Real-time people flow statistics method and system based on Internet of Things (IoT) information collection

Abstract

This application proposes a method and system for real-time people flow statistics based on Internet of Things (IoT) information collection. The method involves using a camera to capture classroom monitoring images, processing the classroom monitoring images using a smart classroom information board, saving and uploading the processed results to a local data center, which then generates real-time people flow statistics and displays them to teachers and students via a visualization interface. The system corresponding to this method includes a monitoring module, a processing module, and a display module. The method and system can help students quickly find vacant classrooms, increase classroom resource utilization, standardize student punctuality, and assist in teacher attendance statistics. Furthermore, it can improve monitoring resource utilization, avoid unnecessary monitoring, and reduce the burden on cameras and devices storing monitoring resources.

Description

[Technical Field]

[0001] This application proposes a method and system for real-time people flow statistics based on Internet of Things (IoT) information collection, which belongs to the application field of Internet of Things (IoT) technology. [Background technology]

[0002] In recent years, with the development of science and technology and the Internet, many industries have had great demands for people flow statistics, such as places with high traffic volumes, such as train stations, bus stops, subway stations, and department store entrances and exits. People flow statistics systems can conveniently and reliably collect real-time data on people flow in various locations without affecting the public. By incorporating image analysis technology, customer flow dynamics can be clearly and quickly grasped, providing data support for decision makers to make timely decisions. The main advantage of using image processing to complete a people flow statistics system is that image signals are highly intuitive and easy for people to understand.

[0003] For university students, real-time classroom attendance is crucial. Currently, students can access the internet through their devices anytime, anywhere. However, university libraries and study rooms generally lack sufficient resources, making empty classrooms the primary study location for many students. However, due to the uncertainty surrounding classroom resource usage, students often spend a lot of time searching for empty classrooms when classes are not in session. For university teachers, attendance statistics are an effective way to record student learning progress and provide a basis for evaluating grades during normal times. Traditional methods of counting attendance are time-consuming and laborious, and often involve alternate signatures, making the data unreliable. This wastes valuable class time and places a strain on teachers. Summary of the Invention [Means for solving the problem]

[0004] This application provides a real-time people flow statistics method and system based on Internet of Things (IoT) information collection to solve the above problems.

[0005] The present application provides a real-time people flow statistics method based on Internet of Things (IoT) information collection, the method comprising: Obtaining a classroom monitoring screen using a camera; Processing the classroom monitoring screen through the smart classroom information board, saving the processed results and uploading them to a local data center; The local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface.

[0006] Furthermore, capturing classroom surveillance images using cameras is When the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that a person has entered the classroom, the corresponding smart classroom guide board in the classroom checks whether there is a class within the preset time period; When the smart classroom information board finds that there is a class in the corresponding classroom within the preset time period, it sends a control signal to the cameras in the classroom and at the classroom entrance / exit using ZigBee communication standard, and after receiving the control signal, the cameras in the classroom and at the classroom entrance / exit start monitoring and send the monitoring screen to the corresponding smart classroom information board; When the smart classroom information board detects that there is no class in the corresponding classroom within the preset time period, it sends a control signal to the camera at the entrance / exit of the classroom using the ZigBee communication standard, and after receiving the control signal, the camera at the entrance / exit of the classroom starts monitoring and transmits the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model.

[0007] Furthermore, when the smart classroom information board finds that there is no class in the corresponding classroom within the preset time period, it sends a control signal to the camera at the entrance / exit of the classroom using the ZigBee communication standard. After receiving the control signal, the camera at the entrance / exit of the classroom starts monitoring and sends the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model, which is expressed by the following formula:

[0008]

number

[0009] where F is the frequency, FD is the difference between the previous and next frames of the surveillance video, Presence is a binary variable indicating the presence of personnel, which is 1 if the infrared sensor in the classroom detects that a person has entered the classroom, and 0 otherwise, α is a weight for balancing the frame difference and the structural similarity index, k1 is a constant for adjusting the frequency of surveillance transmission, and the value of k1 ranges from [15, 30], S is the structural similarity index, where

[0010]

number

[0011] where x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually

[0012]

number

[0013]

number

[0014] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x,y) is a similarity function to measure the change in light-shadow.

[0015]

number

[0016] where L X and Ly represent the light intensities of the two images, respectively, and L max represents the maximum light intensity.

[0017] Furthermore, the smart classroom information board processes the classroom monitoring screen, stores the processed results, and uploads them to the local data center. After receiving the surveillance images from inside the classroom and at the entrance and exit of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrance and exit of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people. If it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding electronic device of the teacher, so that the teacher can further check the attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; The smart classroom information board receives the classroom entrance / exit monitoring screen, first preprocesses the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, displaying it through a visualization interface and uploading it to the local data center; During data preprocessing, we optimize through the following model:

[0018]

number

[0019] Here, the following equation (7) represents the new morphology process, A represents the input image, B represents the structuring element for defining the shape and size of the dilation and corrosion process, the following equation (8) represents the result after dilation obtained by performing dilation processing on image A, and the following equation (9) represents the output image obtained by performing corrosion processing on the result of the following equation (10).

[0020]

number

[0021]

number

[0022]

number

[0023]

number

[0024] Furthermore, the local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface; The local data center processes the attendance information of the classrooms where classes are held, stores it according to the attendance rate of the corresponding class, and displays the processed attendance rate of each classroom through a visualization interface; The local data center processes the number of people entering and leaving the vacant classrooms, arranges them according to the display format of the smart classroom guide board, and forms the display data of the general smart classroom guide board and the display data of the smart classroom guide board of each building and each floor; The display data of the smart classroom guide board is displayed on the smart classroom guide board, and the display data of the smart classroom guide board on each floor of each building is transmitted to the corresponding smart classroom guide board devices on each floor of each building through a local area network communication method, so that students can register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms.

[0025] Furthermore, the present application proposes a real-time people flow statistics system based on Internet of Things (IoT) information collection, the system comprising: a monitoring module for capturing a monitoring screen of the classroom using a camera; a processing module for processing the classroom monitoring screen through the smart classroom information board, storing the processed results, and uploading them to a local data center; The local data center includes a display module for generating real-time people flow statistics results and displaying them to teachers and students through a visualization interface.

[0026] Furthermore, the monitoring module a classroom entry detection module for detecting whether there is a class within a preset time period when the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that there is a person entering the classroom; a classroom and classroom entrance camera start module for sending a control signal to the cameras in the classroom and at the classroom entrance using ZigBee communication standard when the smart classroom information board finds that a class is held in the corresponding classroom within a preset time period, and the cameras in the classroom and at the classroom entrance start monitoring after receiving the control signal and transmit the monitoring screen to the corresponding smart classroom information board; The smart classroom information board includes a classroom entrance camera starting module for sending a control signal to the classroom entrance camera using ZigBee communication standard when it detects that there is no class in the corresponding classroom within the preset time period, and the classroom entrance camera starts monitoring after receiving the control signal and transmits the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model.

[0027] Furthermore, the classroom entrance / exit camera starting module includes a frequency calculation model module;

[0028]

number

[0029] where F is the frequency, FD is the dissimilarity between the previous and next frames of the surveillance video, Presence is a binary variable indicating the presence of personnel, which is 1 if the infrared sensor in the classroom detects that a person has entered the classroom, and 0 otherwise, α is a weight for balancing the frame dissimilarity and the structural similarity index, the value of α ranges from (0,1), k1 is a constant for adjusting the frequency of surveillance transmission, the value of k1 ranges from [15,30], S is the structural similarity index,

[0030]

number

[0031] where x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually

[0032]

number

[0033]

number

[0034] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x,y) is a similarity function to measure the change in light-shadow.

[0035]

number

[0036] where L X and Ly represent the light intensities of the two images, respectively, and L max represents the maximum light intensity.

[0037] Furthermore, the processing module After receiving the surveillance images from inside the classroom and at the entrances and exits of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrances and exits of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people; if it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding teacher's electronic device to further confirm the teacher's attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; and a class response processing module. The smart classroom information board includes an empty classroom response processing module that receives the classroom entrance / exit monitoring screen, first preprocesses the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, displaying it through a visualization interface, and uploading it to the local data center. During data preprocessing, we optimize through the following model:

[0038]

number

[0039] Here, the following equation (Equation 17) represents the new morphology process, A represents the input image, B represents the structuring element for defining the shape and size of the dilation and corrosion process, the following equation (Equation 18) represents the result after dilation obtained by performing dilation processing on image A, and the following equation (Equation 19) represents the output image obtained by performing corrosion processing on the result of the following equation (Equation 20).

[0040]

number

[0041]

number

[0042]

number

[0043]

number

[0044] Furthermore, the display module The local data center processes the attendance information of the classrooms where classes are held, and then stores the information according to the attendance rate of the corresponding class. A corresponding class attendance rate display module is provided to display the processed attendance rate of each classroom through a visualization interface. a local data center processing the number of people entering and leaving the vacant classrooms, arranging them according to the display format of the smart classroom guide board, and forming the display data of the general smart classroom guide board and the display data of the smart classroom guide board of each building and each floor; and a transfer module for displaying the display data of the smart classroom guide board on the smart classroom guide board, and transmitting the display data of the smart classroom guide board on each floor of each building to the corresponding smart classroom guide board devices on each floor of each building through a local area network communication method, allowing students to register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms. [Effects of the Invention]

[0045] This application has the following beneficial effects: it can help students quickly find vacant classrooms, improve classroom resource utilization, standardize student attendance rates, and assist in teacher attendance statistics. Target detection network technology can be used to process and analyze classroom surveillance video footage, build a visualization interface, and provide relevant classroom data information. Monitoring can be automatically started or stopped based on classroom conditions, reducing unnecessary monitoring time and saving resources. By combining infrared sensors with smart classroom information boards, the system can intelligently determine whether a classroom is having classes during a specific time period. Automated monitoring switches can avoid unnecessary monitoring, thereby saving energy and reducing the burden on equipment. Real-time monitoring screens can be provided as needed, and monitoring frequencies can be adjusted as needed, effectively managing monitoring data according to specific needs. Since the target of monitoring is people, infrared sensors can start monitoring when they detect a person entering the classroom, reducing the burden on equipment and the workload of reviewing surveillance video. Furthermore, when monitoring begins, the monitoring needs differ depending on whether there are classes in the classroom or not. When classes are in session, the classroom needs to check attendance and monitor the number of students in class to prevent students from skipping classes. When there are no classes, the classroom mainly needs to monitor the number of people in the classroom and whether there are currently vacant seats in the classroom, and there is no need to monitor the situation inside the classroom. The system can automatically switch monitoring modes according to the school timetable and school schedule, thereby meeting the monitoring needs of different scenarios. The intelligent monitoring system allows schools to make better use of monitoring resources, avoid unnecessary monitoring, and reduce the burden on cameras and devices that store monitoring resources. [Brief explanation of the drawings]

[0046] [Figure 1] 1 is a flowchart of a method for real-time people flow statistics based on Internet of Things (IoT) information collection according to the present application; DETAILED DESCRIPTION OF THE INVENTION

[0047] Hereinafter, preferred embodiments of the present application will be described with reference to the accompanying drawings. However, it should be understood that the preferred embodiments described here are used only for the description and explanation of the present application and are not used to limit the present application.

[0048] As one embodiment of the present application, there is provided a real-time people flow statistics method based on Internet of Things (IoT) information collection, the method comprising: Obtaining a classroom monitoring screen using a camera; Processing the classroom monitoring screen through the smart classroom information board, saving the processed results and uploading them to a local data center; The local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface.

[0049] The working principle and effect of the above technical means are as follows: cameras are installed in classrooms and at classroom entrances and exits to capture real-time monitoring images containing information on classroom activities, personnel, etc. The smart classroom information board collects and processes the monitoring images, and uses computer vision technology and other image processing methods to detect and track personnel. The processed results are saved and uploaded to a local data center via a network connection for further data storage and processing. The local data center receives the processed monitoring data, statistics on personnel entering and exiting, and generates real-time people flow statistics. Finally, the real-time people flow statistics are displayed to teachers and students through a visualization interface, allowing them to clearly understand the dynamic situation of personnel in the classroom.

[0050] As one embodiment of the present application, there is provided a real-time people flow statistics method based on Internet of Things (IoT) information collection, the method including using a camera to capture a classroom monitoring screen, When the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that a person has entered the classroom, the corresponding smart classroom guide board in the classroom checks whether there is a class within the preset time period; When the smart classroom information board finds that there is a class in the corresponding classroom within the preset time period, it sends a control signal to the cameras in the classroom and at the classroom entrance / exit using ZigBee communication standard, and after receiving the control signal, the cameras in the classroom and at the classroom entrance / exit start monitoring and send the monitoring screen to the corresponding smart classroom information board; When the smart classroom information board detects that there is no class in the corresponding classroom within the preset time period, it sends a control signal to the camera at the entrance / exit of the classroom using the ZigBee communication standard, and after receiving the control signal, the camera at the entrance / exit of the classroom starts monitoring and transmits the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model.

[0051] The preset time period can be set by a person, and since one frame is generally 45 minutes, the preset time is set to 45 minutes as a reference.

[0052] The working principle of the above technical means is that when the infrared sensor inside the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that someone has entered the classroom, the smart classroom information board will check whether there are any classes within the preset time period. It will send a control signal based on the course status. If it finds that there are classes in the corresponding classroom within the preset time period, the smart classroom information board will send control signals to the cameras inside the classroom and at the classroom entrance using ZigBee communication, and the cameras will start monitoring and transmit the monitoring images to the corresponding smart classroom information board. If it finds that there are no classes in the corresponding classroom within the preset time period, the smart classroom information board will send a control signal to the cameras at the classroom entrance using ZigBee communication, and the cameras will start monitoring and transmit the monitoring images to the smart classroom information board at the preset monitoring frequency.

[0053] The effects of the above technical measures are as follows: The system can automatically start or stop monitoring based on the situation in the classroom, reducing unnecessary monitoring time and saving resources. By combining infrared sensors with smart classroom information boards, the system can intelligently determine whether a classroom is in session at a specific time. Automated monitoring switches can avoid unnecessary monitoring, thereby saving energy and reducing the burden on equipment. Real-time monitoring screens can be provided as needed, and monitoring frequencies can be adjusted as needed, effectively managing monitoring data according to specific needs. Because the target of monitoring is people, starting monitoring when the infrared sensor detects a person entering the classroom not only reduces the burden on equipment, but also the workload of reviewing surveillance video. Furthermore, when monitoring begins, the monitoring needs differ depending on whether there are classes in the classroom or not. When classes are in session, the classroom needs to check attendance and monitor the number of students in class to prevent students from skipping classes. When there are no classes, the classroom mainly needs to monitor the number of people in the classroom and whether there are currently vacant seats in the classroom, and there is no need to monitor the situation inside the classroom. The system can automatically switch monitoring modes according to the school timetable and school schedule, thereby meeting the monitoring needs of different scenarios. The intelligent monitoring system allows schools to make better use of monitoring resources, avoid unnecessary monitoring, and reduce the burden on cameras and devices that store monitoring resources.

[0054] One embodiment of the present application is a real-time people flow statistics method based on Internet of Things (IoT) information collection, characterized as follows: when the smart classroom information board detects that there is no class in the corresponding classroom within a preset time period, it sends a control signal to the camera at the entrance / exit of the classroom using ZigBee communication standard, and the camera at the entrance / exit of the classroom starts monitoring after receiving the control signal and sends the monitoring screen to the corresponding smart classroom information board at a monitoring frequency obtained by a frequency calculation model, and the frequency calculation model is expressed by the following formula:

[0055]

number

[0056] where F is the frequency, FD is the difference between the previous and next frames of the surveillance video,

[0057]

number

[0058] Here, m and n represent the height and width of the image, respectively, and A(i,j) and B(i,j) represent the pixel values ​​at locations (i,j) in images A and B, respectively. Presence is a binary variable indicating the presence of personnel. It is set to 1 if the classroom infrared sensor detects that a person has entered the classroom, and 0 otherwise. α is a weight for balancing the frame dissimilarity and the structural similarity index. The value of α ranges from (0,1). k1 is a constant for adjusting the frequency of monitoring transmission. The value of k1 ranges from [15,30]. S is the structural similarity index.

[0059]

number

[0060] where x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually

[0061]

number

[0062]

number

[0063] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x,y) is a similarity function to measure the change in light-shadow.

[0064]

number

[0065] where L X and Ly represent the light intensities of the two images, respectively, and L max represents the maximum light intensity.

[0066] The operating principle and effect of the above technical means are that the model allows the system to dynamically adjust the transmission frequency of the monitoring image according to the actual situation, avoiding waste of bandwidth and storage resources when there is no change in the number of people in the classroom. The model also reduces the transmission frequency of the monitoring image when no one is present or there is little change in the image, thereby saving network transmission and storage resources. By configuring the weight terms and parameters in the model, the monitoring system becomes more intelligent and adaptive, dynamically adjusting the monitoring frequency based on the actual scene and better meeting actual needs. The calculation of F combines frame dissimilarity, structural similarity, and personnel appearance status, allowing the monitoring frequency to be dynamically adjusted according to changes in the image and personnel appearance status. The calculation of S takes into account the structural similarity and light and shadow similarity of the image, making the monitoring image frequency closer to the actual changing situation and avoiding overly frequent or sparse transmission of monitoring images. I(x,y) is a similarity function used to measure changes in light and shadow. The model can more comprehensively consider the effect of light intensity on the monitoring frequency. Changes in light and shadow can also cause differences between two images over time. However, since no human presence exists, the monitoring frequency should not be increased. The model also takes light and shadow changes into account in the monitoring frequency, further improving the accuracy of the monitoring frequency model. The model intelligently adjusts the monitoring frequency according to scene changes, effectively avoiding the transmission and storage of invalid monitoring images. Dynamically adjusting the monitoring frequency avoids resource waste and improves the effectiveness and practicality of the monitoring system. Each formula comprehensively considers factors such as inter-frame differences, structural similarity, and changes in light and shadow, allowing the monitoring frequency to more comprehensively and accurately reflect changes in the monitoring image. I(x,y) takes into account changes in light intensity and matches and standardizes the brightness changes at corresponding image points. Since the value range is [0,1], the differences between them are easy to understand and are not affected by the absolute value of light intensity. The design of the I(x,y) model applies the similarity function to image matching under different lighting conditions, including images captured under different lighting conditions.Since I(x,y) takes into account the maximum light intensity, it can maintain a certain robustness even under different lighting conditions and can better reflect the similarity between images. This model makes the similarity calculation between images more universal and robust.

[0067] As one embodiment of the present application, a real-time people flow statistics method based on Internet of Things (IoT) information collection, wherein a classroom monitoring screen is processed by an intelligent classroom information board, and the processed results are stored and uploaded to a local data center: After receiving the surveillance images from inside the classroom and at the entrance and exit of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrance and exit of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people. If it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding electronic device of the teacher, so that the teacher can further check the attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; The smart classroom information board receives the classroom entrance / exit monitoring screen, first preprocesses the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, displaying it through a visualization interface and uploading it to the local data center; During data preprocessing, we optimize through the following model:

[0068]

number

[0069] Here, the following equation (Equation 28) represents the new morphology process, A represents the input image, B represents the structuring element for defining the shape and size of the dilation and corrosion process, the following equation (Equation 29) represents the result after dilation processing of image A, and the following equation (Equation 30) represents the output image obtained by performing corrosion processing on the result of the following equation (Equation 31).

[0070]

number

[0071]

number

[0072]

number

[0073]

number

[0074] The operating principle of the above technology is that after the smart classroom information board receives the surveillance image from the classroom entrance, it checks the pre-stored facial data of the class to confirm whether the person entering the classroom is the expected person. It then continuously measures the number of people in the classroom surveillance video through the YOLOv5s model to obtain real-time headcount data. If a facial mismatch or the number of monitored faces does not match the actual number of people monitored, the system sends information to the corresponding teacher's electronic device to confirm the teacher's attendance. The information confirmed by the teacher is uploaded to a local data center, where it performs data preprocessing on the surveillance image at the classroom entrance. The YOLOv5s model is used to obtain the number of people entering and exiting the classroom, which is then displayed in a visualization interface and uploaded to the local data center. The data is preprocessed using the morphological processing model Θ(A,B), which first dilates the image and then corrodes the result, thereby removing small noise while preserving the overall shape of the target.

[0075] The effect of the above technology is that through face matching and real-time headcount monitoring, the system can immediately detect whether the person entering the classroom matches expectations, thereby improving the level of campus safety management. If an abnormality occurs, the system automatically sends information to the corresponding teacher, who can check attendance status and enhance the accuracy of student attendance management. The visualization interface intuitively displays the number of people entering and exiting the classroom, providing real-time and effective data support to school management departments. Using morphological processing models to preprocess data helps improve image quality and improve the accuracy of subsequent face matching and headcount monitoring. It also simplifies the subsequent model processing flow, removes small noise in the image, and connects discontinuous parts, allowing for more accurate identification of target contours. Dilation and erosion can be combined to achieve more complex morphological transformations. A combination of opening and closing operations can achieve better image processing results.

[0076] As one embodiment of the present application, a real-time people flow statistics method based on Internet of Things (IoT) information collection, wherein the local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface, The local data center processes the attendance information of the classrooms where classes are held, stores it according to the attendance rate of the corresponding class, and displays the processed attendance rate of each classroom through a visualization interface; The local data center processes the number of people entering and leaving the vacant classrooms, arranges them according to the display format of the smart classroom guide board, and forms the display data of the general smart classroom guide board and the display data of the smart classroom guide board of each building and each floor; The display data of the smart classroom guide board is displayed on the smart classroom guide board, and the display data of the smart classroom guide board on each floor of each building is transmitted to the corresponding smart classroom guide board devices on each floor of each building through a local area network communication method, so that students can register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms.

[0077] The working principle of the above technical means is that the local data center processes the attendance information of classrooms where classes are held, stores it according to the attendance rate of the corresponding class, and displays the processed attendance rate of each classroom to relevant administrators through a visualization interface. The information on available classrooms is processed and arranged according to the display format of the smart classroom signboard, forming display data for the global smart classroom signboard and display data for the smart classroom signboard on each floor of each building. The display data for the global smart classroom signboard is displayed on the global smart classroom signboard, and the display data for the smart classroom signboard on each floor of each building is transmitted to the corresponding smart classroom signboard device on each floor of each building via a local area network communication method. Students can register on the campus network to check the information on the global smart classroom signboard and obtain information on available classrooms.

[0078] The effects of the above technical measures are that, through the processing of attendance information, school administrators can clearly understand the attendance rate of each classroom and corresponding period, providing data support for educational management. Classroom attendance rates are displayed through a visualization interface, allowing administrators to immediately grasp student attendance status and providing a basis for subsequent educational management and intervention. Vacant classroom information is organized into smart classroom guide board format and transmitted to the smart classroom guide board device on the corresponding floor, helping students quickly obtain vacant classroom information. Students can easily obtain information from the smart classroom guide board via the local area network, quickly obtaining vacant classroom information and improving the utilization efficiency of campus resources. Data transmission is completed via the local area network, ensuring the security and timeliness of information.

[0079] As one embodiment of the present application, there is provided a people flow real-time statistics system based on Internet of Things (IoT) information collection, the system comprising: a monitoring module for capturing a monitoring screen of the classroom using a camera; a processing module for processing the classroom monitoring screen through the smart classroom information board, storing the processed results, and uploading them to a local data center; The local data center includes a display module for generating real-time people flow statistics results and displaying them to teachers and students through a visualization interface.

[0080] The working principle and effect of the above technical means are as follows: cameras are installed in classrooms and at classroom entrances and exits to capture real-time monitoring images containing information on classroom activities, personnel, etc. The smart classroom information board collects and processes the monitoring images, and uses computer vision technology and other image processing methods to detect and track personnel. The processed results are saved and uploaded to a local data center via a network connection for further data storage and processing. The local data center receives the processed monitoring data, statistics on personnel entering and exiting, and generates real-time people flow statistics. Finally, the real-time people flow statistics are displayed to teachers and students through a visualization interface, allowing them to clearly understand the dynamic situation of personnel in the classroom.

[0081] As one embodiment of the present application, there is provided a people flow real-time statistics system based on Internet of Things (IoT) information collection, wherein the monitoring module comprises: a classroom entry detection module for detecting whether there is a class within a preset time period when the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that there is a person entering the classroom; a classroom and classroom entrance camera start module for sending a control signal to the cameras in the classroom and at the classroom entrance using ZigBee communication standard when the smart classroom information board finds that a class is held in the corresponding classroom within a preset time period, and the cameras in the classroom and at the classroom entrance start monitoring after receiving the control signal and transmit the monitoring screen to the corresponding smart classroom information board; The smart classroom information board includes a classroom entrance camera starting module for sending a control signal to the classroom entrance camera using ZigBee communication standard when it detects that there is no class in the corresponding classroom within the preset time period, and the classroom entrance camera starts monitoring after receiving the control signal and transmits the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model.

[0082] The working principle of the above technical means is that when the infrared sensor inside the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that someone has entered the classroom, the smart classroom information board checks whether there are any classes within the preset time period. It sends a control signal based on the course status, and if it finds that there are classes in the corresponding classroom within the preset time period, the smart classroom information board uses ZigBee communication to send control signals to the cameras inside the classroom and at the classroom entrance, and the cameras start monitoring and transmit the monitoring images to the corresponding smart classroom information board. If it finds that there are no classes in the corresponding classroom within the preset time period, the smart classroom information board uses ZigBee communication to send control signals to the cameras at the classroom entrance, and the cameras start monitoring and transmit the monitoring images to the smart classroom information board at the preset monitoring frequency.

[0083] The effects of the above technical measures are as follows: the system can intelligently determine whether a classroom is in session during a specific time period, and an automated monitoring switch can avoid unnecessary monitoring, thereby saving energy and reducing the burden on equipment. It can provide real-time monitoring as needed, and adjust the monitoring frequency as needed, effectively managing monitoring data according to specific needs. Because the target of monitoring is people, monitoring can be initiated when an infrared sensor detects a person entering the classroom, not only reducing the burden on equipment but also the workload of reviewing monitoring video. At the same time, the monitoring requirements differ depending on whether a classroom is in session or not. When classes are in session, the classroom needs to check attendance and monitor the number of students in class to prevent students from skipping class. When classes are not in session, the classroom mainly monitors the number of people in the classroom and whether there are currently vacant seats, eliminating the need to monitor the classroom situation. The system can automatically switch monitoring modes according to the school timetable and school schedule, thereby meeting the monitoring needs of different situations. The intelligent monitoring system allows schools to better utilize monitoring resources, avoid unnecessary monitoring, and reduce the burden on cameras and devices storing monitoring resources.

[0084] As an embodiment of the present application, there is provided a real-time people flow statistics system based on Internet of Things (IoT) information collection, wherein the classroom entrance / exit camera initiation module includes a frequency calculation model module;

[0085]

number

[0086] where F is the frequency, FD is the dissimilarity between the previous and next frames of the surveillance video, Presence is a binary variable indicating the presence of personnel, which is 1 if the infrared sensor in the classroom detects that a person has entered the classroom, and 0 otherwise, α is a weight for balancing the frame dissimilarity and the structural similarity index, the value of α ranges from (0,1), k1 is a constant for adjusting the frequency of surveillance transmission, the value of k1 ranges from [15,30], S is the structural similarity index, where

[0087]

number

[0088] where x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually

[0089]

number

[0090]

number

[0091] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x,y) is a similarity function to measure the change in light-shadow.

[0092]

number

[0093] where L Xand Ly represent the light intensities of the two images, respectively, and L max represents the maximum light intensity.

[0094] The operating principle and effect of the above technical means are that the model allows the system to dynamically adjust the transmission frequency of the monitoring image according to the actual situation, avoiding waste of bandwidth and storage resources when there is no change in the number of people in the classroom. The model also reduces the transmission frequency of the monitoring image when no one is present or there is little change in the image, thereby saving network transmission and storage resources. The weight terms and parameter settings in the model make the monitoring system more intelligent and adaptive, dynamically adjusting the monitoring frequency based on the actual scene to better meet actual needs. The calculation of F combines frame dissimilarity, structural similarity, and personnel appearance status, allowing the monitoring frequency to be dynamically adjusted according to changes in the image and personnel appearance status. The calculation of S takes into account image structural similarity and light and shadow similarity, making the monitoring image frequency closer to the actual changing situation and avoiding overly frequent or sparse transmission of monitoring images. I(x,y) is a similarity function used to measure changes in light and shadow. The model can more comprehensively consider the effect of light intensity on the monitoring frequency. Changes in light and shadow can also cause differences between two images over time. However, since no human presence exists, the monitoring frequency should not be increased. The model also takes light and shadow changes into account in the monitoring frequency, further improving the accuracy of the monitoring frequency model. The model intelligently adjusts the monitoring frequency according to scene changes, effectively avoiding the transmission and storage of invalid monitoring images. Dynamically adjusting the monitoring frequency avoids resource waste and improves the effectiveness and practicality of the monitoring system. Each formula comprehensively considers factors such as inter-frame differences, structural similarity, and changes in light and shadow, allowing the monitoring frequency to more comprehensively and accurately reflect changes in the monitoring image. I(x,y) takes into account changes in light intensity and matches and standardizes the brightness changes at corresponding image points. Since the value range is [0,1], the differences between them are easy to understand and are not affected by the absolute value of light intensity. The design of the I(x,y) model applies the similarity function to image matching under different lighting conditions, including images captured under different lighting conditions.Since I(x,y) takes into account the maximum light intensity, it can maintain a certain robustness even under different lighting conditions and can better reflect the similarity between images. This model makes the similarity calculation between images more universal and robust.

[0095] As one embodiment of the present application, there is provided a people flow real-time statistics system based on Internet of Things (IoT) information collection, wherein the processing module comprises: After receiving the surveillance images from inside the classroom and at the entrances and exits of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrances and exits of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people; if it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding teacher's electronic device to further confirm the teacher's attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; and a class response processing module. The smart classroom information board includes an empty classroom response processing module that receives the classroom entrance / exit monitoring screen, first preprocesses the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, displaying it through a visualization interface, and uploading it to the local data center. During data preprocessing, we optimize through the following model:

[0096]

number

[0097] Here, the following equation (Equation 38) represents the new morphology process, A represents the input image, B represents the structuring element for defining the shape and size of the dilation and corrosion process, the following equation (Equation 39) represents the result after dilation obtained by performing dilation processing on image A, and the following equation (Equation 40) represents the output image obtained by performing corrosion processing on the result of the following equation (Equation 41).

[0098]

number

[0099]

number

[0100]

number

[0101]

number

[0102] The operating principle of the above technology is that after the smart classroom information board receives the surveillance image from the classroom entrance, it checks the pre-stored facial data of the class to confirm whether the person entering the classroom is the expected person. It then continuously measures the number of people in the classroom surveillance video through the YOLOv5s model to obtain real-time headcount data. If a facial mismatch or the number of monitored faces does not match the actual number of people monitored, the system sends information to the corresponding teacher's electronic device to confirm the teacher's attendance. The information confirmed by the teacher is uploaded to a local data center, where it performs data preprocessing on the surveillance image at the classroom entrance. The YOLOv5s model is used to obtain the number of people entering and exiting the classroom, which is then displayed in a visualization interface and uploaded to the local data center. The data is preprocessed using the morphological processing model Θ(A,B), which first dilates the image and then corrodes the result, thereby removing small noise while preserving the overall shape of the target.

[0103] The effect of the above technology is that through face matching and real-time headcount monitoring, the system can immediately detect whether the person entering the classroom matches expectations, thereby improving the level of campus safety management. If an abnormality occurs, the system automatically sends information to the corresponding teacher, who can check attendance status and enhance the accuracy of student attendance management. The visualization interface intuitively displays the number of people entering and exiting the classroom, providing real-time and effective data support to school management departments. Using morphological processing models to preprocess data helps improve image quality and improve the accuracy of subsequent face matching and headcount monitoring. It also simplifies the subsequent model processing flow, removes small noise in the image, and connects discontinuous parts, allowing for more accurate identification of target contours. Dilation and erosion can be combined to achieve more complex morphological transformations. A combination of opening and closing operations can achieve better image processing results.

[0104] As one embodiment of the present application, there is provided a people flow real-time statistics system based on Internet of Things (IoT) information collection, wherein the display module comprises: The local data center processes the attendance information of the classrooms where classes are held, and then stores the information according to the attendance rate of the corresponding class. A corresponding class attendance rate display module is provided to display the processed attendance rate of each classroom through a visualization interface. a local data center processing the number of people entering and leaving the vacant classrooms, arranging them according to the display format of the smart classroom guide board, and forming the display data of the general smart classroom guide board and the display data of the smart classroom guide board of each building and each floor; and a transfer module for displaying the display data of the smart classroom guide board on the smart classroom guide board, and transmitting the display data of the smart classroom guide board on each floor of each building to the corresponding smart classroom guide board devices on each floor of each building through a local area network communication method, allowing students to register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms.

[0105] The effects of the above technical measures are that, through the processing of attendance information, school administrators can clearly understand the attendance rate of each classroom and corresponding period, providing data support for educational management. Classroom attendance rates are displayed through a visualization interface, allowing administrators to immediately grasp student attendance status and providing a basis for subsequent educational management and intervention. Vacant classroom information is organized into smart classroom guide board format and transmitted to the smart classroom guide board device on the corresponding floor, helping students quickly obtain vacant classroom information. Students can easily obtain information from the smart classroom guide board via the local area network, quickly obtaining vacant classroom information and improving the utilization efficiency of campus resources. Data transmission is completed via the local area network, ensuring the security and timeliness of information.

[0106] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application also intends to include these modifications and variations.

Claims

1. A real-time people flow statistics method based on Internet of Things (IoT) information collection, comprising: The method comprises: Obtaining a classroom monitoring screen using a camera; Processing the classroom monitoring screen through the smart classroom information board, saving the processed results and uploading them to a local data center; The local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface; Obtaining classroom surveillance images using cameras is When the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that a person has entered the classroom, the corresponding smart classroom guide board in the classroom checks whether there is a class within the preset time period; When the smart classroom information board finds that there is a class in the corresponding classroom within the preset time period, it sends a control signal to the cameras in the classroom and at the classroom entrance / exit using ZigBee communication standard, and after receiving the control signal, the cameras in the classroom and at the classroom entrance / exit start monitoring and send the monitoring screen to the corresponding smart classroom information board; When the smart classroom information board detects that there are no classes in the corresponding classroom within the preset time period, it sends a control signal to the camera at the entrance / exit of the classroom using the ZigBee communication standard, and after receiving the control signal, the camera at the entrance / exit of the classroom starts monitoring and transmits the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model.

2. When the smart classroom information board finds that there is no class in the corresponding classroom within the preset time period, it sends a control signal to the camera at the entrance of the classroom using the ZigBee communication standard. After receiving the control signal, the camera at the entrance of the classroom starts monitoring and sends the monitoring screen to the corresponding smart classroom information board at the monitoring frequency obtained by the frequency calculation model, which is expressed by the following formula: [0.001] where F is the frequency, FD is the dissimilarity between the previous and next frames of the surveillance video, Presence is a binary variable indicating the presence of personnel, which is 1 if the infrared sensor in the classroom detects that a person has entered the classroom, and 0 otherwise, α is a weight for balancing the frame dissimilarity and the structural similarity index, k1 is a constant for adjusting the frequency of surveillance transmission, and the value range of k1 is [15, 30], S is the structural similarity index, [Equation 43] Here, x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually [0.0000] [Equation 45] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x, y) is a similarity function to measure the change in light-shadow. [Equation 46] Here, L Ｘ and Ly represent the light intensities of the two images, respectively, and L ｍａｘ The real-time people flow statistics method based on Internet of Things (IoT) information collection according to claim 1, characterized in that:

3. The smart classroom information board processes the classroom monitoring screen, stores the processed results, and uploads them to the local data center. After receiving the surveillance images from inside the classroom and at the entrance and exit of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrance and exit of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people; if it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding electronic device of the teacher, so that the teacher can further check the attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; The smart classroom information board receives the monitoring screen of the classroom entrance and exit, and then first pre-processes the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, and displays it through a visualization interface, and uploads it to the local data center; During data preprocessing, we optimize through the following model: [Equation 47] where: [Number 48] represents a new morphological process, A represents the input image, and B represents the structuring element for defining the shape and size of the dilation and erosion process. [Number 49] represents the result of dilation obtained by dilation processing on image A, [Number 50] teeth [0.51] The real-time people flow statistics method based on Internet of Things (IoT) information collection according to claim 1, characterized in that the output diagram obtained by performing corrosion processing on the results of the above is displayed.

4. The local data center generates real-time people flow statistics results and displays them to teachers and students through a visualization interface; The local data center processes the attendance information of the classrooms where classes are held, stores it according to the attendance rate of the corresponding class, and displays the processed attendance rate of each classroom through a visualization interface; The local data center processes the number of people entering and leaving the vacant classrooms, arranges them according to the display format of the smart classroom guide board, and forms the general smart classroom guide board display data and the smart classroom guide board display data of each building and each floor; The real-time people flow statistics method based on Internet of Things (IoT) information collection according to claim 1, further comprising: displaying the display data of the smart classroom guide board on the smart classroom guide board; and transmitting the display data of the smart classroom guide board on each floor of each building to the corresponding smart classroom guide board devices on each floor of each building through a local area network communication method; and allowing students to register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms.

5. A real-time people flow statistics system based on Internet of Things (IoT) information collection, The system comprises: a monitoring module for capturing a monitoring screen of the classroom using a camera; a processing module for processing the classroom monitoring screen through the smart classroom information board, storing the processed results, and uploading them to a local data center; The local data center includes a display module for generating real-time people flow statistics results and displaying them to teachers and students through a visualization interface.

6. The monitoring module a classroom entry detection module for detecting whether there is a class within a preset time period when the infrared sensor in the classroom detects that there is no one in the classroom and the infrared sensor at the classroom entrance detects that there is a person entering the classroom; a classroom and classroom entrance camera start module for sending a control signal to the cameras in the classroom and at the classroom entrance using ZigBee communication standard when the smart classroom information board finds that a class is scheduled in the corresponding classroom within a preset time period, and the cameras in the classroom and at the classroom entrance start monitoring after receiving the control signal and transmit the monitoring screen to the corresponding smart classroom information board; and a classroom entrance camera starting module for sending a control signal to a classroom entrance camera using ZigBee communication standard when the smart classroom information board detects that there is no class in the corresponding classroom within a preset time period, and the classroom entrance camera starts monitoring after receiving the control signal and transmits the monitoring screen to the corresponding smart classroom information board at a monitoring frequency obtained by a frequency calculation model.

7. The classroom entrance / exit camera starting module includes a frequency calculation model module; [Number 52] where F is the frequency, FD is the dissimilarity between the previous and next frames of the surveillance video, Presence is a binary variable indicating the presence of personnel, which is 1 if the infrared sensor in the classroom detects that a person has entered the classroom, and 0 otherwise, α is a weight for balancing the frame dissimilarity and the structural similarity index, and the value range of α is (0, 1), k1 is a constant for adjusting the frequency of surveillance transmission, and the value range of k1 is [15, 30], and S is the structural similarity index. [Number 53] Here, x and y are the two images to be matched, μ1 and μ2 are the average values ​​of the two images, σ1*σ1 represents the variance of image x, σ2*σ2 represents the variance of image y, σ12 is the covariance of the two images, C1 and C2 are constants to avoid the case where the denominator is 0. The values ​​of the constants vary depending on the color range, and are usually [Number 54] [Number 55] where L is the range of pixel numbers, L-1 is the maximum pixel value, K1 and K2 are constants less than 1, β is a weighting factor to balance the influence of structural similarity and light-shadow similarity on the overall similarity, and I(x, y) is a similarity function to measure the change in light-shadow. [Number 56] Here, L Ｘ and Ly represent the light intensities of the two images, respectively, and L ｍａｘ The real-time people flow statistics system based on Internet of Things (IoT) information collection according to claim 5, characterized in that:

8. The processing module includes: After receiving the surveillance images of the inside of the classroom and the entrance of the classroom, the smart classroom information board first compares the faces captured on the surveillance screen at the entrance of the classroom with the faces of the class attending class stored on the smart classroom information board, and then uses the YOLOv5s model to continuously measure the number of people in the surveillance video in the classroom to obtain the number of people; if it finds that there is a face mismatch or the number of faces does not match the number of people in the classroom, it will send the information to the corresponding teacher's electronic device to further confirm the teacher's attendance status, and after obtaining the teacher's confirmation information, it will upload it to the local data center; and a class response processing module. The smart classroom information board includes an empty classroom response processing module, which receives the classroom entrance / exit monitoring screen, first preprocesses the data, and then uses the YOLOv5s model to obtain the number of people entering and exiting the classroom, display it through a visualization interface, and upload it to a local data center; During data preprocessing, we optimize through the following model: [Number 57] where: [Number 58] represents a new morphological process, A represents the input image, and B represents the structuring element for defining the shape and size of the dilation and erosion process. [Number 59] represents the result of dilation obtained by dilation processing on image A, [Number 60] teeth [Number 61] The real-time people flow statistics system based on Internet of Things (IoT) information collection according to claim 5, characterized in that the output diagram obtained by performing corrosion processing on the results of the above is displayed.

9. The display module includes: The local data center processes the attendance information of the classrooms where classes are held, and then stores the information according to the attendance rate of the corresponding class. A corresponding class attendance rate display module is provided to display the processed attendance rate of each classroom through a visualization interface. a local data center processing the number of people entering and leaving the vacant classrooms, arranging them according to the display format of the smart classroom guide board, and forming the display data of the general smart classroom guide board and the display data of the smart classroom guide board of each building and each floor; and a transmission module for displaying the display data of the smart classroom guide board on the smart classroom guide board, and transmitting the display data of the smart classroom guide board on each floor of each building to the corresponding smart classroom guide board device on each floor of each building through a local area network communication method, and allowing students to register on the campus network to check the information on the smart classroom guide board and obtain information on available classrooms.

Images