CN112819984B - Classroom multi-person roll-call sign-in method based on face recognition - Google Patents
Classroom multi-person roll-call sign-in method based on face recognition Download PDFInfo
- Publication number
- CN112819984B CN112819984B CN202110043061.6A CN202110043061A CN112819984B CN 112819984 B CN112819984 B CN 112819984B CN 202110043061 A CN202110043061 A CN 202110043061A CN 112819984 B CN112819984 B CN 112819984B
- Authority
- CN
- China
- Prior art keywords
- face
- student
- students
- class
- facenet
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/10—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people together with the recording, indicating or registering of other data, e.g. of signs of identity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K19/00—Record carriers for use with machines and with at least a part designed to carry digital markings
- G06K19/06—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
- G06K19/06009—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking
- G06K19/06037—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking multi-dimensional coding
-
- 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
- G06V40/161—Detection; Localisation; Normalisation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a classroom multi-person roll-call sign-in method based on face recognition, which comprises the following steps: generating a face frame and face feature points through a face detection and face feature extraction network mtcnn; then improving a feature extraction framework inclusion-ResNet-v 1 of a face clustering network facenet, and converting facial feature vectors of the face of the student into an Euclidean vector space by utilizing the facenet to perform face clustering; the student numbers are made into two-dimensional codes, and the two-dimensional codes are used for assisting in rapid comparison of face information of multiple people in a classroom and verification of class roll names; the camera is called to take pictures of students in class, and the process of calling and signing in by multiple people in class is completed through information provided by different pictures. The invention not only greatly improves the working efficiency of attendance registration of the class, but also effectively solves the problems of course escape of students and class attendance of deputy at present.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a classroom multi-person roll-call sign-in method based on face recognition.
Background
The phenomenon is comparatively serious for college's lesson, for the people to answer at present, along with the suggestion of face identification algorithm, this work of calling roll has lightened the burden for the tedious class of mr. The face recognition mainly comprises two processes of face detection alignment and face comparison, the face detection mainly frames the face position, the face detection methods include SSD, S3FD, MTCNN and the like, although the SSD is high in speed, the detection of dense small targets such as faces is poor, S3FD improves the detection performance of the small targets, the speed is low, and the MTCNN-based method has good comprehensive performance, so that the face recognition method has good practical application; after the MTCNN is used for face detection, the process of face comparison is to map the facial feature points provided by the MTCNN to an euclidean space through facenet to perform distance comparison.
At present, a face detection algorithm is mature, a distance comparison method in a face comparison process is low in efficiency, a plurality of methods, such as a KNN (K nearest neighbor) algorithm, an SVM (support vector machine) and the like, are used at present, the KNN nearest neighbor algorithm needs to compare current face image data with faces in a database, and time expenditure is large when a plurality of faces exist; the SVM method realizes the face classification process by finding a decision boundary, however, a new face data needs to be retrained every time, and its practicability is limited.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a classroom multi-person roll-call sign-in method based on face recognition.
In order to achieve the purpose, the technical scheme provided by the invention is as follows: a classroom multi-person roll-call sign-in method based on face recognition comprises the following steps:
s1, making the student numbers into two-dimensional codes, scanning the codes to obtain the numbers of each student, and searching certificate photos of each student by using the numbers;
s2, keeping the face of the student still in the whole photographing process, calling a camera in a classroom to photograph a front face of the student in class, enabling each student to use the two-dimensional code to shield the face of the student, and taking a picture of which all faces are shielded by the two-dimensional code;
s3, obtaining the face position of each student through the first photo, and simultaneously cutting the face parts of the two photos to obtain the actual face image of each student and the face image shielded by the two-dimensional code;
s4, acquiring identification photo information of students by scanning face images shielded by two-dimensional codes, comparing the actual face images of each student with the identification photo by human faces to finish the process of confirming the arrival of the students, and comparing the school number of the student who arrives at the site with the school number of the student of the class to finish the final process of attendance roll registration and sign-in.
In step S1, the study numbers of the students are obtained by scanning the two-dimensional code, and the identification photo of each student is quickly retrieved by using the study numbers, so that the face comparison process is changed from 1 to 1, and the two-dimensional code can be conveniently obtained by the mobile phone applet.
In step S2, the camera is used to take pictures of students in class under different situations, and face detection and identity information acquisition of all students are performed simultaneously, thereby saving time greatly.
In step S3, the face position of the student is completed using the face detection and facial feature extraction network mtcnn, through which the position of the face appearing in the image and the facial feature of each face are obtained.
In step S4, comparing the actual faces of the students with the certificates to make face comparison is performed on the improved face clustering network facenet; the facenet is improved and the face clustering is specifically as follows:
the global average value pooling of the feature extraction network increment-ResNet-v 1 of facenet is changed into convolution calculation, so that the loss of face feature information is reduced, and the face clustering performance is further improved;
the facial features of the human face are converted into Euclidean space vectors through the improved facenet, when the distance between the Euclidean space vectors corresponding to the facial features of the two human face images is smaller than 1, the two human face images are considered to be of the same person, otherwise, the two human face images are not of the same person.
In step S4, the specific process of face comparison is as follows:
acquiring the face position of each student certificate photo through mtcnn, and cutting the face part of each student certificate photo to obtain a certificate photo face image;
the facial features of the actual face image and the identification photo face image are obtained through mtcnn, the corresponding facial features are subjected to European space vectorization through facenet, and the face comparison process is finished by calculating the distance between vectors and judging whether the face images are the same face.
In step S4, the confirmation of the presence of the student is to compare the face image of the student with the identification photo image of the student, and in this way, the report of the student on behalf of the student is stopped.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the problem of roll call of many people on class can be dealt with simultaneously, carry out authentication to many people simultaneously through the form of shooing, the face verification process is accomplished fast with the comparison of individual rather than the certificate photograph, need not extra network training to handle newly-increased personal data.
2. The personal information of the students is only required to be quoted through the school number, so that the personal privacy of the students is protected, and the problem of student information change (such as high school) can be solved.
3. The roll call sign-in process of the students is completed by referring to the student status information of the students, meanwhile, the problem that students attend classes on behalf of people can be prevented, and good guarantee is made for the students to correct the learning attitude.
4. The method improves the feature extraction network of facenet, not only improves the performance of face clustering, but also reduces the dependence on face data of students.
5. The operation process is simple, and the roll call sign-in work of students can be carried out at any time, so that the phenomenon that the students escape from classes in the midway is prevented.
6. And the classroom monitoring camera is only required to be called without depending on other hardware resources, so that the cost is greatly saved.
Drawings
FIG. 1 is a flow chart of roll call sign-in.
Fig. 2 is a flow chart of face comparison.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
As shown in fig. 1 and fig. 2, as a specific implementation case, the training set of the face detection and facial feature point extraction network mtcnn is a widget _ face and CelebA database, and the training set of the face clustering network facenet is an LFW data set. The classroom multi-person roll-call sign-in method based on face recognition comprises the following steps:
1) make into two-dimensional code with student's school code, through scanning the sign indicating number and obtain every student's school code after, utilize the certificate photo of every student of school number retrieval, specifically as follows:
the student number information is obtained by scanning the two-dimensional code, and the certificate photo of each student is retrieved quickly by the student number, so that the face comparison process is changed from 1 pair of changes to 1 pair of changes, and the two-dimensional code can be obtained conveniently through a small mobile phone program.
2) The method comprises the steps that students are kept face still in the whole photographing process, a camera in a classroom is called to photograph a front face of each student in class, each student is enabled to shield the face by a two-dimensional code, and a photo with all faces shielded by the two-dimensional code is photographed. The cameras are called under different scenes to shoot students in class in the classroom, face detection and identity information acquisition of all students are carried out simultaneously, and therefore time is greatly saved.
3) And acquiring the face position of each student through the first photo, and simultaneously cutting the face parts of the two photos to obtain the actual face image of each student and the face image shielded by the two-dimensional code.
The face position of the student is completed by using a face detection and facial feature extraction network mtcnn, and the position of the face appearing in the image and the facial features of each face are obtained through the network. The specific training process of mtcnn is to respectively train three sub-network modules of P-Net, O-Net and R-Net, and carry out face detection and face feature extraction on the best model obtained by training.
4) The identification photo information of students is obtained by scanning the face image shielded by the two-dimensional code, the actual face image of each student is compared with the identification photo through the face to complete the process of confirming the arrival of the students, and the final process of calling and signing in class is completed by comparing the school number of the student who arrives at the scene with the school number of the student of the class.
The actual face of the student and the face comparison of the certificate photography are completed on an improved face clustering facenet network, and the facenet is improved and face clustering conditions are as follows:
the global average value pooling of the feature extraction network increment-ResNet-v 1 of facenet is changed into convolution calculation, so that the loss of face feature information is reduced, and the face clustering performance is further improved. The specific process of training the improved facenet is to face crop the LFW dataset by mtcnn and specify a crop size of 160x160 (due to facenet input of 160x160), and then train the improved facenet using the triplet loss function.
The facial features of the human face are converted into Euclidean space vectors through the improved facenet, when the distance between the Euclidean space vectors corresponding to the facial features of the two human face images is smaller than 1, the two images are considered to be of the same person, otherwise, the two images are of different persons.
The specific process of face comparison is as follows:
and acquiring the face position of each student certificate photo through mtcnn, and cutting the face part of each certificate photo to obtain a certificate photo face image.
The facial features of the actual face image and the identification photo face image are obtained through mtcnn, the corresponding facial features are subjected to European space vectorization through facenet respectively, and the face comparison process is finished by calculating the distance between vectors and judging whether the face images are the same face or not.
The confirmation of the arrival of the student is to compare the face image of the student with the identification photo image of the student, and the report of the student on behalf of the student is prevented by the mode.
In conclusion, the method generates the face frame and the face feature points through the face detection and the face feature extraction network mtcnn; then improving a feature extraction framework inclusion-ResNet-v 1 of a face clustering network facenet, and converting facial feature vectors of the face of the student into an Euclidean vector space by utilizing the facenet to perform face clustering; the student numbers are made into two-dimensional codes, and the two-dimensional codes are used for assisting in rapid comparison of face information of multiple people in a classroom and verification of class roll names; the camera is called to take pictures of students in class, and the process of calling and signing in by multiple people in class is completed through information provided by different pictures. The invention not only greatly improves the working efficiency of attendance registration of the class, but also effectively solves the problems of course escape of students and class attendance of deputy, and is worth popularizing.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (6)
1. A classroom multi-person roll-call sign-in method based on face recognition is characterized by comprising the following steps:
s1, making the student numbers into two-dimensional codes, scanning the codes to obtain the numbers of each student, and searching certificate photos of each student by using the numbers;
s2, keeping the face of the student still in the whole photographing process, calling a camera in a classroom to photograph a front face of the student in class, enabling each student to use the two-dimensional code to shield the face of the student, and taking a picture of which all faces are shielded by the two-dimensional code;
s3, obtaining the face position of each student through the first photo, and simultaneously cutting the face parts of the two photos to obtain the actual face image of each student and the face image shielded by the two-dimensional code;
s4, acquiring identification photo information of students by scanning face images shielded by two-dimensional codes, comparing the actual face images of each student with the identification photo by human faces to finish the process of confirming the arrival of the students, and comparing the school number of the student who arrives at the site with the school number of the student of the class to finish the final process of attendance roll registration and sign-in.
2. The classroom multi-person roll-call sign-in method based on face recognition as claimed in claim 1, wherein: in step S1, the study numbers of the students are obtained by scanning the two-dimensional code, and the identification photo of each student is quickly retrieved by using the study numbers, so that the face comparison process is changed from 1 to 1, and the two-dimensional code can be conveniently obtained by the mobile phone applet.
3. The classroom multi-person roll-call sign-in method based on face recognition as claimed in claim 1, wherein: in step S2, the camera is used to take pictures of students in class under different situations, and face detection and identity information acquisition of all students are performed simultaneously, thereby saving time greatly.
4. The classroom multi-person roll-call sign-in method based on face recognition as claimed in claim 1, wherein: in step S3, the face position of the student is completed using the face detection and facial feature extraction network mtcnn, through which the position of the face appearing in the image and the facial feature of each face are obtained.
5. The classroom multi-person roll-call sign-in method based on face recognition as claimed in claim 1, wherein: in step S4, comparing the actual faces of the students with the certificates to make face comparison is performed on the improved face clustering network facenet; the facenet is improved and the face clustering is specifically as follows:
the global average value pooling of the feature extraction network increment-ResNet-v 1 of facenet is changed into convolution calculation, so that the loss of face feature information is reduced, and the face clustering performance is further improved;
the facial features of the human face are converted into Euclidean space vectors through the improved facenet, when the distance between the Euclidean space vectors corresponding to the facial features of the two human face images is smaller than 1, the two human face images are considered to be of the same person, otherwise, the two human face images are not of the same person.
6. The classroom multi-person roll-call sign-in method based on face recognition as claimed in claim 1, wherein: in step S4, the specific process of face comparison is as follows:
acquiring the face position of each student certificate photo through mtcnn, and cutting the face part of each student certificate photo to obtain a certificate photo face image;
the facial features of the actual face image and the identification photo face image are obtained through mtcnn, the corresponding facial features are subjected to European space vectorization through facenet, and the face comparison process is finished by calculating the distance between vectors and judging whether the face images are the same face.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110043061.6A CN112819984B (en) | 2021-01-13 | 2021-01-13 | Classroom multi-person roll-call sign-in method based on face recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110043061.6A CN112819984B (en) | 2021-01-13 | 2021-01-13 | Classroom multi-person roll-call sign-in method based on face recognition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112819984A CN112819984A (en) | 2021-05-18 |
CN112819984B true CN112819984B (en) | 2022-01-18 |
Family
ID=75869327
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110043061.6A Active CN112819984B (en) | 2021-01-13 | 2021-01-13 | Classroom multi-person roll-call sign-in method based on face recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112819984B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113656843B (en) * | 2021-08-18 | 2022-08-12 | 北京百度网讯科技有限公司 | Information verification method, device, equipment and medium |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111985310A (en) * | 2020-07-08 | 2020-11-24 | 华南理工大学 | Training method of deep convolutional neural network for face recognition |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102831412B (en) * | 2012-09-11 | 2015-01-07 | 魏骁勇 | Teaching attendance checking method and device based on face recognition |
US9076270B2 (en) * | 2013-05-14 | 2015-07-07 | Google Inc. | Generating compositions |
CN104103109A (en) * | 2014-06-13 | 2014-10-15 | 北京汇智大成文化传媒有限公司 | Method for realizing multi-person synchronous self-help conference registration through two-dimensional code scanning |
CN206147769U (en) * | 2016-09-28 | 2017-05-03 | 郑州北斗七星通讯科技有限公司 | Classroom management equipment of registering based on two -dimensional code and face identification |
CN108022318A (en) * | 2017-12-28 | 2018-05-11 | 上海享服信息技术有限公司 | More people's recognition of face attendance checking systems and its Work attendance method |
CN108764149B (en) * | 2018-05-29 | 2022-02-18 | 北京中庆现代技术股份有限公司 | Training method for class student face model |
CN110659397B (en) * | 2018-06-28 | 2022-10-04 | 杭州海康威视数字技术股份有限公司 | Behavior detection method and device, electronic equipment and storage medium |
CN109359548B (en) * | 2018-09-19 | 2022-07-08 | 深圳市商汤科技有限公司 | Multi-face recognition monitoring method and device, electronic equipment and storage medium |
CN109544716A (en) * | 2018-10-31 | 2019-03-29 | 深圳市商汤科技有限公司 | Student registers method and device, electronic equipment and storage medium |
CN110136287B (en) * | 2019-05-13 | 2021-07-13 | 广州唐发信息科技有限公司 | Short-time high-concurrency human face attendance system and method |
CN110545378B (en) * | 2019-09-06 | 2020-09-11 | 广州市奥威亚电子科技有限公司 | Intelligent recognition shooting system and method for multi-person scene |
-
2021
- 2021-01-13 CN CN202110043061.6A patent/CN112819984B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111985310A (en) * | 2020-07-08 | 2020-11-24 | 华南理工大学 | Training method of deep convolutional neural network for face recognition |
Also Published As
Publication number | Publication date |
---|---|
CN112819984A (en) | 2021-05-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109344727B (en) | Identity card text information detection method and device, readable storage medium and terminal | |
CN110210276A (en) | A kind of motion track acquisition methods and its equipment, storage medium, terminal | |
US20200380246A1 (en) | Virtual avatar generation method and apparatus, and storage medium | |
CN105022999B (en) | A kind of adjoint real-time acquisition system of people's code | |
CN109858392B (en) | Automatic face image identification method before and after makeup | |
CN108830559A (en) | A kind of Work attendance method and device based on recognition of face | |
CN105303161A (en) | Method and device for shooting multiple people | |
JP2015219797A (en) | Image collation device, image retrieval system, image collation method, image retrieval method, and program | |
CN108921038A (en) | A kind of classroom based on deep learning face recognition technology is quickly called the roll method of registering | |
CN113591562B (en) | Image processing method, device, electronic equipment and computer readable storage medium | |
CN107766864B (en) | Method and device for extracting features and method and device for object recognition | |
Vadlapati et al. | Facial recognition using the OpenCV Libraries of Python for the pictures of human faces wearing face masks during the COVID-19 pandemic | |
Alburaiki et al. | Mobile based attendance system: face recognition and location detection using machine learning | |
CN113205002B (en) | Low-definition face recognition method, device, equipment and medium for unlimited video monitoring | |
CN112819984B (en) | Classroom multi-person roll-call sign-in method based on face recognition | |
CN110826534B (en) | Face key point detection method and system based on local principal component analysis | |
CN115115552B (en) | Image correction model training method, image correction device and computer equipment | |
CN110633631B (en) | Pedestrian re-identification method based on component power set and multi-scale features | |
CN112001285A (en) | Method, device, terminal and medium for processing beautifying image | |
CN106156739A (en) | A kind of certificate photo ear detection analyzed based on face mask and extracting method | |
CN117496019B (en) | Image animation processing method and system for driving static image | |
CN116205952B (en) | Face recognition and tracking method and device, electronic equipment and storage medium | |
CN112232221A (en) | Method, system and program carrier for processing human image | |
CN112348008A (en) | Certificate information identification method and device, terminal equipment and storage medium | |
CN111611911A (en) | Class attendance checking method based on convolutional neural network and multi-face recognition |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |