
When the camera is accessed, hardware interruption is triggered, the kernel identifies and finds a device driver matched with the kernel according to the device descriptor, and therefore the camera hardware can be operated by using a bottom function interface in an application layer.
The classifier can also be designed to vary with size, since the distance of an object can lead to uncertainty in the size of the target region. It is possible to achieve this with different ratios of scanning windows.
In the embodiment, an OpenCV self-contained cascade classifier harrcade _ front face _ alt2.xml is used, and the face can be well detected, loaded through a Load function of a cascadeClassifier class, and then subjected to face detection.
Before face recognition, a face library needs to be created and trained, in this embodiment, 10 pictures of 10 persons are selected for training, that is, front face images of each person with different postures, expressions and illumination are selected, and the size of the images is 100 × 100. A series of functions of OpenCV can greatly simplify the creation and loading work of a face library, and the codes are as follows:
Ptr<FaceRecognizer>model；
Model＝createEigenFaceRecognizer()；model->train(images，labels)；
model->save("ORL_PCAModel.xml")；model->load("ORL_PCAModel.xml")；
for better detection, the image needs to be preprocessed, and graying is performed first, and a cvtColor function in OpenCV may be used, and the function may convert colors. Histogram equalization, which uses the equalizeHist function provided in OpenCV, is then performed, which enhances image contrast. Summarizing the two processes, the cvtColor is used for converting the picture into the gray-scale image, the contrast of the gray-scale image is improved, the light and shade part of the image can be better distinguished, and the picture is easier to be analyzed by the cascade classifier.
After image preprocessing, face detection can be started. In this embodiment, a detectMultiScale function is adopted, and corresponding parameters are introduced as needed to adjust the detection effect, but a trained high-level classifier is used to detect the required target and return the target to a matrix. The function will search through the image using different sized scanning windows, each time detecting and combining overlapping areas in the image. And after the face area is obtained, adjusting the size of the face area, comparing the face area with a trained face library, judging the size of a threshold value to determine personal identity information, and if the face image cannot be found in the face library, alarming and starting an automatic video recording function.
In summary, the remote face recognition method provided by the invention has the beneficial effects that: according to the invention, the weak classifier, the strong classifier and the cascade classifier are combined for use, so that the identification accuracy of the matched face can be ensured, and the screening speed of the unmatched face can be accelerated, thereby enhancing the safety and efficiency of identity authentication; firstly, classifying and recognizing the Haar features of the face image to be recognized by using a weak classifier, judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier or not, and rapidly recognizing and screening the non-face and the face with low matching degree by using fewer face sample features in the weak classifier; the human face with higher matching degree can enter the next step through the screening of the weak classifiers, and the strong classifiers are used for carrying out feature classification and identification; after the characteristic recognition of the strong classifier is passed, the cascade classifier is applied to classify and recognize the Haar characteristic of the face to be recognized, so that the face characteristic matching precision is further improved, and the safety of identity authentication is well guaranteed.
Referring to fig. 2, the present embodiment provides a remote face recognition system, including:
a detection module: the method is used for carrying out Haar feature detection on the face to be recognized to obtain a face image to be recognized, and an integral graph algorithm is applied to accelerate Haar feature evaluation in the detection process.
The detection module is also used for creating the classifier and a trained face recognition library;
projecting Haar features of training samples in the face recognition library to a PCA subspace to obtain Haar feature vectors of the training samples;
adding Haar feature vectors of the training samples to the classifier.
The detection module is further configured to pre-process the facial image to be recognized, where the pre-processing mode includes: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
A first identification module: the face recognition method comprises the steps of applying a weak classifier to classify and recognize the Haar features of the face image to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the weak classifier, wherein the weak classifier is obtained by training the face features of a sample through an Adaboost algorithm.
The first recognition module is also used for applying the weak classifier to calculate the category of the Haar feature vector of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
A second identification module: and if the Haar features of the face to be recognized pass the recognition of the weak classifiers, applying a strong classifier to classify and recognize the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers.
The second recognition module is also used for applying the strong classifier to calculate the category of the Haar feature vector of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
A third identification module: and if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series.
The third recognition module is also used for calculating the category of the Haar feature vector of the face image to be recognized by applying the cascade classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the cascade classifiers.
An identity display module: and the personal identity information corresponding to the face image to be recognized is displayed if the Haar features of the face to be recognized are recognized by the cascade classifier.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (10)

1. A remote face recognition method, the method comprising:
carrying out Haar feature detection on a face to be recognized to obtain a face image to be recognized, and accelerating Haar feature evaluation by applying an integral graph algorithm in the detection process;
classifying and recognizing the Haar features of the face image to be recognized by using a weak classifier, and judging whether the Haar features of the face to be recognized pass the classification and recognition of the weak classifier, wherein the weak classifier is obtained by training the facial features of a sample through an Adaboost algorithm;
if the Haar features of the face to be recognized pass the recognition of the weak classifiers, applying a strong classifier to classify and recognize the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers;
if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series;
and if the Haar features of the face to be recognized pass through the recognition of the cascade classifier, displaying the personal identity information corresponding to the face image to be recognized.
2. The remote face recognition method according to claim 1, wherein the Haar feature detection is performed on the face to be recognized to obtain the image of the face to be recognized, and the step of applying an integral graph algorithm to accelerate the Haar feature evaluation in the detection process further comprises:
preprocessing the face image to be recognized, wherein the preprocessing mode comprises the following steps: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
3. The remote face recognition method according to claim 2, wherein the step of applying a weak classifier to perform classification recognition on the Haar features of the face image to be recognized and determining whether the Haar features of the face to be recognized pass the classification recognition of the weak classifier comprises:
calculating the category of the Haar feature vector of the face image to be recognized by applying the weak classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
4. The remote face recognition method of claim 2, wherein the step of applying a strong classifier to perform classification recognition on the Haar features of the face to be recognized and determining whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier comprises:
calculating the category of the Haar characteristic vector of the face image to be recognized by applying the strong classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
5. The remote face recognition method according to claim 2, wherein the step of applying a cascade classifier to perform classification recognition on the Haar features of the face to be recognized and determining whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier comprises:
calculating the category of the Haar feature vector of the face image to be recognized by applying the cascade classifier;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the cascade classifiers.
6. The remote face recognition method of claim 1, wherein the step of performing Haar feature detection on the face to be recognized to obtain the face image to be recognized further comprises:
creating the classifier and a trained face recognition library;
projecting Haar features of training samples in the face recognition library to a PCA subspace to obtain Haar feature vectors of the training samples;
adding Haar feature vectors of the training samples to the classifier.
7. A remote face recognition system, comprising:
a detection module: the system comprises a face to be recognized, a face image acquisition unit, a face image processing unit and a face feature evaluation unit, wherein the face to be recognized is subjected to Haar feature detection to obtain a face image to be recognized, and an integral graph algorithm is applied to accelerate Haar feature evaluation in the detection process;
a first identification module: the system comprises a weak classifier, a face recognition module and a face recognition module, wherein the face recognition module is used for applying the weak classifier to carry out classification recognition on the Haar features of the face image to be recognized and judging whether the Haar features of the face to be recognized pass the classification recognition of the weak classifier, and the weak classifier is obtained by training the face features of a sample through an Adaboost algorithm;
a second identification module: if the Haar features of the face to be recognized pass the recognition of the weak classifiers, applying a strong classifier to classify and recognize the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the strong classifier, wherein the strong classifier is obtained by combining a plurality of weak classifiers;
a third identification module: if the Haar features of the face to be recognized pass the recognition of all the strong classifiers, applying a cascade classifier to carry out classification recognition on the Haar features of the face to be recognized, and judging whether the Haar features of the face to be recognized pass the classification recognition of the cascade classifier, wherein the cascade classifier is obtained by connecting all the strong classifiers in series;
an identity display module: and the personal identity information corresponding to the face image to be recognized is displayed if the Haar features of the face to be recognized are recognized by the cascade classifier.
8. The remote face recognition system of claim 7, wherein the detection module is further configured to pre-process the face image to be recognized, and the pre-processing manner includes: graying processing and histogram equalization processing;
and projecting the preprocessed face image to be recognized to a PCA subspace to obtain a Haar characteristic vector of the face image to be recognized.
9. The remote face recognition system of claim 8, wherein the first recognition module is further configured to apply the weak classifier to calculate a class of Haar feature vectors of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
10. The remote face recognition system of claim 8, wherein the second recognition module is further configured to apply the strong classifier to calculate a class of Haar feature vectors of the face image to be recognized;
and judging whether the categories of the Haar feature vectors of the face image to be recognized are matched with the categories of the Haar feature vectors of the training samples in all the weak classifiers.
CN202110156623.8A 2021-02-04 2021-02-04 Remote face recognition method and system Pending CN112784808A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110156623.8A CN112784808A (en) 2021-02-04 2021-02-04 Remote face recognition method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110156623.8A CN112784808A (en) 2021-02-04 2021-02-04 Remote face recognition method and system

Publications (1)

Publication Number Publication Date
CN112784808A true CN112784808A (en) 2021-05-11

Family

ID=75760929

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110156623.8A Pending CN112784808A (en) 2021-02-04 2021-02-04 Remote face recognition method and system

Country Status (1)

Country Link
CN (1) CN112784808A (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103093250A (en) * 2013-02-22 2013-05-08 福建师范大学 Adaboost face detection method based on new Haar- like feature
EP2879078A2 (en) * 2013-12-02 2015-06-03 Huawei Technologies Co., Ltd. Method and apparatus for generating strong classifier for face detection
CN105760881A (en) * 2016-02-01 2016-07-13 南京斯图刻数码科技有限公司 Facial modeling detection method based on Haar classifier method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103093250A (en) * 2013-02-22 2013-05-08 福建师范大学 Adaboost face detection method based on new Haar- like feature
EP2879078A2 (en) * 2013-12-02 2015-06-03 Huawei Technologies Co., Ltd. Method and apparatus for generating strong classifier for face detection
CN105760881A (en) * 2016-02-01 2016-07-13 南京斯图刻数码科技有限公司 Facial modeling detection method based on Haar classifier method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王锋: "基于Adaboost算法的人脸检测与识别技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Similar Documents

Publication Publication Date Title
US7430315B2 (en) Face recognition system
KR100944247B1 (en) System and method for face recognition
US7492926B2 (en) Method for identifying a person from a detected eye image
US20060104504A1 (en) Face recognition method and apparatus
CN102902959B (en) Face recognition method and system for storing identification photo based on second-generation identity card
US7856122B2 (en) Method and device for collating biometric information
JP2006146626A (en) Pattern recognition method and device
JP5726596B2 (en) Image monitoring device
Jayaswal et al. Comparative analysis of human face recognition by traditional methods and deep learning in real-time environment
Basthomi et al. Implementation of RFID attendance system with face detection using validation viola-jones and local binary pattern histogram method
Gangopadhyay et al. FACE DETECTION AND RECOGNITION USING HAAR CLASSIFIER AND LBP HISTOGRAM.
del Rio et al. Face-based recognition systems in the ABC e-gates
CN112307453A (en) Personnel management method and system based on face recognition
CN112784808A (en) Remote face recognition method and system
Mccullagh Face detection by using haar cascade classifier
Ki et al. Face identification for an in-vehicle surveillance system using near infrared camera
JP7337541B2 (en) Information processing device, information processing method and program
Weda et al. Automatic children detection in digital images
Borkar et al. Comparison of real-time face detection and recognition algorithms
Sharma et al. Study and implementation of face detection algorithm using Matlab
CN108487820B (en) Electronic detection closing device
Mohibullah et al. Face Detection and Recognition from Real Time Video or Recoded Video using Haar Features with Viola Jones Algorithm and Eigenface Approach with PCA
Ki et al. Face Identification Using a Near-Infrared Camera in a Nonrestrictive In-Vehicle Environment
Fritz et al. Entropy based saliency maps for object recognition
Jain et al. Face Recognition and Face Detection Using Open Computer Vision Classifiers and Python

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination