Online proctoring methods and related equipment based on electronic fences

Abstract

This application provides an online proctoring method and related equipment based on an electronic fence. The method includes: acquiring a standard posture image of the examinee; obtaining an electronic fence for the examinee based on the standard posture image; acquiring a real-time posture image of the examinee at preset time intervals; performing human skeleton point detection on the real-time posture image to obtain multiple human skeleton points and the pixel coordinates of each human skeleton point in the real-time posture image; and monitoring the examinee online by comparing the pixel coordinates of each human skeleton point with the electronic fence. The above method can accurately identify abnormal behavior of examinees, further explore the hidden information in the proctoring video, improve proctoring efficiency, and monitor abnormal behavior of examinees in a timely manner.

Description

Technical Field

[0001] This application relates to the field of monitoring technology, specifically to online monitoring methods and related equipment based on electronic fences. Background Technology

[0002] With the development of internet technology, online examination platforms have been widely used across various industries. As online examinations become more prevalent, online examination proctoring is also receiving increasing attention.

[0003] Currently, the most common method for online exam proctoring involves installing surveillance equipment in the exam room. Proctors then manually analyze students' facial and visual behaviors, judging whether there are any cheating behaviors such as leaving the exam room without permission, impersonating others, or multiple people taking the exam based on changes in the faces on the screen. In this method, the surveillance equipment in the exam room only serves to record the exam process; proctoring is still primarily done manually.

[0004] As the number of test takers increases, the work of invigilators in reviewing the surveillance footage becomes increasingly burdensome, especially when reviewing live video, as it is difficult for invigilators to monitor a large number of test takers' footage in a real-time and efficient manner. Summary of the Invention

[0005] In view of this, this application provides an online proctoring method and related equipment based on electronic fences, which can solve the problem of not being able to intelligently identify abnormal behavior of candidates during the proctoring process.

[0006] The first aspect of this application provides an online proctoring method based on an electronic fence. The method includes: acquiring a standard posture image of a candidate; obtaining an electronic fence for the candidate based on the standard posture image; acquiring a real-time posture image of the candidate at preset time intervals; performing human skeleton point detection on the real-time posture image to obtain multiple human skeleton points in the real-time posture image and the pixel coordinates of each human skeleton point; and monitoring the candidate online by comparing the pixel coordinates of each human skeleton point with the electronic fence.

[0007] The online proctoring method based on electronic fences provided in this application first obtains the electronic fence for the examinee based on the examinee's standard posture image. Then, it obtains the examinee's global human skeleton points by performing human skeleton point detection on the examinee's real-time posture image. Finally, it compares each human skeleton point with the electronic fence and monitors the examinee online based on the comparison results. In the above method, if the examinee's human skeleton points are outside the electronic fence, it indicates that the examinee is exhibiting abnormal behavior. By combining the comparison of the examinee's global human skeleton points with the electronic fence to monitor the examinee's potential abnormal behavior, it is possible to accurately identify situations where the examinee is exhibiting abnormal behavior. This allows for further in-depth mining of hidden information in the proctoring video, improves proctoring efficiency, and enables timely monitoring of the examinee's abnormal behavior, creating a better and fairer examination environment for examinees and providing proctors with more efficient and intelligent real-time proctoring technology.

[0008] In one embodiment, the online monitoring of the examinee includes determining that any one of the plurality of human skeleton points is an abnormal skeleton point if the pixel coordinates of any one of the human skeleton points are outside the electronic fence; outputting prompt information including standardized posture; and sending abnormal information including the abnormal skeleton point to the invigilator's monitoring terminal.

[0009] In one embodiment, the online monitoring of the examinee includes: if the pixel coordinates of any one of the plurality of human skeleton points are outside the electronic fence, determining that any one human skeleton point as an abnormal skeleton point; storing the abnormal skeleton point; performing fine-grained subdivision on the abnormal skeleton point to obtain the human body part corresponding to the abnormal skeleton point; obtaining the examinee's cumulative out-of-bounds score based on a preset out-of-bounds score for the human body part corresponding to the abnormal skeleton point; if the cumulative out-of-bounds score is greater than a preset score threshold, outputting a prompt message including correct posture, and sending abnormal information including the abnormal skeleton point to the invigilator's monitoring terminal, so that the invigilator can judge the examinee's violation.

[0010] In one embodiment, the human body part is one of the head, shoulder, hand, leg, and foot.

[0011] In one embodiment, the electronic fence includes a planar electronic fence and a vertical electronic fence. The online monitoring of the examinee includes: comparing the pixel coordinates of each human skeleton point with the planar electronic fence and the vertical electronic fence respectively; if the pixel coordinates of any one of the multiple human skeleton points are outside the planar electronic fence, and / or, the pixel coordinates of any one human skeleton point are outside the vertical electronic fence, the any one human skeleton point is determined to be an abnormal skeleton point; outputting prompt information including standardized posture, and sending abnormal information including the abnormal skeleton point to the invigilator's monitoring terminal.

[0012] In one embodiment, obtaining the electronic fence of the examinee based on the standard pose image includes: performing background subtraction processing on the standard pose image to obtain a mask image of the examinee's body in the standard pose image; and performing multiple dilation operations on the mask image to obtain the electronic fence of the examinee.

[0013] In one embodiment, obtaining the electronic fence of the examinee based on the standard pose image further includes: performing depth estimation on the mask image to obtain depth estimation information of the mask image; obtaining the farthest distance and the closest distance of the examinee from the host based on the depth estimation information; and adding a preset offset to the farthest distance and the closest distance to obtain the vertical electronic fence.

[0014] A second aspect of this application provides an online proctoring device based on an electronic fence, comprising: a first acquisition module for acquiring a standard posture image of a candidate and obtaining an electronic fence for the candidate based on the standard posture image; a second acquisition module for acquiring a real-time posture image of the candidate at preset time intervals; a human body detection module for performing human skeleton point detection on the real-time posture image to obtain multiple human skeleton points in the real-time posture image and the pixel coordinates of each human skeleton point; and an anomaly monitoring module for online monitoring of the candidate by comparing the pixel coordinates of each human skeleton point with the electronic fence.

[0015] A third aspect of this application provides an electronic device, the electronic device comprising: a memory for storing at least one instruction; and a processor for executing the at least one instruction to implement the online monitoring method based on an electronic fence as described in the above embodiments.

[0016] A fourth aspect of this application provides a computer-readable storage medium storing at least one instruction that is executed by a processor in an electronic device to implement the online monitoring method based on an electronic fence as described in the above embodiments. Attached Figure Description

[0017] Figure 1 This is an application scenario diagram of an online examination monitoring method based on electronic fences provided in the embodiments of this application.

[0018] Figure 2 This is a flowchart of an online examination monitoring method based on electronic fences provided in the embodiments of this application.

[0019] Figure 3 This is a monitoring scene diagram of an online examination monitoring method based on electronic fences provided in one embodiment of this application.

[0020] Figure 4 This is a partially detailed flowchart of step S400 provided in one embodiment of this application.

[0021] Figure 5 This is a partially detailed flowchart of step S400 provided in another embodiment of this application.

[0022] Figure 6 This is a partially detailed flowchart of step S400 provided in another embodiment of this application.

[0023] Figure 7 This is a monitoring scene diagram of an online examination monitoring method based on electronic fences, provided in another embodiment of this application.

[0024] Figure 8 This is a schematic diagram of an online examination monitoring device based on an electronic fence, provided in an embodiment of this application.

[0025] Figure 9 This is a schematic diagram of an electronic device provided in an embodiment of this application.

[0026] Explanation of main component symbols

[0027] Candidate Terminal 10

[0028] Host 11

[0029] Auxiliary machine 12

[0030] 30 proctoring terminals

[0031] Online proctoring device 100

[0032] First Acquisition Module 110

[0033] Second acquisition module 120

[0034] Human body detection module 130

[0035] Anomaly monitoring module 140

[0036] Electronic devices 20

[0037] Memory 21

[0038] Processor 22 Detailed Implementation

[0039] To better understand the above-mentioned objectives, features, and advantages of this application, the application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing an embodiment in one instance only and is not intended to be limiting of the application.

[0041] It should be noted that in this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and drawings of this application are used to distinguish similar objects, not to describe a specific order or sequence.

[0042] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0043] It should also be noted that the methods disclosed in the embodiments of this application or the methods shown in the flowcharts include one or more steps for implementing the method. Without departing from the scope of the claims, the execution order of multiple steps can be interchanged, and some steps can also be deleted.

[0044] Some embodiments will now be described with reference to the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0045] Currently, most online exams rely on dual-camera or multi-camera setups with multiple cameras. One camera captures the examinee's face, while another one or more cameras capture the examinee's surroundings. Invigilators then review the recorded videos or live feeds from these multiple devices to determine if any violations have occurred during the exam.

[0046] However, with the increase in the number of examinees, the workload for invigilators in reviewing monitoring footage from multiple devices becomes increasingly heavy, especially when reviewing live video streams, where it is difficult for invigilators to monitor a large number of examinees in real time. Therefore, online examinations require a method to constrain examinee behavior and issue alerts for abnormal behavior to alleviate the burden on invigilators.

[0047] To address the aforementioned issues, this application provides an online proctoring method based on electronic fences, which can accurately identify abnormal behavior by examinees, further mine hidden information in the proctoring video, improve proctoring efficiency, and monitor examinees' abnormal behavior in a timely manner.

[0048] Figure 1 This is an application scenario diagram of an online examination monitoring method based on electronic fences provided in the embodiments of this application.

[0049] The online proctoring method provided in this application can be applied to a scenario consisting of a candidate terminal 10 and a proctoring terminal 30 that are interconnected. The candidate terminal 10 is used to detect abnormal behavior of candidates and send abnormal information to the proctoring terminal 30 when abnormal behavior is detected. The proctoring terminal 30 is used to provide proctoring personnel with proctoring video and abnormal information. In one embodiment, the candidate terminal 10 can be a mobile phone, computer equipment, camera, or other terminal device with camera and communication functions, and the proctoring terminal 30 can be a mobile phone, computer equipment, large screen, or other terminal device with communication and display functions. In practical applications, the device form of the candidate terminal 10 and the proctoring terminal 30 is not limited.

[0050] In this embodiment, the candidate terminal 10 may include a host 11 and an auxiliary device 12. The host 11 is located in front of the candidate and is used to detect abnormal behavior of the candidate. The auxiliary device 12 is located to the side of the candidate and is used to monitor the candidate's surrounding environment.

[0051] Figure 2 This is a flowchart illustrating an online proctoring method based on an electronic fence, as provided in an embodiment of this application. The method is applied to the examinee terminal 10, for example... Figure 1 The host 11 is shown. Depending on different requirements, the order of the steps in this flowchart can be changed, and some steps can be omitted.

[0052] Specifically, the above-mentioned online proctoring methods include:

[0053] S100: Obtain the candidate's standard posture image, and obtain the candidate's electronic fence based on the standard posture image.

[0054] In one embodiment, before executing step S100, the host 11 may prompt the examinee to begin the exam. After the examinee confirms the start of the exam, the host 11 will display a prompt for the student to maintain a standard sitting posture for a preset time, such as 5 seconds, and then take an image of the examinee's standard posture after 5 seconds. Then, the host 11 performs background subtraction processing on the standard posture image to obtain a mask image of the examinee's body in the standard posture image. The host 11 dilates the mask image multiple times to obtain the examinee's electronic fence. Specifically, as an example, when dilating the mask image, the host 11 treats areas with a pixel value of 1 as human body areas and areas with a pixel value of 0 as non-human body areas. The host 11 dilates the mask image N times to obtain the electronic fence corresponding to the examinee. For example, N can be 1000, meaning that dilating the mask image 1000 times can obtain the examinee's electronic fence.

[0055] In the above embodiments, portrait matting technology is used to generate a transparency mask for the human body in the image, retaining only the human body and removing the background. Image dilation involves an image morphology processing algorithm that can increase bright areas. An electronic fence is represented as an area enclosed by a series of points, which can be understood as the activity range of a test-taker during the exam. In this embodiment, the electronic fence is the image obtained after multiple dilations of the mask image.

[0056] S200: Acquire the candidate's real-time posture image at preset time intervals.

[0057] For example, the preset time can be 1 second, and the host 11 acquires the candidate's real-time posture image every 1 second. In practical applications, it is not limited to this and can be set and adjusted according to the exam type / exam content, etc.

[0058] S300, Perform human skeleton point detection on the real-time posture image to obtain multiple human skeleton points in the real-time posture image and the pixel coordinates of each human skeleton point.

[0059] Specifically, host 11 uses real-time pose images as input data for the Alphapose model. The Alphapose model then uses the input data to identify human skeletal points, ultimately identifying multiple human skeletal points and the pixel coordinates of each point. The Alphapose model employs a top-down approach, first using the Faster R-CNN (faster region-based convolutional neural network) object detector to detect the candidate's human body region in the real-time pose image, and then using the single-person poseestimator (SPPE) algorithm to detect skeletal points in the detected human body region, ultimately identifying multiple human skeletal points and the pixel coordinates of each point.

[0060] S400: The examinee is monitored online by comparing the pixel coordinates of each human skeleton point with the electronic fence.

[0061] Specifically, host 11 determines the pixel coordinate range of the electronic fence area and, based on the pixel coordinates of each human skeleton point, determines whether each human skeleton point is outside the electronic fence. For example, if all human skeleton points are outside the electronic fence, it is considered that the examinee has not engaged in any abnormal behavior, and the examination continues normally. Figure 3 As shown in a.

[0062] For example, if the pixel coordinates of a human skeleton point are located on the edge of the electronic fence or inside the electronic fence, it indicates that the human skeleton point is not outside the electronic fence.

[0063] The online proctoring method provided in the above embodiments first obtains the electronic fence of the examinee based on the examinee's standard posture image. Then, it obtains the examinee's global human skeleton points by performing human skeleton point detection on the examinee's real-time posture image. Finally, it compares each human skeleton point with the electronic fence and monitors the examinee online based on the comparison results. In the above method, if the examinee's human skeleton points are outside the electronic fence, it indicates abnormal behavior. By combining the comparison of the examinee's global human skeleton points with the electronic fence, potential cheating behaviors can be monitored, accurately identifying abnormal behavior, further deepening the mining of hidden information in the proctoring video, improving proctoring efficiency, and timely monitoring of abnormal behavior, thus creating a better and fairer examination environment for examinees and providing proctors with more efficient and intelligent real-time proctoring technology.

[0064] Figure 4 This application provides an embodiment of the method for... Figure 2A detailed flowchart of step S400 is provided. Specifically, the online monitoring of the examinee in step S400 includes:

[0065] S410, if the pixel coordinates of any one of the plurality of human skeleton points are outside the electronic fence, then the arbitrary human skeleton point is determined to be an abnormal skeleton point.

[0066] Specifically, if, during the detection process, the pixel coordinates of any one of the multiple human skeleton points are outside the electronic fence, such as... Figure 3 As shown in b and c, the abnormal behavior of the examinee is indicated, and the human skeleton points outside the electronic fence are identified as abnormal skeleton points.

[0067] S411, output prompts including correct posture, and send abnormal information including the abnormal skeletal points to the invigilator's terminal so that the invigilator can judge the candidate's violation.

[0068] Specifically, if during the detection process, if the pixel coordinates of any one of the multiple human skeletal points are detected outside the electronic fence, the host 11 outputs a prompt message including a correct posture to remind the examinee to maintain the correct posture. For example, the host 11's display interface pops up a prompt message saying "Please maintain a correct sitting posture" and outputs a voice prompt message saying "Please maintain a correct sitting posture." At the same time, it sends abnormal information including abnormal skeletal points to the invigilator's monitoring terminal 30 so that the invigilator can focus on observing the examinee. The invigilator combines the monitoring video provided by the host 11 and the auxiliary machine 12 to judge the degree of violation of the examinee and remind or punish the examinee according to the degree of violation. For examinees with serious violations, the examination can be terminated directly. For example, the abnormal information may include the examinee's information on the abnormal skeletal points, the information on the abnormal skeletal points, and the monitoring video of the examinee with abnormal skeletal points.

[0069] The online proctoring method provided in the above embodiments treats all abnormal skeletal points equally, regardless of which part of the human body the abnormal skeletal point belongs to. When the pixel coordinates of any human skeletal point are detected outside the electronic fence, an abnormal alarm is immediately sent to the teacher's terminal. This method can comprehensively identify abnormal behavior of candidates, monitor such behavior in a timely manner, create a more rigorous and fair examination environment for candidates, and provide proctors with more efficient and intelligent real-time proctoring technology.

[0070] Figure 5 This is another embodiment of the present application that provides a solution for... Figure 2 A detailed flowchart of step S400 is provided. Specifically, online monitoring of the examinee includes:

[0071] S420, if the pixel coordinates of any one of the plurality of human skeleton points are outside the electronic fence, then the arbitrary human skeleton point is determined to be an abnormal skeleton point.

[0072] Specifically, if during the detection process, the pixel coordinates of any one of the multiple human skeleton points are outside the electronic fence, such as... Figure 3 As shown in b and c, the abnormal behavior of the examinee is indicated, and the human skeleton points outside the electronic fence are identified as abnormal skeleton points.

[0073] S421, Store the abnormal bone points.

[0074] Specifically, the host temporarily stores abnormal skeletal points outside the electronic fence, for example, by storing image information of the abnormal skeletal points.

[0075] S422, perform fine-grained division on the abnormal bone points to obtain the human body parts corresponding to the abnormal bone points.

[0076] Specifically, the method for fine-grained segmentation of the abnormal skeletal points includes: first, detecting the location of the abnormal skeletal points in the real-time pose image using preset human body blocks to obtain the target block; then, detecting the discriminative regions within the preset human body blocks; and finally, simultaneously feeding the target block and the discriminative preset human body blocks into a CNN classification system to obtain the human body part corresponding to the abnormal skeletal points. For example, the human body part may be one of the following: head, shoulder, hand, leg, or foot.

[0077] S423, the candidate's cumulative out-of-bounds score is obtained based on the preset out-of-bounds score of the human body part corresponding to the abnormal skeletal point.

[0078] In one embodiment of this application, a preset boundary score for each body part can be set in advance. The cumulative boundary score of candidates with abnormal bone points can be calculated by setting the preset boundary score for each body part. For example, the rules for dividing the preset boundary score for each body part are as follows: the boundary score for abnormal bone points of the head is 40, the boundary score for abnormal bone points of the shoulder is 10, the boundary score for abnormal bone points of the hand is 30, the boundary score for abnormal bone points of the leg is 10, and the boundary score for abnormal bone points of the foot is 10.

[0079] In some embodiments, such as Figure 3 As shown in b, the candidate has 3 abnormal bone points in their head, so the cumulative score for exceeding the limit is 3 * 40 = 120.

[0080] In some embodiments, such as Figure 3 As shown in c, the candidate has 1 abnormal bone point on the head and 3 abnormal bone points on the hands. Therefore, the cumulative score for exceeding the limit is 1*40+3*30=130.

[0081] S424, if the cumulative out-of-bounds score is greater than a preset score threshold, output a prompt message including the correct posture, and send abnormal information including the abnormal skeletal points to the invigilator's invigilation terminal so that the invigilator can judge the candidate's violation.

[0082] For example, if the score threshold is 60, then Figure 3 In scenarios b and c, if the cumulative score for exceeding the limit is greater than 60, the host 11 will output a prompt message including a correct posture to remind the examinee to maintain the correct posture. For example, the host 11's display interface will pop up a prompt message saying "Please maintain a correct sitting posture" and output a voice prompt message saying "Please maintain a correct sitting posture". At the same time, it will send abnormal information including abnormal skeletal points to the invigilator's monitoring terminal 30 so that the invigilator can focus on observing the examinee. The invigilator will judge the degree of violation of the examinee based on the monitoring video provided by the host 11 and the auxiliary machine 12, and will remind or punish the examinee according to the degree of violation. For examinees with serious violations, the examination can be terminated directly. For example, the abnormal information may include the examinee's information on the abnormal skeletal points, the information on the abnormal skeletal points, and the monitoring video of the examinee with abnormal skeletal points.

[0083] The online proctoring method provided in the above embodiments assigns different out-of-bounds scores to abnormal skeletal points on different parts of the human body according to the likelihood of cheating on different parts of the body. The out-of-bounds scores of abnormal skeletal points are accumulated, and when the accumulated out-of-bounds scores exceed a preset score threshold, an abnormal alarm is sent to the teacher's terminal. It can accurately identify abnormal behavior of candidates based on the actual situation, monitor candidates' abnormal behavior in a timely manner, create a more rigorous and fair examination environment for candidates, and provide proctors with more efficient and intelligent real-time proctoring technology.

[0084] Figure 6 This is another embodiment of the present application that provides a solution for... Figure 2 A detailed flowchart of step S400 is provided. In this embodiment, the electronic fence includes a planar electronic fence and a vertical electronic fence.

[0085] In one embodiment, the method for the host 11 to acquire a vertical electronic fence includes: performing depth estimation on a mask image to obtain depth estimation information of the mask image; obtaining the farthest distance and the closest distance between the examinee and the host 11 based on the depth estimation information; and adding a preset offset to the farthest distance and the closest distance to obtain the vertical electronic fence.

[0086] In one embodiment, the method for the host 11 to acquire a planar electronic fence includes: performing background subtraction processing on a standard pose image to obtain a mask image of the examinee's body in the standard pose image, and then performing multiple dilation operations on the mask image to obtain the planar electronic fence of the examinee.

[0087] Specifically, monitoring the candidates includes:

[0088] S430, compare the pixel coordinates of each of the human skeleton points with the planar electronic fence and the vertical electronic fence respectively.

[0089] Specifically, the pixel coordinates of each human skeleton point are compared to see if they are outside the planar electronic fence. If the pixel coordinates of multiple human skeleton points are not outside the planar electronic fence, the pixel coordinates of each human skeleton point need to be compared with the vertical electronic fence. If the pixel coordinates of any one of the multiple human skeleton points are outside the planar electronic fence, then it is not necessary to compare the pixel coordinates of that one human skeleton point with the vertical electronic fence, and the pixel coordinates of that one human skeleton point can be directly identified as an abnormal skeleton point.

[0090] S431, if the pixel coordinates of any one of the plurality of human skeleton points are within the planar electronic fence, and / or the pixel coordinates of any one of the human skeleton points are outside the vertical electronic fence, then the any one of the human skeleton points is determined to be an abnormal skeleton point.

[0091] Specifically, if the pixel coordinates of any human skeleton point are outside the horizontal electronic fence or the pixel coordinates of any human skeleton point are outside the vertical electronic fence, then that human skeleton point is determined to be an abnormal skeleton point.

[0092] For example, Figure 7 As shown in b and c, if the pixel coordinates of a human skeleton point are outside the planar electronic fence, then it is determined that... Figure 7 In points b and c, the human skeleton points outside the planar electronic fence are abnormal skeleton points.

[0093] For example, Figure 7 As shown in B and C, if the pixel coordinates of a human skeleton point are outside the vertical electronic fence, then it is determined that... Figure 7 B and C, which are human skeletal points outside the planar electronic fence, are considered abnormal skeletal points. S432, output prompts including correct posture and send abnormal information including the abnormal skeletal points to the proctor's monitoring terminal.

[0094] Specifically, the host 11 outputs prompts including correct posture to remind candidates to maintain the correct posture. For example, the host 11's display interface pops up a prompt message saying "Please maintain correct posture" and outputs a voice prompt message saying "Please maintain correct posture." At the same time, it sends abnormal information, including abnormal skeletal points, to the invigilator's monitoring terminal 30 so that the invigilator can closely observe the candidate. The invigilator combines the monitoring video provided by the host 11 and the auxiliary machine 12 to judge the degree of the candidate's violation and remind or punish the candidate according to the degree of violation. For candidates with serious violations, the examination can be terminated directly. For example, the abnormal information may include the candidate's information on the abnormal skeletal points, the information on the abnormal skeletal points, and the monitoring video of the candidate with the abnormal skeletal points.

[0095] In some embodiments, if the pixel coordinates of all the plurality of human skeleton points are within the planar electronic fence, such as Figure 7 As shown in a, and all the pixel coordinates of the multiple human skeleton points are within the vertical electronic fence, such as... Figure 7 As shown in A, if the candidate exhibits no abnormal behavior, the exam continues normally. The online proctoring method provided in the above embodiments extends the electronic fence for candidates to three-dimensional space based on depth estimation. By simultaneously constraining candidates with planar and vertical electronic fences, it can more accurately identify abnormal behavior, monitor such behavior in a timely manner, create a more rigorous and fairer examination environment for candidates, and provide proctors with more efficient and intelligent real-time proctoring technology.

[0096] In some embodiments, the host computer 11 and the auxiliary computer 12 can upload monitoring videos of examinees to a server (not shown). The server simultaneously performs human skeleton point detection on multiple monitoring videos provided by the host computer 11. If no abnormal behavior is detected, the examinee's exam is automatically considered valid. If abnormal behavior is detected in an examinee, the examiners or invigilators will focus on observing that student based on the monitoring videos provided by the host computer 11 and the auxiliary computer 12. If it is confirmed that there is no cheating, the examinee's exam is valid; otherwise, the examinee's exam result is deemed invalid.

[0097] In the above embodiments, the detection of abnormal behavior is moved to the server side, which facilitates random checks on the examination process after the exam and creates a more rigorous and fairer examination environment for candidates.

[0098] Figure 8 This is a schematic diagram of an online examination monitoring device 100 based on an electronic fence provided in an embodiment of this application. Specifically, the online examination monitoring device 100 includes: a first acquisition module 110, a second acquisition module 120, a human body detection module 130, and an anomaly monitoring module 140.

[0099] Specifically, the first acquisition module 110 is used to acquire a standard posture image of the examinee and obtain an electronic fence for the examinee based on the standard posture image. The second acquisition module 120 is used to acquire a real-time posture image of the examinee at preset time intervals. The human body detection module 130 is used to perform human skeleton point detection on the real-time posture image to obtain multiple human skeleton points in the real-time posture image and the pixel coordinates of each human skeleton point. The anomaly monitoring module 140 is used to perform online monitoring of the examinee by comparing the pixel coordinates of each human skeleton point with the electronic fence.

[0100] It is understood that the module division described above is a logical functional division, and there may be other division methods in actual implementation. Furthermore, the functional modules in the various embodiments of this application can be integrated into the same processing unit, or each module can exist physically separately, or two or more modules can be integrated into the same unit. The integrated modules described above can be implemented in hardware or in a combination of hardware and software functional modules.

[0101] The online proctoring device provided in the above embodiment firstly obtains the electronic fence of the examinee based on the examinee's standard posture image using a first acquisition module 110. Then, the human body detection module 130 performs human skeleton point detection on the real-time posture image of the examinee obtained by the second acquisition module 120 to obtain the examinee's global human skeleton points. Finally, the anomaly monitoring module 140 compares each human skeleton point with the electronic fence and monitors the examinee online based on the comparison results. In the above system, if the examinee's human skeleton points are outside the electronic fence, it indicates abnormal behavior. By combining the comparison of the examinee's global human skeleton points with the electronic fence to monitor potential cheating behavior, the system can accurately identify situations where examinees exhibit abnormal behavior. This allows for deeper analysis of hidden information in the proctoring video, improves proctoring efficiency, and enables timely monitoring of examinee abnormal behavior, creating a better and fairer examination environment for examinees and providing proctors with more efficient and intelligent real-time proctoring technology.

[0102] Figure 9 This is a schematic diagram of an electronic device 20 provided in an embodiment of this application.

[0103] In one embodiment of this application, the electronic device 20 includes, but is not limited to, a memory 21 and a processor 22. The electronic device 20 can serve as the host 11 of the examinee terminal 10. The memory 21 stores at least one instruction, and the processor 22 executes the at least one instruction stored in the memory 21 to implement the online proctoring method based on electronic fences described in the above embodiment.

[0104] For example, instructions can be divided into one or more modules / units, one or more of which are stored in memory 21 and executed by processor 22. One or more modules / units can be a series of instruction segments capable of performing a specific function, and these instruction segments describe the execution process of the instructions in electronic device 20. For example, they can be divided into... Figure 8 The first acquisition module 110, the second acquisition module 120, the human body detection module 130, and the anomaly monitoring module 140 are shown.

[0105] Electronic device 20 can be a desktop computer, laptop, handheld computer, industrial computer, tablet computer, server, or other computing device. Those skilled in the art will understand that the schematic diagram is merely an example of electronic device 20 and does not constitute a limitation on electronic device 20. It may include more or fewer components than shown in the diagram, or combine certain components, or use different components. For example, electronic device 20 may also include input / output devices, network access devices, buses, etc.

[0106] The processor 22 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0107] The memory 21 can be used to store computer programs and / or modules. The processor 22 implements various functions of the online proctoring device 100 by running or retrieving the computer programs and / or modules stored in the memory 21, and by calling the data stored in the memory 21. The memory 21 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as image acquisition function, image detection function, etc.), etc.; the data storage area may store data created based on the use of the online proctoring device 100, etc. In addition, the memory 21 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0108] The memory 21 can be the external memory and / or internal memory of the online monitoring device 100. Furthermore, the memory 21 can be a physical memory, such as a memory stick, a TF card (Trans-flash Card), etc.

[0109] This application also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor 22 in an electronic device 20 to implement the online monitoring method based on an electronic fence as described in the above embodiments.

[0110] If the program code and various data in memory 21 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments, such as the online proctoring method, can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), etc.

[0111] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within this application.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application.

Claims

1. An online examination monitoring method based on electronic fences, characterized in that, The method includes: A standard posture image of a candidate is obtained, and an electronic fence for the candidate is obtained based on the standard posture image. The process of obtaining the electronic fence for the candidate based on the standard posture image includes: performing background subtraction processing on the standard posture image to obtain a mask image of the candidate's body in the standard posture image; performing multiple dilation operations on the mask image to obtain the electronic fence for the candidate; the electronic fence includes a planar electronic fence and a vertical electronic fence. The candidate's real-time posture image is acquired at preset time intervals; Human skeleton points are detected in the real-time pose image to obtain multiple human skeleton points in the real-time pose image and the pixel coordinates of each human skeleton point. The examinee is monitored online by comparing the pixel coordinates of each human skeleton point with the electronic fence.

2. The online proctoring method based on electronic fences as described in claim 1, characterized in that, The online monitoring of the examinee includes: If the pixel coordinates of any one of the plurality of human skeleton points are outside the electronic fence, then that human skeleton point is determined to be an abnormal skeleton point. The output includes prompts for proper posture and sends abnormal information, including the abnormal skeletal points, to the proctor's monitoring terminal.

3. The online proctoring method based on electronic fences as described in claim 1, characterized in that, The online monitoring of the examinee includes: If the pixel coordinates of any one of the plurality of human skeleton points are outside the electronic fence, then that human skeleton point is determined to be an abnormal skeleton point. Store the abnormal skeletal points; The abnormal skeletal points are divided into fine-grained segments to obtain the human body parts corresponding to the abnormal skeletal points. Based on the preset out-of-bounds score of the human body part corresponding to the abnormal skeletal point, the candidate's cumulative out-of-bounds score is obtained. If the cumulative out-of-bounds score exceeds a preset score threshold, a prompt message including the correct posture is output, and abnormal information including the abnormal skeletal points is sent to the proctor's monitoring terminal.

4. The online examination monitoring method based on electronic fences as described in claim 3, characterized in that, The body part referred to is one of the following: head, shoulders, hands, legs, or feet.

5. The online proctoring method based on electronic fences as described in claim 1, characterized in that, The electronic fence includes a planar electronic fence and a vertical electronic fence, and the online monitoring of the examinee includes: The pixel coordinates of each of the human skeleton points are compared with the planar electronic fence and the vertical electronic fence, respectively; If the pixel coordinates of any one of the plurality of human skeleton points are outside the planar electronic fence, and / or, the pixel coordinates of any one of the human skeleton points are outside the vertical electronic fence, then the any one of the human skeleton points is determined to be an abnormal skeleton point. The output includes prompts for proper posture and sends abnormal information, including the abnormal skeletal points, to the proctor's monitoring terminal.

6. The online proctoring method based on electronic fences as described in claim 1, characterized in that, The process of obtaining the electronic fence for the examinee based on the standard posture image also includes: Depth estimation is performed on the mask image to obtain depth estimation information of the mask image; Based on the depth estimation information, obtain the farthest distance and the closest distance between the examinee and the host. The vertical electronic fence is obtained by adding a preset offset to the farthest distance and the nearest distance respectively.

7. An online examination monitoring device based on electronic fences, characterized in that, include: The first acquisition module is used to acquire a standard posture image of the examinee and obtain an electronic fence for the examinee based on the standard posture image. The process of obtaining the electronic fence for the examinee based on the standard posture image includes: performing background subtraction processing on the standard posture image to obtain a mask image of the examinee's body in the standard posture image; performing multiple dilation operations on the mask image to obtain the electronic fence for the examinee; the electronic fence includes a planar electronic fence and a vertical electronic fence. The second acquisition module is used to acquire the real-time posture image of the examinee at preset time intervals; The human body detection module is used to detect human skeleton points in the real-time posture image to obtain multiple human skeleton points in the real-time posture image and the pixel coordinates of each human skeleton point. The anomaly monitoring module is used to monitor the examinee online by comparing the pixel coordinates of each human skeleton point with the electronic fence.

8. An electronic device, characterized in that, The electronic device includes: Memory, storing at least one instruction; The processor executes the at least one instruction to implement the online monitoring method based on electronic fences as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which is executed by a processor in an electronic device to implement the online monitoring method based on an electronic fence as described in any one of claims 1 to 6.

Images