CN-122024135-A - Real-time anti-cheating method for online examination

CN122024135ACN 122024135 ACN122024135 ACN 122024135ACN-122024135-A

Abstract

The application discloses a real-time anti-cheating method for an online examination, which comprises the steps of responding to an examination starting instruction, collecting video streams of an examinee in the examination process, calling a preset visual perception model to detect each frame of picture in the video streams, triggering abnormal early warning if abnormal behaviors of the examinee are determined, displaying the abnormal early warning on an examinee interface, and uploading the abnormal early warning to a monitoring background. Therefore, abnormal behaviors (such as changing a person, turning a head, using a mobile phone and the like) are detected in real time by utilizing a preset visual perception model through automatic acquisition of the examination video stream, and early warning is displayed on an examinee end in real time and synchronously uploaded to an examination background, so that automatic identification, real-time intervention and evidence preservation of cheating behaviors are realized under the condition of unmanned on-site examination, and the fairness and the authenticity of on-line examination are effectively ensured.

Inventors

Yang Huankun
Shi Bisong
LIN ZHENGJIE
HUANG WU
YANG TING
Wen Tingying

Assignees

深圳市海云天科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260123

Claims (10)

1. An online examination real-time anti-cheating method, which is characterized by comprising the following steps: responding to an examination starting instruction, and collecting video streams of an examinee in the examination process; Calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined; and displaying the abnormal early warning on an examinee interface, and uploading the abnormal early warning to a monitoring background.
2. The online examination real-time anti-cheating method according to claim 1, wherein the preset visual perception model comprises a face detection model; Before the response of the examination starting instruction and the acquisition of the video stream of the examinee in the examination process, the method comprises the following steps: responding to a login instruction of the examinee, and collecting image information of the examinee through an image collecting module; Inputting the image information into a face detection model for examinee quantity detection, and outputting a first boundary frame and a first confidence coefficient; if only one first boundary box is detected and the first confidence coefficient is larger than a preset threshold value, confirming that only one examinee exists in the examinee interface; inputting image data corresponding to the face area defined by the first boundary frame into a face feature model, and extracting reference face data of the examinee; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: invoking the face detection model to detect the face of each frame of picture in the video stream, and extracting face data of an examinee included in the video stream; And triggering a person changing early warning if the similarity between the face data of the examinee and the reference face data included in the video stream is lower than a first preset value.
3. The online examination real-time anti-cheating method according to claim 1, wherein the preset visual perception model comprises a face detection model; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: Invoking the face detection model to detect the face of each frame of picture in the video stream, and outputting a second boundary frame and a second confidence coefficient if the second boundary frame and the second confidence coefficient are output; And if the second confidence coefficient is smaller than a preset threshold value, judging that the current frame is in an unmanned state, and triggering unmanned early warning.
4. The online examination real-time anti-cheating method of claim 3, further comprising: Recording the number of frames continuously determined to be in an unmanned state through a sliding time window; and if the number of frames judged to be in the unmanned state reaches the preset minimum duration, triggering unmanned early warning.
5. The method for preventing cheating in real time on-line examination according to claim 1, wherein the preset visual perception model comprises a face detection model, and the method comprises the following steps: dynamically constructing a detection area with aligned centers based on the width and the height of each frame of picture in the video stream; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: Invoking the face detection model to detect the face of each frame of picture in the video stream, and outputting a third boundary frame with a third confidence coefficient larger than a preset threshold; Calculating the geometric center coordinates of the third boundary frame; if the geometric center coordinate is detected not to be located in the detection area, the fact that the examinee leaves the examination interface is judged, and the examinee leaves the examination interface early warning is triggered.
6. The method for online examination real-time anti-cheating according to claim 5, further comprising: recording the number of frames continuously judged to leave the examination interface by sliding a time window; And if the number of frames of the test taker leaving the test interface reaches the preset minimum continuous duration, triggering the test taker to leave the test interface for early warning.
7. The online examination real-time anti-cheating method according to claim 1, wherein the preset visual perception model comprises a face detection model; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: Invoking the face detection model to detect the face of each frame of picture in the video stream, and determining left eye coordinate information, right eye coordinate information, left ear coordinate information and right ear coordinate information of an examinee; determining a first distance between the left eye and the right eye of the examinee according to the left eye coordinate information and the right eye coordinate information; Determining a second distance between the left ear and the right ear of the examinee according to the left ear coordinate information and the right ear coordinate information; determining a face yaw angle according to the first distance and the second distance; When the face yaw angle is detected to be larger than a first preset fixed value, the head turning early warning of the examinee is triggered.
8. The online examination real-time anti-cheating method according to claim 1, wherein the preset visual perception model comprises a face detection model; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: Invoking the face detection model to detect the face of each frame of picture in the video stream, and determining nose coordinate information and mouth coordinate information of an examinee; Determining a third distance between the nose and the mouth of the examinee according to the nose coordinate information and the mouth coordinate information; determining a pitch angle according to the third distance; When the pitch angle is detected to be smaller than a second preset fixed value, triggering the head-down early warning of the examinee.
9. The online examination real-time anti-cheating method according to claim 1, wherein the preset visual perception model comprises a target detection model; and calling a preset visual perception model to detect each frame of picture in the video stream, and triggering abnormal early warning if abnormal behaviors of an examinee are determined, wherein the method comprises the following steps: Invoking the target detection model to detect each frame of picture in the video stream, and outputting a fourth boundary box, a category label and a fourth confidence coefficient; and if the class label is a preset class and the fourth confidence coefficient is larger than a preset value, triggering forbidden article early warning.
10. The online examination real-time anti-cheating method of claim 9, further comprising: and if at least two fourth bounding boxes exist, class labels corresponding to the at least two fourth bounding boxes are examinee classes, and the fourth confidence coefficient is larger than a preset value, triggering multi-person early warning.

Description

Real-time anti-cheating method for online examination Technical Field The application relates to the technical field of online prison, in particular to a real-time anti-cheating method for online examination. Background With the popularization of internet technology, online examination has become an important way widely adopted in the fields of education assessment, qualification certification, professional recruitment and the like. Compared with the traditional offline examination room, the online examination room has the remarkable advantages of flexible organization, low cost, no regional limitation and the like. However, virtualization and remodelling of the examination environment also bring serious challenges, and it is difficult to effectively identify and prevent cheating behaviors under the condition of lacking direct supervision of on-site invigilators, so as to ensure fairness, seriousness and achievement authenticity of the examination. Disclosure of Invention In view of the above problems, the present application provides a real-time anti-cheating method for online examination, which can solve the above problems. The embodiment of the application provides a real-time anti-cheating method for an online examination, which comprises the steps of responding to an examination starting instruction, collecting video streams of an examinee in the examination process, calling a preset visual perception model to detect each frame of picture in the video streams, triggering abnormal early warning if abnormal behaviors of the examinee are determined, displaying the abnormal early warning on an examinee interface, and uploading the abnormal early warning to a monitoring background. Therefore, abnormal behaviors (such as changing a person, turning a head, using a mobile phone and the like) are detected in real time by utilizing a preset visual perception model through automatic acquisition of the examination video stream, and early warning is displayed on an examinee end in real time and synchronously uploaded to an examination background, so that automatic identification, real-time intervention and evidence preservation of cheating behaviors are realized under the condition of unmanned on-site examination, and the fairness and the authenticity of on-line examination are effectively ensured. Drawings In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required for the description of the embodiments will be briefly introduced below, and it is apparent that the drawings in the following description are only some embodiments of the present application, not all embodiments. All other embodiments and figures obtained by a person skilled in the art without any inventive effort are within the scope of protection of the present application based on the embodiments of the present application. Fig. 1 shows a flow diagram of an online examination real-time anti-cheating method provided by an embodiment of the application. Fig. 2 shows a schematic structural diagram of an online examination real-time anti-cheating device according to an embodiment of the present application. Fig. 3 shows a schematic structural diagram of an electronic device according to an embodiment of the present application. Fig. 4 is a schematic structural diagram of a computer readable storage medium according to an embodiment of the present application. Detailed Description Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention as detailed in the accompanying claims. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. It should be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It should also be understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations. Furthermore, the terms "first," "second," "third," and the like in the description of the present specification and in the appended claims, are used for distinguishing bet