A multifunctional electronic class board system and method based on face recognition

Abstract

The application relates to a multifunctional electronic class board system and method based on face recognition, and relates to the technical field of face recognition.The system comprises an electronic class board module, a plurality of electronic class boards, the electronic class board is used for generating an attendance record of a user, the user is a teacher or a student, the electronic class board is also used for adjusting the operation parameters of an electronic teaching device in response to recognizing the teacher; a central control module is used for generating a dynamic attendance scheme based on historical attendance data of the electronic class board and a multi-target attendance optimization function, receiving and storing the attendance record of the user generated by the electronic class board, determining a target electronic class board of to-be-pushed information based on historical activity information of a class corresponding to the electronic class board, and sending the to-be-pushed information to the target electronic class board; the electronic class board is also used for performing user attendance according to the dynamic attendance scheme; and the electronic class board is also used for receiving and displaying the to-be-pushed information, thereby improving attendance efficiency and historical attendance abnormal values.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of face recognition, and particularly relates to a multifunctional electronic class board system and method based on face recognition. BACKGROUND

[0002] In today's society, intelligent and information-based education methods are increasingly valued. Traditional education management models have some problems, such as low attendance efficiency and poor information transmission. Under the background of continuous promotion of smart campus construction, electronic class board, as an important terminal device for campus information management, its functions and application scenarios continue to expand.

[0003] Traditional electronic class boards usually use card swiping or manual check-in methods for attendance management, which has significant drawbacks. Cards are easily lost or misused, resulting in inaccurate attendance data. Manual check-in is inefficient and cannot be used for real-time statistics and tracking, making it difficult to meet the management needs of large groups of students.

[0004] Therefore, it is necessary to provide a multifunctional electronic class board system based on face recognition to improve attendance efficiency and historical attendance outliers. SUMMARY

[0005] The present application provides a multifunctional electronic class board system based on face recognition, comprising: an electronic class board module comprising a plurality of electronic class boards, wherein one electronic class board is installed in each classroom, the electronic class board is used to generate attendance records of users based on face recognition technology, the users are teachers or students, and the electronic class board is also used to adjust the operating parameters of electronic teaching equipment in response to the identification of teachers, the electronic teaching equipment at least includes a motorized podium, a podium lighting device and a podium temperature control device; a central control module for generating a dynamic attendance scheme based on historical attendance data of the electronic class board and a multi-target attendance optimization function, wherein the dynamic attendance scheme includes dynamic attendance times of each electronic class board and attendance auxiliary electronic class boards of each electronic class board, the central control module is also used to receive and store the attendance records of users generated by the electronic class board, obtain the information to be pushed, determine the target electronic class board of the information to be pushed based on the historical activity information of the class corresponding to the electronic class board, and send the information to be pushed to the target electronic class board; the electronic class board is also used to perform user attendance according to the dynamic attendance scheme; and the electronic class board is also used to receive and display the information to be pushed.

[0006] Further, the central control module generates a dynamic attendance scheme based on historical attendance data of the electronic class boards and a multi-objective attendance optimization function, including: for each electronic class board, determining a single attendance computing power requirement of a class corresponding to the electronic class board and historical attendance abnormal values of a plurality of attendance time periods based on historical attendance data of the electronic class board; and generating the dynamic attendance scheme based on the single attendance computing power requirement of the class corresponding to each electronic class board, the historical attendance abnormal values of the plurality of attendance time periods, and the multi-objective attendance optimization function.

[0007] Further, the dynamic attendance scheme is generated based on the single attendance computing power requirement of the class corresponding to each electronic class board, the historical attendance abnormal values of the plurality of attendance time periods, and the multi-objective attendance optimization function, including: for each electronic class board, determining a candidate attendance time period corresponding to the electronic class board based on the historical attendance abnormal values of the plurality of attendance time periods; for any two electronic class boards, calculating an attendance complementary parameter of the two electronic class boards based on the historical attendance abnormal values of the plurality of attendance time periods of the two electronic class boards; establishing a constraint condition set, wherein the constraint condition set at least includes a sampling range constraint of a dynamic attendance time and a computing power load constraint of a single electronic class board; generating a plurality of candidate dynamic attendance schemes based on the constraint condition set and the attendance complementary parameter of any two electronic class boards; for each candidate dynamic attendance scheme, calculating a multi-objective attendance optimization function value of the candidate dynamic attendance scheme; and generating the dynamic attendance scheme based on the multi-objective attendance optimization function value of each candidate dynamic attendance scheme.

[0008] Further, the multi-objective attendance optimization function is at least related to a computing power load balance degree of the plurality of electronic class boards, a shortest time required for attendance of each electronic class board, and a historical attendance abnormal value of a dynamic attendance time of each electronic class board.

[0009] Further, the candidate attendance time period corresponding to the electronic class board is determined based on the historical attendance abnormal values of the plurality of attendance time periods, including: for any two electronic class boards, determining an attendance similarity of the two electronic class boards based on the historical attendance abnormal values of the plurality of attendance time periods of the two electronic class boards; for each electronic class board, determining a similar electronic class board of the electronic class board based on the attendance similarity of any two electronic class boards; for each electronic class board, determining a to-be-corrected attendance time period based on an attendance frequency of the plurality of attendance time periods of the electronic class board; correcting a historical attendance abnormal value of the to-be-corrected attendance time period of the electronic class board based on the historical attendance abnormal value of the to-be-corrected attendance time period of the similar electronic class board of the electronic class board; and determining the candidate attendance time period corresponding to the electronic class board based on the corrected historical attendance abnormal values of the plurality of attendance time periods.

[0010] Further, the electronic class board generates an attendance record of the user based on a face recognition technology, including: extracting a plurality of face features of a registered user corresponding to the electronic class board; determining a feature matching order based on the plurality of face features of the registered user corresponding to the electronic class board; grouping the registered user corresponding to the electronic class board based on the face features of the registered user corresponding to the electronic class board and the feature matching order, to determine a plurality of user groups corresponding to the electronic class board; collecting a face image of a to-be-identified person, and extracting a plurality of face features of the to-be-identified person; and generating an attendance record of the user based on the plurality of face features of the to-be-identified person, the feature matching order, and the plurality of user groups corresponding to the electronic class board.

[0011] Further, the feature matching order is determined based on the plurality of face features of the registered user corresponding to the electronic class board, including: determining a plurality of face feature pairs, wherein each face feature pair includes two face features; determining a feature difference value of each face feature pair based on the plurality of face features of the registered user corresponding to the electronic class board; and determining the feature matching order based on the feature difference value of each face feature pair.

[0012] Further, the electronic class board adjusts the operating parameters of the electronic teaching device, including: collecting real-time environmental features; obtaining environmental features of a historical teaching time period corresponding to a teacher; determining whether there is a similar historical teaching time period based on the real-time environmental features and the environmental features of the historical teaching time period corresponding to the teacher, and if so, retrieving operating parameters of the electronic teaching device of the similar historical teaching time period, adjusting the operating parameters of the electronic teaching device, and if not, determining a similar teacher based on the operating parameters of the electronic teaching device of the historical teaching time period corresponding to the teacher, determining a similar historical teaching time period based on environmental features of the historical teaching time period corresponding to the similar teacher, and adjusting the operating parameters of the electronic teaching device based on the operating parameters of the electronic teaching device of the similar historical teaching time period.

[0013] Further, the target electronic class board of the to-be-pushed information is determined based on the historical activity information of the class corresponding to the electronic class board, including: establishing a class portrait of the class corresponding to the electronic class board based on the historical activity information of the class corresponding to the electronic class board; and determining the target electronic class board of the to-be-pushed information based on the class portrait of the class corresponding to each electronic class board.

[0014] The application provides a multi-functional electronic class board method based on face recognition, which is applied to the multi-functional electronic class board system based on face recognition, and comprises the following steps of: generating a dynamic attendance scheme based on historical attendance data of the electronic class board and a multi-target attendance optimization function, wherein the dynamic attendance scheme comprises dynamic attendance time of each electronic class board and an attendance auxiliary electronic class board of each electronic class board; performing user attendance according to the dynamic attendance scheme, and generating an attendance record of the user based on face recognition technology, wherein the user is a teacher or a student; in response to recognizing the teacher, adjusting operation parameters of an electronic teaching device, wherein the electronic teaching device at least comprises a motorized podium, podium lighting equipment and podium temperature control equipment; obtaining to-be-pushed information, determining a target electronic class board of the to-be-pushed information based on historical activity information of a class corresponding to the electronic class board, and sending the to-be-pushed information to the target electronic class board; and receiving and displaying the to-be-pushed information by the electronic class board.

[0015] Compared with the prior art, the multi-functional electronic class board system and method based on face recognition provided by the application have at least the following beneficial effects:

[0016] 1. The electronic class board module generates the attendance record of the teacher or the student by using the face recognition technology, thereby avoiding problems such as proxy signing and greatly improving the accuracy and authenticity of the attendance compared with the traditional attendance mode. Meanwhile, the central control module generates the dynamic attendance scheme based on the historical attendance data and the multi-target attendance optimization function, and the dynamic attendance scheme comprises the dynamic attendance time of each electronic class board and the attendance auxiliary electronic class board, so that the attendance arrangement is more suitable for the actual teaching situation, and the efficiency and flexibility of the attendance management are improved.

[0017] 2. After recognizing the teacher, the electronic class board can automatically adjust the operation parameters of the electronic teaching device (such as the motorized podium, the podium lighting equipment and the podium temperature control equipment). This function can create a suitable teaching environment according to the teaching habits of different teachers and the real-time environmental conditions, and is helpful to improve the teaching quality and the teaching comfort of the teacher.

[0018] 3. After obtaining the to-be-pushed information, the central control module accurately determines the target electronic class board and sends the information based on the historical activity information of the class corresponding to the electronic class board. The electronic class board timely receives and displays the to-be-pushed information, so that the students and the teachers can quickly obtain the content such as the notification and the activity arrangement related to themselves, and the circulation and communication efficiency of the information is improved. BRIEF DESCRIPTION OF DRAWINGS

[0019] The application will be further described in the form of exemplary embodiments, which will be described in detail with reference to the drawings. These embodiments are not restrictive, and in these embodiments, the same numbers represent the same structures, wherein:

[0020] Figure 1is a module diagram of a multifunctional electronic class board system based on face recognition shown in an embodiment of the present application;

[0021] Figure 2 is a flowchart of generating a dynamic attendance scheme shown in an embodiment of the present application;

[0022] Figure 3 is a flowchart of a multifunctional electronic class board method based on face recognition shown in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description.

[0024] Figure 1 is a module diagram of a multifunctional electronic class board system based on face recognition shown in an embodiment of the present application, as Figure 1 shown, a multifunctional electronic class board system based on face recognition can include an electronic class board module and a central control module.

[0025] The electronic class board module includes a plurality of electronic class boards, wherein one electronic class board is installed in each classroom, the electronic class board is used to generate an attendance record of a user based on face recognition technology, the user is a teacher or a student, the electronic class board is also used to adjust operating parameters of an electronic teaching device in response to recognizing a teacher, the electronic teaching device at least includes a motorized podium, a podium lighting device and a podium temperature control device.

[0026] Specifically, the electronic class board as a core interactive terminal in a smart campus scene integrates a main screen, an auxiliary screen, a card reader and a multifunctional control module, supports functions such as touch writing, device Internet of Things, identity recognition and remote management. Its hardware structure is based on the concept of integrated design, and high-efficiency teaching service and device management and control are realized through multi-module cooperation.

[0027] The main screen adopts a 23.8-inch IPS liquid crystal panel, with a resolution of 1920x1080 (16:9 ratio), a contrast ratio of 1000:1 and a brightness of 250 cd / m², which can clearly present course information, announcements and multimedia content. The screen supports capacitive+electromagnetic dual-touch technology, compatible with finger and electromagnetic pen operation, realizes 10-point touch and 5080LPI handwriting resolution, and the pressure level reaches 8192 levels, meeting the needs of fine writing and annotation. The electromagnetic touch layer accurately captures the trajectory of the handwriting through electromagnetic induction technology, which is suitable for blackboard teaching, electronic homework correction and other scenes; the capacitive touch layer supports gesture sliding, zooming and other fast operations, which improves the interaction efficiency.

[0028] The auxiliary screen is a 10.1-inch vertical screen design (10:16 ratio) with a resolution of 1200x1920 and a brightness of 350 cd / m². It uses capacitive touch technology and is built-in with an Android 11 operating system. It is equipped with a quad-core 1.8GHz processor, 4GB of memory, and 32GB of storage space, and can run teaching applications independently or serve as a device control hub. The auxiliary screen realizes three major functions through pre-installed management software:

[0029] 1. Multimedia Internet of Things Management and Control: After connecting with the intelligent central control host, it can control classroom projectors, electronic whiteboards, lighting, air conditioning, and other devices, and support one-key switching of scene modes (such as automatically turning on the projector and dimming the lights in class mode);

[0030] 2. Interface Custom Configuration: Supports remote or local programming of background, operation interface, and keys, and configuration files can be backed up and distributed on the cloud to meet the individual needs of different schools or classrooms;

[0031] 3. Dynamic QR Code Generation: Real-time display of dynamic identity verification QR code, user completes permission authentication by scanning the code with a mobile phone, realizes "scan to class" or device access control.

[0032] Integrated IC card reading and writing module, compatible with school card system, supports card swiping / insertion to start devices, and records user operation logs for audit. Equipped with a 5 million-pixel wide-angle camera, supports live detection and 1:N fast comparison, used for device use permission identification and visual intercom function. The camera has an infrared fill light, which can realize high-definition imaging in low-light environments. Integrated with high-sensitivity pickup microphone and stereo speaker, after connecting with the IP voice intercom host, it supports two-way real-time communication between the classroom and the control room, with call quality reaching VoIP standard and delay below 200ms.

[0033] The electronic class board is built-in with a high-integration control board, providing the following interfaces:

[0034] Power management: ≥8 programmable AC / DC switches, supporting device timing start / stop and energy consumption monitoring;

[0035] Serial communication: ≥1 RS232 interface for connecting traditional teaching devices (such as power amplifiers, exhibition stands);

[0036] Infrared control: ≥4 infrared emission interfaces that can learn and simulate air conditioner, curtain, and other device remote control signals;

[0037] Data transmission: ≥4 USB 3.0 interfaces, 3 HDMI input / output interfaces, and built-in 8-port gigabit switch (including 3 physical ports) to meet multimedia content transmission and device cascading needs;

[0038] Bidirectional IO interface: ≥2 programmable interfaces supporting linkage with electronic door locks, sensors, and other external devices.

[0039] Device integration network module, supporting wired / wireless dual-mode access to campus network, realizing three major remote functions:

[0040] 1. Centralized visual management: real-time monitoring of device status, online rate and fault alarm through management platform;

[0041] 2. Remote parameter configuration: support batch modification of device time, network settings, touch sensitivity and other parameters;

[0042] 3. Firmware OTA upgrade: automatically detect and download the latest firmware version to reduce on-site maintenance costs.

[0043] Electronic class board integrates Wi-Fi 6 and Bluetooth 5.0 dual-mode wireless communication chips to realize stable connection with campus network and short-distance communication with electronic teaching equipment. The Wi-Fi module supports 2.4GHz / 5GHz dual-band, and the data transmission rate can reach 1.2Gbps, ensuring that the attendance records are uploaded to the central control module in real time. The Bluetooth module is used to pair with local devices such as electric podiums, lighting equipment, etc. Control instructions are transmitted through low-power protocols to reduce wiring costs.

[0044] The electronic class board contains multiple relay output interfaces and PWM dimming interfaces, which can directly control the lifting of the electric podium, the switching and brightness adjustment of the podium lighting equipment, and the start-stop and temperature setting of the temperature control equipment. The module receives the teacher identity recognition result through the core processing unit, automatically calls the preset parameters to complete the equipment adjustment, and supports manual override of automatic settings through the touch screen to meet individual needs.

[0045] The central control module is used to generate a dynamic attendance scheme based on the historical attendance data of the electronic class board and a multi-target attendance optimization function, wherein the dynamic attendance scheme includes a dynamic attendance time for each electronic class board and an attendance auxiliary electronic class board for each electronic class board, and is also used to receive and store the attendance records of the users generated by the electronic class board, and to obtain the to-be-pushed information and determine the target electronic class board of the to-be-pushed information based on the historical activity information of the class corresponding to the electronic class board, and send the to-be-pushed information to the target electronic class board.

[0046] In some embodiments, the central control module generates a dynamic attendance scheme based on the historical attendance data of the electronic class board and a multi-target attendance optimization function, including:

[0047] For each electronic class board, based on the historical attendance data of the electronic class board, determine the single attendance computing power requirement of the class corresponding to the electronic class board and the historical attendance abnormal value of the multiple attendance time periods, wherein a day can be divided into multiple attendance time periods, for example, 08:00-08:30 is an attendance time period, 12:00-12:30 is an attendance time period, etc.

[0048] Based on the single attendance computing power demand of each electronic class board corresponding to the class, the historical attendance abnormal value of multiple attendance time periods, and the multi-objective attendance optimization function, a dynamic attendance scheme is generated.

[0049] Specifically, the time consumption of the electronic class board to complete a face recognition, i.e. single attendance time consumption, can be calculated. The product of the number of students in the class corresponding to the electronic class board and the single attendance time consumption is taken as the single attendance computing power demand of the class corresponding to the electronic class board.

[0050] For each attendance of the class corresponding to the electronic class board, if there is at least one absent student, it is determined that attendance abnormality occurs in the attendance time period.

[0051] The historical attendance abnormal value of the attendance time period can be calculated according to the following formula:

[0052]

[0053] Wherein, is the historical attendance abnormal value of the i-th attendance time period, is the number of absent students in the n-th attendance of the i-th attendance time period, is the total number of students in the class corresponding to the electronic class board, is the total number of attendances of the i-th attendance time period, is the number of times of attendance abnormality occurring in the i-th attendance time period.

[0054] It can be understood that, The absence proportions of all N attendances in the i-th attendance time period are summed up. This step summarizes the absence of each attendance in the time period to obtain a comprehensive absence proportion sum. For example, if the absence proportions of 3 attendances in the time period are 0.1, 0.15 and 0.05 respectively, the sum is 0.1+0.15+0.05=0.3. Dividing the above sum by the total number of attendances N, the average absence proportion of each attendance in the i-th attendance time period is obtained. The proportion of the number of times of attendance abnormality in the total number of attendances in the i-th attendance time period is calculated, which reflects the frequency of occurrence of attendance abnormality in the time period. The average absence proportion of each attendance is multiplied by the frequency of attendance abnormality to obtain the historical attendance abnormal value of the i-th attendance time period . This value comprehensively considers the severity of absence and the frequency of abnormality, and can more comprehensively reflect the abnormality of the attendance time period. For example, the average absence proportion of each attendance is 0.1, and the frequency of attendance abnormality is 0.2, then =0.1×0.2=0.02. The greater the value, the more serious the abnormality of the attendance time period.

[0055] Figure 2 is a flowchart of generating a dynamic attendance scheme in an embodiment of the present application, as shown, in some embodiments, based on the single attendance computing power requirement of each electronic class board corresponding to the class, the historical attendance abnormal value of the plurality of attendance time periods, and the multi-objective attendance optimization function, a dynamic attendance scheme is generated, including: Figure 2

[0056] For each electronic class board, based on the historical attendance abnormal value of the plurality of attendance time periods, a candidate attendance time period corresponding to the electronic class board is determined;

[0057] For any two electronic class boards, based on the historical attendance abnormal value of the plurality of attendance time periods of the two electronic class boards, the attendance complementary parameter of the two electronic class boards is calculated;

[0058] A constraint condition set is established, wherein the constraint condition set at least includes the sampling range constraint of the dynamic attendance time and the computing power load constraint of a single electronic class board;

[0059] Based on the constraint condition set and the attendance complementary parameter of any two electronic class boards, a plurality of candidate dynamic attendance schemes are generated;

[0060] For each candidate dynamic attendance scheme, the multi-objective attendance optimization function value of the candidate dynamic attendance scheme is calculated;

[0061] Based on the multi-objective attendance optimization function value of each candidate dynamic attendance scheme, a dynamic attendance scheme is generated.

[0062] In some embodiments, based on the historical attendance abnormal value of the plurality of attendance time periods, the candidate attendance time period corresponding to the electronic class board is determined, including:

[0063] For any two electronic class boards, based on the historical attendance abnormal value of the plurality of attendance time periods of the two electronic class boards, the attendance similarity of the two electronic class boards is determined;

[0064] For each electronic class board, based on the attendance similarity of any two electronic class boards, the similar electronic class board of the electronic class board is determined;

[0065] For each electronic class board, the attendance frequency of the plurality of attendance time periods of the electronic class board is determined, and the attendance time period to be corrected is determined;

[0066] Based on the historical attendance abnormal value of the attendance time period to be corrected of the similar electronic class board of the electronic class board, the historical attendance abnormal value of the attendance time period to be corrected of the electronic class board is corrected;

[0067] Based on the historical attendance abnormal value of the plurality of corrected attendance time periods, the candidate attendance time period corresponding to the electronic class board is determined. ​

[0068] Specifically, for each electronic class board, a check-in vector can be generated based on historical check-in abnormal values of multiple check-in time periods of the electronic class board, where the i-th element of the check-in vector is the historical check-in abnormal value of the i-th check-in time period of the electronic class board.

[0069] For any two electronic class boards, a cosine similarity of check-in vectors of the two electronic class boards is calculated as a check-in similarity of the two electronic class boards. If the check-in similarity of the two electronic class boards is greater than a check-in similarity threshold (e.g., 0.5), the two electronic class boards are similar electronic class boards.

[0070] A check-in time period with a check-in frequency less than a check-in frequency threshold (e.g., 5) can be regarded as a check-in time period to be corrected of an electronic class board.

[0071] A historical check-in abnormal value of a check-in time period to be corrected of a similar electronic class board of an electronic class board can be averaged as a corrected historical check-in abnormal value of the check-in time period to be corrected.

[0072] A check-in time period with a historical check-in abnormal value greater than a historical check-in abnormal value threshold (e.g., 0.3) can be regarded as a candidate check-in time period corresponding to an electronic class board.

[0073] A check-in complementary parameter of two electronic class boards can be calculated according to absolute values of differences of historical check-in abnormal values of each check-in time period of the two electronic class boards.

[0074] For example, the check-in complementary parameter of two electronic class boards can be calculated according to the following formula:

[0075]

[0076] wherein, is a check-in complementary parameter of an h-th electronic class board and a g-th electronic class board, is a historical check-in abnormal value of the h-th electronic class board in an i-th check-in time period, is a historical check-in abnormal value of the g-th electronic class board in the i-th check-in time period, is a total number of check-in time periods.

[0077] The sampling range constraint of dynamic check-in time can be to sample in the candidate check-in time period corresponding to an electronic class board.

[0078] The computing power load constraint of a single electronic class board can be that a real-time load of the single electronic class board needs to be less than a maximum load of the electronic class board. The maximum load of the electronic class board is determined based on hardware devices of the electronic class board.

[0079] The larger the attendance complementarity parameter of the two electronic attendance panels, the higher the probability that one of the electronic attendance panels serves as an attendance auxiliary electronic attendance panel for the other electronic attendance panel.

[0080] In combination with the above set of constraints and the attendance complementarity parameter, a certain algorithm or rule is used to generate a plurality of different candidate dynamic attendance schemes. For example, an exhaustive method can be used to try all possible combinations of attendance times under the premise of meeting the constraints. And using a roulette wheel selection algorithm, the complementarity parameter of each electronic attendance panel is used as the weight to randomly select the attendance auxiliary electronic attendance panel. For example, the complementarity parameter of electronic attendance panel A and electronic attendance panel B is 0.6, and the complementarity parameter of electronic attendance panel C is 0.4, so the probability of electronic attendance panel B being selected as the attendance auxiliary electronic attendance panel for electronic attendance panel A is higher.

[0081] In some embodiments, the multi-objective attendance optimization function is at least related to the load balancing degree of the plurality of electronic attendance panels, the shortest time required for attendance of each electronic attendance panel, and the historical attendance abnormal value of the dynamic attendance time of each electronic attendance panel.

[0082] Specifically, for each candidate dynamic attendance scheme, the number of attendances that each electronic attendance panel needs to process can be calculated, and the variance of the number of attendances that each electronic attendance panel needs to process is calculated. The larger the variance, the smaller the load balancing degree of the plurality of electronic attendance panels.

[0083] For each electronic attendance panel, when performing attendance, the attendance task can be assigned to the attendance auxiliary electronic attendance panel, i.e., the collected image is sent to the attendance auxiliary electronic attendance panel for face recognition. The shortest time required for attendance of each electronic attendance panel can be the maximum of the time for the electronic attendance panel to complete the attendance and the time for the attendance auxiliary electronic attendance panel to complete the attendance.

[0084] For example only, the multi-objective attendance optimization function can be:

[0085]

[0086] wherein, is the multi-objective attendance optimization function, is the normalized load balancing degree of the plurality of electronic attendance panels, is the normalized, is the normalized shortest time required for attendance of the hth electronic attendance panel, is the normalized historical attendance abnormal value of the dynamic attendance time of the hth electronic attendance panel.

[0087] The function integrates the computing power load balancing degree, the total length of attendance and the historical attendance abnormal value three important factors, and guides the generation of the attendance scheme through a unified optimization target, can comprehensively and systematically consider the influence of various factors on the attendance, find a better attendance scheme, and improve the quality and level of the overall attendance management.

[0088] The candidate dynamic attendance scheme with the maximum multi-objective attendance optimization function value can be used as the dynamic attendance scheme.

[0089] In some embodiments, the central control module determines the target electronic class board of the information to be pushed based on the historical activity information of the class corresponding to the electronic class board, including:

[0090] Based on the historical activity information of the class corresponding to the electronic class board, a class portrait of the class corresponding to the electronic class board is established.

[0091] Based on the class portrait of the class corresponding to each electronic class board, the target electronic class board of the information to be pushed is determined.

[0092] Specifically, the central control module will widely collect various types of historical activity information related to the class corresponding to the electronic class board. These information sources are diverse, and may include records of class participation in campus activities such as sports meetings, cultural performances, and club activities; class schedule and learning situation, such as test scores of different subjects, homework completion, classroom participation, etc.; daily behavior management data of the class, such as attendance records, discipline cases, reward cases, etc. By comprehensively collecting these information, it can provide rich materials for building a class portrait.

[0093] In-depth analysis of the collected historical activity information to mine the characteristics and laws contained therein. For example, by analyzing the class participation in campus activities, it can be found that a certain class is active in sports and often participates in various sports competitions and achieves excellent results; or a class has a strong interest in cultural and artistic activities and actively participates in painting, music and other club activities. From the course learning information, it can be understood that the overall learning style of the class is inclined to active inquiry or passive acceptance, and the strengths and weaknesses in various subjects. Daily behavior management data can reflect the discipline and team cooperation spirit of the class.

[0094] According to the characteristics and laws obtained by analysis, a unique portrait is constructed for each class. The class portrait can be presented in various forms, such as a series of keywords to describe the characteristics of the class, such as "sports special class", "cultural active class", "learning diligent class", etc.

[0095] The central control module will carefully match the information to be pushed with the portraits of each class. For example, if the information to be pushed is a notification about an upcoming sports match, the classes whose portraits show an interest in sports activities, good sports performance, or students with sports special talents will be screened as the target classes; if it is information about a cultural and artistic exhibition, classes that are active in the arts, have related social activities, or students have a high interest in art will be selected as the target classes. Through this precise matching, it can be ensured that the information to be pushed is highly consistent with the needs and interests of the class.

[0096] According to the matching results of the information and the class portraits, the central control module finally determines the target electronic class board of the information to be pushed. The classes corresponding to these target electronic class boards are the groups that are most likely to be interested in the pushed information and most in need of receiving the information. By accurately pushing the information to these target electronic class boards, the effective transmission rate of the information can be improved, unnecessary diffusion and waste of the information can be avoided, and the needs of the class can be better met, thereby improving the effect and value of information pushing.

[0097] The electronic class board is also used to perform user attendance according to a dynamic attendance scheme.

[0098] In some embodiments, the electronic class board generates an attendance record of a user based on a face recognition technology, including:

[0099] Extracting multiple face features of a registered user corresponding to the electronic class board, wherein the multiple face features can include at least an interocular distance, a face width ratio, a facial contour shape, a philtrum position, and a horizontal position relationship between ears and eyes;

[0100] Determining a feature matching order based on the multiple face features of the registered user corresponding to the electronic class board;

[0101] Grouping the registered user corresponding to the electronic class board based on the face features of the registered user corresponding to the electronic class board and the feature matching order, and determining multiple user groups corresponding to the electronic class board;

[0102] Collecting a face image of a person to be identified, and extracting multiple face features of the person to be identified;

[0103] Generating an attendance record of a user based on the multiple face features of the person to be identified, the feature matching order, and the multiple user groups corresponding to the electronic class board.

[0104] In some embodiments, determining a feature matching order based on the multiple face features of the registered user corresponding to the electronic class board includes:

[0105] Determining a pair of multiple face features, wherein the pair of face features includes two face features;

[0106] determine a feature difference value of each face feature pair based on the plurality of face features of the registered user corresponding to the electronic class list;

[0107] determine a feature matching order based on the feature difference value of each face feature pair.

[0108] Specifically, two features are randomly selected from the plurality of face features of the registered user of the electronic class list to form a pair. For example, the face features of the registered user include eye size, nose shape, mouth curvature, and eyebrow density, and then a plurality of face feature pairs can be formed, such as (eye size, nose shape), (eye size, mouth curvature), (nose shape, mouth curvature), and the like.

[0109] For each face feature, the variance of the value of the face feature of the registered user corresponding to the electronic class list can be calculated, and the average of the variances of the values of the two face features included in the face feature pair is taken as the feature difference value of the face feature pair.

[0110] The greater the feature difference value of the face feature pair, the higher the feature matching order.

[0111] Select the top-ranked plurality of face feature pairs as key face feature pairs, for example, select the top three face feature pairs, calculate the similarity of the registered users corresponding to any two electronic class lists in the key face feature pairs, and group the registered users corresponding to the electronic class lists according to the similarity of the registered users corresponding to any two electronic class lists in the key face feature pairs by using the K-means clustering algorithm to determine a plurality of user groups corresponding to the electronic class lists, so as to ensure that the users in each user group have a certain similarity in the key face feature pairs, and there are obvious differences between different user groups in the key face feature pairs.

[0112] According to the feature with the highest priority in the feature matching order, the feature of the to-be-identified person is compared with the feature range of each user group to preliminarily screen out the user groups that are likely to be matched. For example, if the feature with the highest priority is eye size, and the eye size of the to-be-identified person belongs to a certain range, then the user group with the eye size in the range is taken as the candidate group.

[0113] In the preliminarily screened user groups, other features are used in sequence according to the feature matching order for fine matching. Each feature of the to-be-identified person is compared with the corresponding feature of the registered user in the user group, and the similarity is calculated. According to a preset similarity threshold (for example, 0.6), it is judged whether the to-be-identified person is successfully matched with a certain registered user in the user group.

[0114] If the matching is successful, the identity information and the attendance time of the to-be-identified person are recorded, and an attendance record is generated; if the matching is unsuccessful, the person is marked as an unidentified person, and relevant information is recorded for subsequent processing.

[0115] In some embodiments, the electronic class board adjusts the operating parameters of the electronic teaching device, including:

[0116] Collecting real-time environmental characteristics;

[0117] Obtaining the environmental characteristics of the historical teaching time period corresponding to the teacher;

[0118] Based on the real-time environmental characteristics and the environmental characteristics of the historical teaching time period corresponding to the teacher, it is determined whether there is a similar historical teaching time period. If so, the operating parameters of the electronic teaching device in the similar historical teaching time period are retrieved, the operating parameters of the electronic teaching device are adjusted, and if not, based on the operating parameters of the electronic teaching device in the historical teaching time period corresponding to the teacher, a similar teacher is determined, based on the environmental characteristics of the historical teaching time period corresponding to the similar teacher, a similar historical teaching time period is determined, based on the operating parameters of the electronic teaching device in the similar historical teaching time period are retrieved, the operating parameters of the electronic teaching device are adjusted.

[0119] Specifically, the electronic class board uses various sensors such as temperature sensors, humidity sensors, light sensors, and noise sensors to collect temperature, humidity, and light intensity in the current teaching environment in real time. Real-time environmental characteristics are a direct basis for determining the current teaching environment, providing basic data for subsequent comparison with historical environmental characteristics to determine similar teaching time periods, and ensuring that the adjusted device operating parameters can adapt to the current actual environment.

[0120] From the database of the electronic class board or the related teaching management system, the environmental characteristic data recorded by the sensors in each teaching time period in the past by the teacher is extracted. These data are stored in categories according to teaching time, facilitating query and call.

[0121] Similarity algorithms such as Euclidean distance algorithm and cosine similarity algorithm are used to calculate the real-time environmental characteristic vector and the historical teaching time period environmental characteristic vector, and the similarity value is obtained. The calculated similarity value is compared with the preset similarity threshold value. If the similarity value is greater than or equal to the threshold value, it is determined that there is a similar historical teaching time period; if it is less than the threshold value, it is determined that there is no similar historical teaching time period. If it is determined that there is a similar historical teaching time period, the operating parameters of the electronic teaching device (such as the brightness, contrast of the projector, the volume of the sound, the temperature and wind speed of the air conditioner, etc.) in that time period are retrieved from the database, and the current operating parameters of the electronic teaching device are adjusted according to the retrieved parameters.

[0122] When the historical teaching period of the teacher similar to the current real-time environment cannot be directly found, the similar teaching period is indirectly determined by finding similar teachers and using the historical data of the similar teachers, the data reference range is expanded, and the possibility of finding suitable parameters is improved. The running parameters of the electronic teaching equipment in the historical teaching period of the teacher are analyzed, such as the preferred settings of the projector brightness, the sound volume and the like. Then, the historical equipment running parameters of other teachers are compared, and the similarity between the parameters is calculated. The teacher with high similarity is selected as the similar teacher. The environment feature data of the historical teaching period corresponding to the similar teacher is obtained, and the similarity algorithm is used again to compare and calculate the real-time environment features, and the historical teaching period with the highest similarity is found out, which is determined as the similar historical teaching period. The running parameters of the electronic teaching equipment in the determined similar historical teaching period are retrieved from the database, and the running parameters of the current electronic teaching equipment are adjusted accordingly. When there is no direct matching historical data, the running parameters of the electronic teaching equipment can also be adapted to the current teaching environment through the data of the similar teachers and the similar period, so as to ensure the smooth progress of the teaching activities.

[0123] The electronic class board is also used for receiving and displaying the to-be-pushed information.

[0124] Figure 3 is a flowchart of a multifunctional electronic class board method based on face recognition shown in an embodiment of the present application, as shown in Figure 3 The multifunctional electronic class board method based on face recognition can include the following steps:

[0125] Based on the historical attendance data of the electronic class board and the multi-target attendance optimization function, a dynamic attendance scheme is generated, wherein the dynamic attendance scheme includes the dynamic attendance time of each electronic class board and the attendance auxiliary electronic class board of each electronic class board;

[0126] The electronic class board performs user attendance according to the dynamic attendance scheme, and generates the attendance record of the user based on the face recognition technology, wherein the user is a teacher or a student;

[0127] In response to identifying the teacher, the electronic class board adjusts the running parameters of the electronic teaching equipment, wherein the electronic teaching equipment at least includes a motorized podium, a podium lighting device and a podium temperature control device;

[0128] The to-be-pushed information is obtained, and based on the historical activity information of the class corresponding to the electronic class board, the target electronic class board of the to-be-pushed information is determined, and the to-be-pushed information is sent to the target electronic class board;

[0129] The electronic class board receives and displays the to-be-pushed information.

[0130] A multifunctional electronic class list method based on face recognition can be applied to the multifunctional electronic class list system based on face recognition described above, which will not be repeated here.

[0131] Finally, it should be understood that the embodiments described herein are merely for the purpose of illustrating the principles of the embodiments described herein. Other variations can also be within the scope of the embodiments described herein. Thus, for example, alternative configurations of the embodiments described herein can be considered as being within the teachings of the embodiments described herein. Accordingly, the embodiments described herein are not limited to the embodiments explicitly described and illustrated herein.

Claims

1. A multifunctional electronic class sign system based on facial recognition, characterized in that, include: The electronic class sign module includes multiple electronic class signs, with one electronic class sign installed in each classroom. The electronic class sign is used to generate attendance records for users, such as teachers or students, based on facial recognition technology. The electronic class sign is also used to adjust the operating parameters of the electronic teaching equipment in response to teacher recognition. The electronic teaching equipment includes at least an electric lectern, a lectern lighting device, and a lectern temperature control device. The central control module is used to generate a dynamic attendance scheme based on the historical attendance data of the electronic class card and a multi-objective attendance optimization function. The dynamic attendance scheme includes the dynamic attendance time of each electronic class card and the attendance auxiliary electronic class card of each electronic class card. It is also used to receive and store the attendance records of users generated by the electronic class card, and to obtain the information to be pushed. Based on the historical activity information of the class corresponding to the electronic class card, it determines the target electronic class card to which the information to be pushed is to be sent and sends the information to be pushed to the target electronic class card. The electronic class sign is also used to track user attendance according to a dynamic attendance scheme; The electronic class sign is also used to receive and display information to be pushed; The central control module generates a dynamic attendance scheme based on historical attendance data from the electronic class sign and a multi-objective attendance optimization function, including: For each electronic class sign, based on the historical attendance data of the electronic class sign, determine the single attendance computing power requirement of the class corresponding to the electronic class sign and the historical attendance anomalies for multiple attendance time periods; Based on the single attendance computing power requirement of each class corresponding to each electronic class card, the historical attendance anomalies of multiple attendance time periods, and the multi-objective attendance optimization function, a dynamic attendance scheme is generated. The multi-objective attendance optimization function is related to at least the computing power load balance of multiple electronic class signs, the shortest attendance time required for each electronic class sign, and the historical attendance anomalies of the dynamic attendance time for each electronic class sign.

2. The multifunctional electronic class sign system based on face recognition according to claim 1, characterized in that, Based on the single attendance computing power requirement of each class corresponding to each electronic class card, historical attendance anomalies across multiple attendance time periods, and a multi-objective attendance optimization function, a dynamic attendance scheme is generated, including: For each electronic class sign, candidate attendance time periods are determined based on historical attendance anomalies from multiple attendance time periods. For any two electronic class signs, calculate the complementary attendance parameters of the two electronic class signs based on the historical attendance anomalies of multiple attendance time periods of the two electronic class signs. Establish a set of constraints, which includes at least the sampling range constraint of dynamic attendance time and the computing power load constraint of a single electronic class card. Based on the constraint set and the complementary attendance parameters of any two electronic class cards, multiple candidate dynamic attendance schemes are generated. For each candidate dynamic attendance scheme, calculate the multi-objective attendance optimization function value of the candidate dynamic attendance scheme; A dynamic attendance scheme is generated based on the multi-objective attendance optimization function value of each candidate dynamic attendance scheme.

3. The multifunctional electronic class sign system based on face recognition according to claim 1, characterized in that, Based on historical attendance anomalies across multiple attendance time periods, candidate attendance time periods corresponding to the electronic class sign are determined, including: For any two electronic class signs, the attendance similarity between the two electronic class signs is determined based on the historical attendance anomalies of multiple attendance time periods of the two electronic class signs. For each electronic class sign, determine the similar electronic class signs based on the attendance similarity between any two electronic class signs; For each electronic class sign, the attendance frequency of multiple attendance time periods of the electronic class sign is used to determine the attendance time period to be corrected. Based on the historical attendance anomalies of similar electronic class signs for the attendance period to be corrected, the historical attendance anomalies of the electronic class signs for the attendance period to be corrected are corrected. Based on the corrected historical attendance anomalies for multiple attendance periods, candidate attendance periods corresponding to the electronic class sign are determined.

4. A multifunctional electronic class sign system based on face recognition according to any one of claims 1-3, characterized in that, The electronic class sign uses facial recognition technology to generate user attendance records, including: Extract multiple facial features of registered users corresponding to the electronic class sign; Based on the facial features of multiple registered users corresponding to the electronic class card, the feature matching order is determined. Based on the facial features and feature matching order of the registered users corresponding to the electronic class card, the registered users corresponding to the electronic class card are grouped to determine multiple user groups corresponding to the electronic class card. Collect facial images of the person to be identified and extract multiple facial features of the person to be identified; Based on the multiple facial features of the person to be identified, the feature matching order, and the multiple user groups corresponding to the electronic class card, the user's attendance record is generated.

5. A multifunctional electronic class sign system based on face recognition according to claim 4, characterized in that, Based on the facial features of multiple registered users corresponding to the electronic class sign, the feature matching order is determined, including: Identify multiple face feature pairs, where each face feature pair includes two face features; Based on the multiple facial features of registered users corresponding to the electronic class card, determine the feature difference value of each facial feature pair; The feature matching order is determined based on the feature difference values ​​of each facial feature pair.

6. A multifunctional electronic class sign system based on face recognition according to any one of claims 1-3, characterized in that, The electronic class sign adjusts the operating parameters of the electronic teaching equipment, including: Collect real-time environmental characteristics; Obtain the environmental characteristics of the corresponding historical teaching period for the teacher; Based on real-time environmental characteristics and the environmental characteristics of the teacher's corresponding historical teaching time period, determine whether there are similar historical teaching time periods. If so, retrieve the operating parameters of the electronic teaching equipment for similar historical teaching time periods and adjust the operating parameters of the electronic teaching equipment. If not, based on the operating parameters of the electronic teaching equipment for the teacher's corresponding historical teaching time period, identify similar teachers, determine similar historical teaching time periods based on the environmental characteristics of the similar teachers' corresponding historical teaching time periods, and adjust the operating parameters of the electronic teaching equipment based on the retrieved operating parameters of the electronic teaching equipment for similar historical teaching time periods.

7. A multifunctional electronic class sign system based on face recognition according to any one of claims 1-3, characterized in that, The process of determining the target electronic class board for which information is to be pushed, based on the historical activity information of the class corresponding to the electronic class board, includes: Based on the historical activity information of the class corresponding to the electronic class card, a class profile is established for the class corresponding to the electronic class card; Based on the class profile corresponding to each electronic class sign, the target electronic class sign for the information to be pushed is determined.

8. A method for creating a multifunctional electronic class sign based on facial recognition, characterized in that, The multifunctional electronic class sign system based on face recognition, as described in claim 1, comprises: Based on the historical attendance data of the electronic class cards and the multi-objective attendance optimization function, a dynamic attendance scheme is generated, wherein the dynamic attendance scheme includes the dynamic attendance time of each electronic class card and the attendance auxiliary electronic class card of each electronic class card; The electronic class sign uses a dynamic attendance scheme to track user attendance and generates attendance records based on facial recognition technology. The user can be a teacher or a student. In response to teacher identification, the electronic class sign adjusts the operating parameters of the electronic teaching equipment, which includes at least an electric lectern, lectern lighting equipment, and lectern temperature control equipment. Obtain the information to be pushed, and based on the historical activity information of the class corresponding to the electronic class board, determine the target electronic class board to which the information to be pushed, and send the information to the target electronic class board; The electronic class sign receives and displays information to be pushed.

Images