CN108198262A - A kind of attendance checking system and implementation method - Google Patents
A kind of attendance checking system and implementation method Download PDFInfo
- Publication number
- CN108198262A CN108198262A CN201810129025.XA CN201810129025A CN108198262A CN 108198262 A CN108198262 A CN 108198262A CN 201810129025 A CN201810129025 A CN 201810129025A CN 108198262 A CN108198262 A CN 108198262A
- Authority
- CN
- China
- Prior art keywords
- attendance
- information
- image
- face
- student
- 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.)
- Granted
Links
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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
Abstract
The invention discloses a kind of attendance checking system and implementation method, system is designed as five layers.Most five layers are hardware layers, camera, campus network and server including being mounted on classroom, server provide computing resource and operating system support for attendance checking system, and camera provides the image information of acquisition, the network equipment uses campus network, is used for transmission image information and control information.4th layer is database layer, provides database environment for attendance checking system, preserves student information, teacher's information, classroom information and attendance information etc..Third layer is component layer, including image collection assembly, Face datection component, recognition of face component, attendance component, course management component, class management component and student-directed component.The second layer includes setup module, acquisition module, attendance module and mark module.First layer is user interface.It is realized using recognition of face optimization algorithm, including three steps:Original shape selection, establishes shape at data set pretreatment.
Description
Technical field
The present invention relates to a kind of attendance checking system design and implementation methods, particularly Scientific Research in University Laboratory and classroom to automate attendance
Occasion uses, and belongs to image procossing attendance technical field.
Background technology
College teaching is the premiere feature of colleges and universities.Due to instructional mode, the diversification of learning hierarchy, mode of learning, education
Teaching system is complicated, and attendance information is an important data resource in college teaching activity.Attendance work is also that school examines
One frequently managed routinely important work.
Majority colleges and universities attendance data statistic is larger at present, needs to occupy teacher and teach teacher's extensive work time, shadow
Ring other teaching operations;Statistical result is not prompt enough, easily omits attendance technical data, education and self-education to student are not
It can carry out in time, so that it cannot level occurs the phenomenon that punishment student;Cause counsellor person, teacher, student can not be timely
Attendance situation is solved, forms failure monitoring.
For a long time, attendance work, for the increasing student attendance information of curricula, adds dependent on manual operation
It is big to manage this workload, and waste a lot of manpower and material resources, and cause lack of standardization in more attendance management.Root
According to the demand that real system is applied, this system utilizes video capture technology and UML, RUP technology, completes to attendance checking system of imparting knowledge to students
Information collection, record.Realize the class hour to experimental teaching and the intelligent management of attendance technical information.
Invention content
The number of the invention encountered in the process for Scientific Research in University Laboratory and teacher's attendance is more, it is more to take, can not accurately in real time
Multi-party monitoring problem proposes, achievees the purpose that automatic attendance using video image acquisition technology and network technology.
In view of the above-mentioned problems, technical solution provided by the present invention is:
A kind of attendance checking system is designed as five layers.Layer 5 is hardware layer, camera, campus network including being mounted on classroom
And server, server provide computing resource and operating system support for attendance checking system, camera provides the image information of acquisition,
The network equipment uses campus network, is used for transmission image information and control information.4th layer is database layer, is provided for attendance checking system
Database environment preserves student information, teacher's information, classroom information and attendance information etc..Third layer is component layer, including image
Acquisition component, Face datection component, recognition of face component, attendance component, course management component, class management component and student's pipe
Manage component.The second layer includes setup module, acquisition module, attendance module and mark module.First layer is user interface.
Layer 5 hardware layer includes camera, server and network.Camera obtain realtime image data, by network with
Server is connected.4th layer data library layer storage student information, teacher's information etc., installing DB is in server.Third
Layer be component layer, including image collection assembly, Face datection component, recognition of face component, attendance component, course management component,
Class management component and student-directed component are coordinations, and be stored in server between each component.The second layer includes setting
Module, acquisition module, attendance module and mark module are put, is mainly directed towards attendance management person, second layer modules can be passed through
To set the merits and demerits pattern of attendance checking system.First layer is user interface, and user interface is mainly directed towards user, can be divided into teacher circle
Face, student interface etc..
System uses camera that can control camera by program in the process at school using the camera with holder
Rotation, camera scanning is multiple to acquire all student's facial images for after.Since classroom internal light is more sufficient and stablizes,
Movement range after student takes a seat on seat is smaller, can obtain the image of high quality.
Attendance checking system is designed as four function modules, and modules can be with independent operating.Image capture module is responsible for control
Camera shoots original image, is transmitted to server, and image is analyzed after original image is carried out image preprocessing, carries
Take human face image information therein;Mark module only uses before student participates in attendance for the first time, and the face information of student is existed
Label is accomplished fluently in system, is then face feature database by the face information " training " for accomplishing fluently label convenient for identifying later;Attendance mould
Block completes identification, and complete according to identity information for collected facial image and face characteristic library to be compared
Attendance works;Setup module is mainly configured student information, curriculum information etc..
Continuity between video frame is utilized in attendance checking system and similitude makes improvements, it is proposed that face location is pre-
Method of determining and calculating tracks the VJ algorithms in video using SVM response diagrams and detects to obtain human face target, and obtain the tracing positional of target.
In terms of face alignment, shape part in CLM algorithm models is optimized, employs that speed is fast, and calculation amount is small, accurately
Slightly lower FPS3000 algorithm constructions original shape is spent to replace the average shape of script, is fitted so as to reduce data in CLM algorithms
Iterations.
Its realization is specific as follows:
Recognition of face optimization algorithm of the present invention mainly includes following three steps:
First, original shape selects
Shape is calculated by FPS3000 to substitute the average shape being obtained by script training sample set to script
CLM algorithms optimize, and improve the speed of CLM algorithms, and keep original accuracy rate constant.
2nd, data set pre-processes
The face shape marked in training sample set is represented using MarkedShape.So MarkedShape is aligned to
The process of FpsShape just contains following step:
Each mark point of MarkedShape is subtracted MarkedShape corresponding positions by the 1st, place normalization
Point, removes the position difference between shape, and formula represents as follows:
2nd, dimension normalization carries out decentralization processing, by the characteristic point of each image and image shape center to data
Subtract each other, representation formula is as follows
Wherein n represents the number of the characteristic point in a shape.Represent MarkedShape in removal displacement respectively
The transverse and longitudinal coordinate of difference ith feature point corresponding with after different scale, andWithSimilar, expression is
FpsShape carries out the transverse and longitudinal coordinate of the ith feature point after decentralization.
Both 3rd, the general formula distance of MarkedShape and FpsShape is defined, final solve needs to find a transformation, make
Between general formula distance it is minimum.The defined formula of general formula distance is as follows:
4th, least square method pair is utilizedIt is rotated, that is, is asked:
When wherein a and b is carries out affine transformation, corresponding rotation parameter can represent as follows:
5th, derivative operation is carried out to above formula, acquires the value of corresponding a and b, formula is as follows:、
6th, the standardized data of MarkedShape is subjected to rotation transformation, obtains the standard shape close to MarkedShape
StandardShape.Operational formula is as follows:
7th, 4-6 processes are repeated, until the difference of process convergence or operation twiceLess than certain threshold value.
3rd, shape is established
After image is normalized, needs to calculate obtained data, acquire MarkedShape and exist
Typical deformation on FpsShape, processing mode are exactly by the mark shape and FpsShape by alignment all in data set
On feature point coordinates subtract each other, the difference a of each imageiIt represents, wherein i refers to data and concentrates the i-th width image.It is specific to calculate
Step is as follows:
1st, the average shape of difference is calculated, by a in data setiCharacter pair point coordinates sum it up and seek its mean value.
2nd, covariance matrix is calculated
3rd, SVD operations are carried out, the characteristic value of corresponding covariance matrix is sought, is represented with λ.Then t feature before being chosen in λ
Amount, i.e. P=(p1,p2,…,pk).T feature vector needs to meet following condition before selection:
Wherein q represents the number of the feature vector after SVD operations, and θ is weight proportion coefficient, and the setting of value generally exists
Between 95%-98%.
4th, on this basis, each mark shape can be acquired by FpsShape via certain operation, formula
It is expressed as follows:
MarkedShapei≈FpsShapei+Pb
Wherein
By above-mentioned process, the shape based on FpsShape is just successfully established.
Advantageous effect
Using computer network, required attendance data is transferred in quick, convenient, accurate transmission;Statistics is late, leaves early, is spacious
Class information.It realizes attendance management automation, changes current attendance management means backwardness, ineffective situation, avoid repeating
Attendance grade, false attendance information, can automatically form various enquiry forms etc., provide attendance data analysis, comparative analysis.Make
Each correlation teacher and leader can understand and grasp student attendance information in time, provide attendance foundation.
Description of the drawings
Fig. 1 is this system schematic diagram.
Specific embodiment
1st, Image Acquisition
Image Acquisition is mainly completed by image capture module.
Acquisition module module is mainly made of two parts, i.e. Image Acquisition and face extraction.
Image Acquisition is completed by camera, the JS-AP2812 monopod video cameras that the present invention selects, in religion indoor deployment one
Camera with holder, at school during camera is controlled to rotate by program, camera acquires after scanning multiple positions
All student's facial images.
Traditional attendance often has fixed attendance time point, such as some time before the class, in class or after class, this be because
Student is needed to participate in for traditional attendance, be the process with student's interaction.Traditional attendance need of work student cooperates on one's own initiative, and attendance is complete
It also needs to result feeding back to student after finishing, student is prompted to complete attendance or absence from duty.It is different from traditional approach, this system
Using the method for recognition of face, student's participation is may not need, system carries out attendance work automatically, need to only be provided before End-of-Course
Checking-in result.The camera of this system must not necessarily be deployed in classroom entrance, also can carry out while progress in course
Image Acquisition.
Since classroom internal light is more sufficient and stablizes, movement range is smaller after student takes a seat on seat in addition,
Cradle head camera is easier to obtain the image of high quality.
Camera shooting acquisition process is run in the operating system of background server, and system calls the API of camera, and realizing will figure
As information is transmitted to the formulation catalogue of background server, and saves as picture file.In order to improve picture quality, need to image
It is pre-processed.System loads camera the image collected automatically, carries out equalization processing and smoothing processing to image respectively,
The contrast of image can be thus improved, inhibits parts of images noise, is conducive to human face detection and recognition.
Face extraction is realized by Face datection algorithm.One face identifier of system loads passes through polling mechanism timing
The catalogue of detection storage acquisition image, if the catalogue includes image file, using face identifier to all areas of image
It is scanned, the facial image scanned is hit storage to specific face catalogue.After some image file end of scan, it is
System will delete its original image, in case same file is again used to face extraction.
2nd, system banner
System banner is mainly completed by mark module, and mark module believes identified face for identifying face information
Breath is trained in " face characteristic library ".The purpose of mark is to add face information for student, and it is special to establish the face for needing checking and studying diligently raw
Library is levied, once student has face information, attendance checking system can just be identified according to face information.According to demand above point
Analysis, the precondition that attendance checking system carries out authentication is must to possess the human face image information of identified object, and to image
The quality of information has certain requirement.From the point of view of recognition effect, facial image is preferably " recent photograph " and image imaging clearly.According to
Existence conditions, system are designed as the image collected generation face characteristic library during the first task of attendance checking system.It does so
Benefit is that the sampling environment of facial image is similar with environment-identification, and time span is small, and the discrimination of face will greatly improve.
Face identification functions include an identification method, in interface identification reality facial image to be identified, are carried for user
For an input frame to input the student number of student to be identified, and a button is provided and is operated for preserving.User, which observes, to be waited to mark
The facial image of knowledge inputs the face according to actual conditions and corresponds to student's student number, and after clicking " preservation " button, system is according to student number
Facial image is identified, forms the correspondence of facial image and student's student number.
The essence of face training is to extract the characteristic information of facial image, its purpose is to establish the face of a student
Feature database, so that facial image is identified.Face training needs to prepare two category informations:Facial image and mark, this two class
Information has generated in face identification function.The correspondence of ready facial image and student number before face training function is loaded into
Relationship is extracted the feature of image by algorithm, feature and mark is saved, generation face characteristic library.
3 system attendances
System attendance is completed by attendance module, and attendance module includes two parts of identification and classroom attendance.
Identification is the Core Feature of attendance checking system.Attendance checking system is loaded into the face characteristic library generated in face training,
Facial image to be identified and the face characteristic in face characteristic library are compared.According to face recognition algorithms, system judgement
Whether there are face and face to be identified to belong to same person in face characteristic library.Face characteristic library can be obtained after identifying successfully
In corresponding face identification information (student number), complete the identification procedure from image to student number.
It is responsible for the attendance logic of system in classroom attendance part.System is according to curriculum information, classroom information and camera first
Information starts the camera specified in the time of course offered and starts attendance work, and the image in classroom passes through image liniment module
It is taken, and takes out facial image wherein included.Then, collected facial image will carry out in face characteristic library
Identification identifies that successful result is updated in Attendance Sheet.Identification success and student number are present in Attendance Sheet
Life is identified as " turning out for work ", and the student not being identified to state in Attendance Sheet remains " not arriving ".Finally, system is according to class
The journey time is automatically stopped attendance work, generates checking-in result.
According to previously described Face datection and face recognition algorithms research shows that, in order to improve attendance accuracy rate and knowledge
Other success rate, by acquisition face direct picture, left avertence side image and right avertence side image.Camera is mounted on classroom middle, when
During the image information of camera acquisition front student, it is matching to transfer face direct picture in image library.When camera is adopted
During the image information of collection left side student, the right avertence side image for transferring image library middle school student is matching.When camera acquires right side
The left avertence side image that image library middle school student are transferred during the image information of student is matching.
4th, background system is set
Setup module includes student's setting, Attendance Sheet management, classroom setting, course offered and camera setting.
The essential information of student is safeguarded in student's setting, includes the multinomial letters such as student number, name, affiliated institute and gender
Breath.
Attendance checking system is required to increase student information, deletes, changes, query function.
Attendance Sheet management mainly includes the setting and inquiry of Attendance Sheet.Attendance Sheet includes curriculum information and student information, class
Journey information corresponds to curriculum schedule, and student information is screened from student's table according to curricula-variable situation.Attendance checking system is according to recognition of face
As a result student corresponding in Attendance Sheet is labeled as " turning out for work ".During End-of-Course, attendance checking system stops the update of Attendance Sheet, newest
Attendance Sheet become the course checking-in result.Attendance query is supplied to one query interface of user, user can by classroom,
The condition queries such as student, course are to checking-in result.
Classroom setting provides a classroom maintenance of information function, and information, the users such as place, coding comprising classroom can be right
Classroom table carries out increasing deletion and changes operation.
Course offered function for setting curriculum information, curriculum information include course coding, course name, classroom of giving lessons,
The information such as time started, end time and classroom, system can increase curriculum information, be deleted, changing operation.
Camera setting is the classroom marked where camera.Camera and classroom are one-to-one relationships, attendance system
The coding in classroom and the IP address of camera where system setting camera, in order to which the camera shooting in specified classroom can be opened during attendance
Head performs acquisition image operation.
Technical solution of the present invention is illustrated below in conjunction with attached drawing.Fig. 1 is this system schematic diagram.
The number of the invention encountered in the process for Scientific Research in University Laboratory and teacher's attendance is more, it is more to take, can not accurately in real time
Multi-party monitoring problem proposes, achievees the purpose that automatic attendance using video image acquisition technology and network technology.
In view of the above-mentioned problems, technical solution provided by the present invention is:
A kind of attendance checking system is designed as five layers.Most five layers are hardware layers, camera, campus network including being mounted on classroom
And server, server provide computing resource and operating system support for attendance checking system, camera provides the image information of acquisition,
The network equipment uses campus network, is used for transmission image information and control information.4th layer is database layer, is provided for attendance checking system
Database environment preserves student information, teacher's information, classroom information and attendance information etc..Third layer is component layer, including image
Acquisition component, Face datection component, recognition of face component, attendance component, course management component, class management component and student's pipe
Manage component.The second layer includes setup module, acquisition module, attendance module and mark module.First layer is user interface.
System uses camera that can control camera by program in the process at school using the camera with holder
Rotation, camera scanning is multiple to acquire all student's facial images for after.Since classroom internal light is more sufficient and stablizes,
Movement range after student takes a seat on seat is smaller, can obtain the image of high quality.
Attendance checking system is designed as four function modules, and modules can be with independent operating.Image capture module is responsible for control
Camera shoots original image, is transmitted to server, and image is analyzed after original image is carried out image preprocessing, carries
Take human face image information therein;Mark module only uses before student participates in attendance for the first time, and the face information of student is existed
Label is accomplished fluently in system, is then face feature database by the face information " training " for accomplishing fluently label convenient for identifying later;Attendance mould
Block completes identification, and complete according to identity information for collected facial image and face characteristic library to be compared
Attendance works;Setup module is mainly configured student information, curriculum information etc..
Continuity between video frame is utilized in attendance checking system and similitude makes improvements, it is proposed that face location is pre-
Method of determining and calculating tracks the VJ algorithms in video using SVM response diagrams and detects to obtain human face target, and obtain the tracing positional of target.
In terms of face alignment, shape part in CLM algorithm models is optimized, employs that speed is fast, and calculation amount is small, accurately
Slightly lower FPS3000 algorithm constructions original shape is spent to replace the average shape of script, is fitted so as to reduce data in CLM algorithms
Iterations.
The recognition of face optimization algorithm that the method for the realization of the system uses, mainly including following three steps:
First, original shape selects
Shape is calculated by FPS3000 to substitute the average shape being obtained by script training sample set to script
CLM algorithms optimize, and improve the speed of CLM algorithms, and keep original accuracy rate constant.
2nd, data set pre-processes
The face shape marked in training sample set is represented using MarkedShape.So MarkedShape is aligned to
The process of FpsShape just contains following step:
Each mark point of MarkedShape is subtracted MarkedShape corresponding positions by the 1st, place normalization
Point, removes the position difference between shape, and formula represents as follows:
2nd, dimension normalization carries out decentralization processing, by the characteristic point of each image and image shape center to data
Subtract each other, representation formula is as follows
Wherein n represents the number of the characteristic point in a shape.Represent MarkedShape in removal displacement respectively
The transverse and longitudinal coordinate of difference ith feature point corresponding with after different scale, andWithSimilar, expression is
FpsShape carries out the transverse and longitudinal coordinate of the ith feature point after decentralization.
Both 3rd, the general formula distance of MarkedShape and FpsShape is defined, final solve needs to find a transformation, make
Between general formula distance it is minimum.The defined formula of general formula distance is as follows:
4th, least square method pair is utilizedIt is rotated, that is, is asked:
When wherein a and b is carries out affine transformation, corresponding rotation parameter can represent as follows:
5th, derivative operation is carried out to above formula, acquires the value of corresponding a and b, formula is as follows:、
6th, the standardized data of MarkedShape is subjected to rotation transformation, obtains the standard shape close to MarkedShape
StandardShape.Operational formula is as follows:
7th, 4-6 processes are repeated, until the difference of process convergence or operation twiceLess than certain threshold value.
3rd, shape is established
After image is normalized, needs to calculate obtained data, acquire MarkedShape and exist
Typical deformation on FpsShape, processing mode are exactly by the mark shape and FpsShape by alignment all in data set
On feature point coordinates subtract each other, the difference a of each imageiIt represents, wherein i refers to data and concentrates the i-th width image.It is specific to calculate
Step is as follows:
5th, the average shape of difference is calculated, by a in data setiCharacter pair point coordinates sum it up and seek its mean value.
6th, covariance matrix is calculated
7th, SVD operations are carried out, the characteristic value of corresponding covariance matrix is sought, is represented with λ.Then t feature before being chosen in λ
Amount, i.e. P=(p1,p2,…,pk).T feature vector needs to meet following condition before selection:
Wherein q represents the number of the feature vector after SVD operations, and θ is weight proportion coefficient, and the setting of value generally exists
Between 95%-98%.
8th, on this basis, each mark shape can be acquired by FpsShape via certain operation, formula
It is expressed as follows:
MarkedShapei≈FpsShapei+Pb
Wherein
By above-mentioned process, the shape based on FpsShape is just successfully established.
Claims (2)
1. a kind of attendance checking system, which is characterized in that including five layers, layer 5 is hardware layer, including be mounted on classroom camera,
Campus network and server, server provide computing resource and operating system support for attendance checking system, and camera provides the figure of acquisition
As information, the network equipment uses campus network, is used for transmission image information and control information;4th layer is database layer, is attendance
System provides database environment, preserves student information, teacher's information, classroom information and attendance information;Third layer is component layer, packet
Include image collection assembly, Face datection component, recognition of face component, attendance component, course management component, class management component and
Student-directed component is coordination, and be stored in server between each component;The second layer includes setup module, acquisition mould
Block, attendance module and mark module set the merits and demerits mould of attendance checking system by second layer modules towards attendance management person
Formula;First layer is user interface, is divided into teacher interface, student interface;
This system be four function modules, image capture module be responsible for control camera shooting original image, by original image into
Server is transmitted to after row image preprocessing, and image is analyzed, extracts human face image information therein;Mark module is only
It is used before student participates in attendance for the first time, the face information of student is accomplished fluently into label in systems, convenient for identifying later, then
It is face feature database by the face information " training " for accomplishing fluently label;Attendance module is used for collected facial image and face is special
Sign library is compared, and completes identification, and complete attendance work according to identity information;Setup module is to student information, course
Information is configured.
2. the implementation method of the system as claimed in claim 1, which is characterized in that system uses camera to use with holder
Camera controls camera to rotate by program in the process at school, and camera scanning is multiple to acquire all student people later
Face image;It is made improvements using the continuity between video frame and similitude, it is proposed that face location prediction algorithm uses
VJ algorithms in SVM response diagrams tracking video detect to obtain human face target, and obtain the tracing positional of target;In face alignment side
Face optimizes shape part in CLM algorithm models, is replaced originally using FPS3000 algorithm constructions original shape
Average shape, so as to reduce the iterations that data in CLM algorithms are fitted, realization includes three steps:
A, original shape selects
Shape is calculated by FPS3000, the CLM of script is calculated by the average shape that script training sample set is obtained to substitute
Method optimizes, and improves the speed of CLM algorithms, and keeps original accuracy rate constant;
B, data set pre-processes
The face shape marked in training sample set is represented using MarkedShape, then is aligned to MarkedShape
The process of FpsShape just contains following step:
Each mark point of MarkedShape is subtracted the point of MarkedShape corresponding positions by b-1), place normalization,
The position difference between shape is removed, formula represents as follows:
B-2), dimension normalization carries out decentralization processing, by the characteristic point of each image and image shape center phase to data
Subtract, representation formula is as follows
Wherein n represents the number of the characteristic point in a shape,Represent MarkedShape in removal displacement difference respectively
The transverse and longitudinal coordinate of ith feature point corresponding with after different scale, andWithIt is similar, expression be FpsShape into
The transverse and longitudinal coordinate of ith feature point after row decentralization;
Both b-3 the general formula distance of MarkedShape and FpsShape), is defined, final solve needs to find a transformation, make
Between general formula distance it is minimum;The defined formula of general formula distance is as follows:
B-4 least square method pair), is utilizedIt is rotated, that is, is asked:
When wherein a and b is carries out affine transformation, corresponding rotation parameter represents as follows:
B-5 derivative operation), is carried out to above formula, acquires the value of corresponding a and b, formula is as follows:、
B-6 the standardized data of MarkedShape), is subjected to rotation transformation, obtains the standard shape close to MarkedShape
StandardShape.Operational formula is as follows:
B-7 b-4), is repeated) to b-6) process, until the difference of process convergence or operation twiceLess than certain threshold value;
C) shape is established
After image is normalized, needs to calculate obtained data, acquire MarkedShape and exist
Typical deformation on FpsShape, processing mode are exactly by the mark shape and FpsShape by alignment all in data set
On feature point coordinates subtract each other, the difference a of each imageiIt represents, wherein i refers to data and concentrates the i-th width image;It is specific to calculate
Step is as follows:
C-1 the average shape of difference) is calculated, by a in data setiCharacter pair point coordinates sum it up and seek its mean value,
C-2 covariance matrix) is calculated
C-3 SVD operations) are carried out, the characteristic value of corresponding covariance matrix is sought, is represented with λ;Then t feature before being chosen in λ
Amount, i.e. P=(p1,p2,…,pk);T feature vector needs to meet following condition before selection:
Wherein q represents the number of the feature vector after SVD operations, and θ is weight proportion coefficient, and the setting of value is generally in 95%-
Between 98%;
C-4) on this basis, each mark shape can be acquired by FpsShape via certain operation, formula table
It states as follows:
MarkedShapei≈FpsShapei+Pb
Wherein
By above-mentioned process, the shape based on FpsShape is just successfully established.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810129025.XA CN108198262B (en) | 2018-02-08 | 2018-02-08 | Attendance system and implementation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810129025.XA CN108198262B (en) | 2018-02-08 | 2018-02-08 | Attendance system and implementation method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108198262A true CN108198262A (en) | 2018-06-22 |
CN108198262B CN108198262B (en) | 2020-05-15 |
Family
ID=62593342
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810129025.XA Active CN108198262B (en) | 2018-02-08 | 2018-02-08 | Attendance system and implementation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108198262B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109040267A (en) * | 2018-08-13 | 2018-12-18 | 河南亚视软件技术有限公司 | A kind of Education Administration Information System based on video |
CN109064578A (en) * | 2018-09-12 | 2018-12-21 | 山西巨擘天浩科技有限公司 | A kind of attendance system and method based on cloud service |
CN109461223A (en) * | 2018-10-29 | 2019-03-12 | 江苏环宇臻视智能科技有限公司 | A kind of classroom roll-call system and method based on recognition of face |
CN109934949A (en) * | 2019-03-12 | 2019-06-25 | 上海商汤智能科技有限公司 | Work attendance method and device, equipment, storage medium |
CN111583251A (en) * | 2020-05-15 | 2020-08-25 | 国网浙江省电力有限公司信息通信分公司 | Video image analysis method and device and electronic equipment |
CN112200924A (en) * | 2020-09-30 | 2021-01-08 | 广东技术师范大学 | Class attendance checking method and system based on face recognition |
JP2021018649A (en) * | 2019-07-22 | 2021-02-15 | パナソニックi−PROセンシングソリューションズ株式会社 | Information processor, attendance management method, and program |
JP7189097B6 (en) | 2019-07-22 | 2022-12-13 | i-PRO株式会社 | Attendance management system, attendance management method, and program |
CN117078219A (en) * | 2023-10-13 | 2023-11-17 | 青岛酒店管理职业技术学院 | School information security management system and method based on Internet of things |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101819687A (en) * | 2010-04-16 | 2010-09-01 | 阜新力兴科技有限责任公司 | Face recognition student attendance device and method |
KR20110053763A (en) * | 2009-11-16 | 2011-05-24 | 박성하 | The automatic measuring instrument manufacturing device for short and middle running |
CN202904683U (en) * | 2012-12-03 | 2013-04-24 | 福州开普软件信息技术有限公司 | Class attendance checking system for campus |
CN105205646A (en) * | 2015-08-07 | 2015-12-30 | 江苏诚创信息技术研发有限公司 | Automatic roll call system and realization method thereof |
CN204990823U (en) * | 2015-10-08 | 2016-01-20 | 吉林工程技术师范学院 | Mutual teaching system for management information system course |
CN107292583A (en) * | 2017-06-20 | 2017-10-24 | 合肥博之泰电子科技有限公司 | A kind of management system and management method based on campus network |
-
2018
- 2018-02-08 CN CN201810129025.XA patent/CN108198262B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20110053763A (en) * | 2009-11-16 | 2011-05-24 | 박성하 | The automatic measuring instrument manufacturing device for short and middle running |
CN101819687A (en) * | 2010-04-16 | 2010-09-01 | 阜新力兴科技有限责任公司 | Face recognition student attendance device and method |
CN202904683U (en) * | 2012-12-03 | 2013-04-24 | 福州开普软件信息技术有限公司 | Class attendance checking system for campus |
CN105205646A (en) * | 2015-08-07 | 2015-12-30 | 江苏诚创信息技术研发有限公司 | Automatic roll call system and realization method thereof |
CN204990823U (en) * | 2015-10-08 | 2016-01-20 | 吉林工程技术师范学院 | Mutual teaching system for management information system course |
CN107292583A (en) * | 2017-06-20 | 2017-10-24 | 合肥博之泰电子科技有限公司 | A kind of management system and management method based on campus network |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109040267A (en) * | 2018-08-13 | 2018-12-18 | 河南亚视软件技术有限公司 | A kind of Education Administration Information System based on video |
CN109064578A (en) * | 2018-09-12 | 2018-12-21 | 山西巨擘天浩科技有限公司 | A kind of attendance system and method based on cloud service |
CN109064578B (en) * | 2018-09-12 | 2020-11-03 | 山西巨擘天浩科技有限公司 | Attendance system and method based on cloud service |
CN109461223A (en) * | 2018-10-29 | 2019-03-12 | 江苏环宇臻视智能科技有限公司 | A kind of classroom roll-call system and method based on recognition of face |
CN109934949A (en) * | 2019-03-12 | 2019-06-25 | 上海商汤智能科技有限公司 | Work attendance method and device, equipment, storage medium |
JP2021018649A (en) * | 2019-07-22 | 2021-02-15 | パナソニックi−PROセンシングソリューションズ株式会社 | Information processor, attendance management method, and program |
JP7189097B2 (en) | 2019-07-22 | 2022-12-13 | i-PRO株式会社 | Attendance management system, attendance management method, and program |
JP7189097B6 (en) | 2019-07-22 | 2022-12-13 | i-PRO株式会社 | Attendance management system, attendance management method, and program |
CN111583251A (en) * | 2020-05-15 | 2020-08-25 | 国网浙江省电力有限公司信息通信分公司 | Video image analysis method and device and electronic equipment |
CN112200924A (en) * | 2020-09-30 | 2021-01-08 | 广东技术师范大学 | Class attendance checking method and system based on face recognition |
CN117078219A (en) * | 2023-10-13 | 2023-11-17 | 青岛酒店管理职业技术学院 | School information security management system and method based on Internet of things |
Also Published As
Publication number | Publication date |
---|---|
CN108198262B (en) | 2020-05-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108198262A (en) | A kind of attendance checking system and implementation method | |
CN104915643B (en) | A kind of pedestrian based on deep learning identification method again | |
CN106204780A (en) | A kind of based on degree of depth study and the human face identification work-attendance checking system and method for cloud service | |
CN110543811B (en) | Deep learning-based non-cooperative examination personnel management method and system | |
Chakraborty et al. | Automatic student attendance system using face recognition | |
CN109903412A (en) | A kind of intelligent check class attendance system based on face recognition technology | |
CN110478862A (en) | A kind of exercise guide system and its guidance method | |
CN110163567A (en) | Classroom roll calling system based on multitask concatenated convolutional neural network | |
CN108766075A (en) | A kind of individualized education analysis system and method based on video analysis | |
CN107659370B (en) | University student classroom total management system and method | |
CN113269903A (en) | Face recognition class attendance system | |
CN116226821B (en) | Teaching data center management system | |
CN211827379U (en) | Access control cluster management system suitable for school uses | |
CN110378261A (en) | A kind of student's recognition methods and device | |
CN115810163A (en) | Teaching assessment method and system based on AI classroom behavior recognition | |
CN107967660A (en) | A kind of safe examination system of automatic face identification | |
CN115273582A (en) | Classroom teaching application system based on artificial intelligence | |
CN110322086A (en) | A kind of education writing on the blackboard checking system based on Intelligent hardware and cloud platform | |
CN112907775A (en) | Attendance system based on face recognition | |
CN114581994A (en) | Class attendance management method and system | |
CN113268512A (en) | Enterprise post professional skill training system based on internet platform | |
Chao et al. | Analysis of the Current Situation of Teaching and Learning of Ideological and Political Theory Courses by Deep Learning | |
Xu | Application of face recognition technology in university educational administration management | |
Sylla et al. | Contribution to attendance and fraud management in engineering schools and digital universities with Deep Learning | |
Patil et al. | Smart Attendance Management System using IoT |
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 | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20230811 Address after: Room A1006, Block A, R&D Building, No. 2 Lijing Road, High tech Zone, Nanjing, Jiangsu Province, 210000 Patentee after: Nanjing Tianqi Fengyun Electronic Technology Co.,Ltd. Address before: 210044 No. 219 Ningliu Road, Nanjing City, Jiangsu Province Patentee before: Nanjing University of Information Science and Technology |