Examination room monitoring system based on artificial intelligence

Abstract

The present invention relates to the technical field of examination room monitoring systems. Disclosed is an examination room monitoring system based on artificial intelligence. The system comprises: a first image collection module, which is used for performing human body image collection on an examinee in an examination room, and uploading collected image data to a server terminal for data processing; an examination terminal, which is used for verifying the identity of an examination participant and issuing an examination paper, and transmitting, to the server terminal for processing, the examination paper, answering of which has been completed; the server terminal, which is used for identifying whether the examinee has cheated, and processing examination paper information uploaded by the examination terminal, so as to obtain an examination score; and a display terminal, which is used by an invigilator, wherein data processed by the server terminal is transmitted, by means of a wireless network, to the display terminal for display, and the invigilator performs corresponding handling on the basis of the displayed content. The system can accurately and instantly find a discipline violation behavior, and mark and upload the discipline violation behavior to the display terminal for display.

Description

An examination room monitoring system based on artificial intelligence TECHNICAL FIELD

[0001] The present application relates to the technical field of examination room monitoring systems, and in particular to an examination room monitoring system based on artificial intelligence. BACKGROUND

[0002] A fair examination environment is a basic guarantee of examination quality. In order to improve the rigorous management system of education examination and create a fair and just examination order, an examination room and an examination site must have a sound examination supervision system. At present, most of the examination room invigilation methods use the artificial "1+1" front and rear flow invigilation mode. However, with the development of various information products, the students' initiative is enhanced, and the teachers' visual range is gradually reduced. The traditional mode of one static and one dynamic invigilating is gradually showing the problems of large demand for professional personnel, strong interference in the examination room, fairness and bias due to its instability. Arranging invigilators not only needs human cost, but also cannot identify the examination cheating behavior in time and comprehensively. In addition, in the prior art, whether an examinee cheats is usually judged by identifying the situation that the examinee's line of sight deviates, and the judgment standard is single and one-sided, resulting in misjudgment and omission.

[0003] SUMMARY

[0004] The technical problem to be solved by the present application is how to provide an examination room monitoring system based on artificial intelligence which can accurately judge whether a examinee has a violation behavior.

[0005] To solve the above technical problems, the technical solution adopted by the present application is: an examination room monitoring system based on artificial intelligence, comprising:

[0006] A first image acquisition module is configured to acquire human body images of examinees in an examination room and upload the acquired image data to a server terminal for data processing.

[0007] An examination terminal is configured to verify the identity of an examination personnel and issue an examination paper, and transmit the paper after completing the answer to the server terminal for processing.

[0008] A server terminal is configured to receive the data uploaded by the image acquisition module and the examination terminal, and process the data uploaded thereby, wherein the examinees are monitored by the image acquisition module, whether the examinees cheat is identified by processing the acquired human body images, and the paper information uploaded by the examination terminal is processed to obtain the examination score.

[0009] A display terminal is configured to be used by an invigilator, and the server terminal and the display terminal perform data transmission through a wireless network. The data processed by the server terminal is transmitted to the display terminal through the wireless network for display, and the invigilator makes corresponding processing according to the displayed content.

[0010] Further technical solutions are that the examination terminal comprises an identity verification module, the identity verification module is used for verifying the identity of the examination, only the examinee who passes the verification can enter the system and use the examination terminal.

[0011] Further technical solutions are that the server terminal comprises a first image processing module, the first image processing module processes the human body image through the following method:

[0012] S1: performing video frame extraction processing on the examination monitoring video;

[0013] S2: constructing an examination student irregular behavior identification and prediction model based on a deep learning algorithm;

[0014] S3: inputting the video frame extracted in step S1 into the identification and prediction model for identification processing;

[0015] S4: identifying and predicting the examinee irregular behavior of the input video frame in time sequence, if the prediction model detects that the current video frame has the examinee irregular behavior, marking the examinee in the video frame, saving the image to the local, and then repeating step S3 for the next video frame; if the prediction model does not detect that the current video frame has the examinee irregular behavior, directly repeating step S3 for the next video frame.

[0016] Further technical solutions are that the server terminal further comprises a test paper processing module, the test paper processing module is used for comparing the test paper uploaded by the examinee with the standard test paper with answers, marking the errors in the test paper, and then obtaining the score of the test paper.

[0017] Further technical solutions are that the display terminal comprises an administrator login module, the administrator login module is used for the invigilator to enter the display terminal through the secret key authentication mode, the secret key authentication adopts the encryption mode of U disk certificate and authentication password, when the invigilator enters the display terminal, if the secret key is not inserted in the display terminal, the system refuses to enter; only after the secret key is inserted and the administrator password is input at the same time, the display terminal can be entered for various operations.

[0018] The beneficial effects of the above technical solutions are that the monitoring system comprises a first image acquisition module, an examination terminal and a server terminal, the examinee is monitored through the first image acquisition module, and the image information is transmitted to the server terminal for processing, the server terminal is identified and processed through the identification and prediction model, the irregular behavior can be accurately and timely found, and the irregular behavior is marked and uploaded to the display terminal for display, and the invigilator processes the irregular personnel according to the related information. BRIEF DESCRIPTION OF DRAWINGS

[0019] The application will be described in further detail below with reference to the drawings and specific embodiments.

[0020] Fig. 1 is a schematic block diagram of the system according to an embodiment of the application;

[0021] Fig. 2 is a flowchart of the image processing of the first image processing module in the system according to an embodiment of the application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only some of the embodiments of the application, but not all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0023] In the following description, many specific details are set forth in order to provide a thorough understanding of the application. However, the application can be practiced without the specific details, which are not described in the present application. In other instances, well-known methods, procedures, components and circuits have not been described in detail so as not to obscure the application. Therefore, the present application is not intended to be limited to the embodiments described herein which can be changed as many times as possible by those skilled in the art without departing from the scope of the application.

[0024] As shown in Fig. 1, the application discloses an examination room monitoring system based on artificial intelligence, comprising:

[0025] The first image acquisition module is configured to acquire human body images of the examinees in the examination room and upload the acquired image data to the server terminal for data processing.

[0026] The examination terminal is configured to verify the identity of the examination personnel and issue examination papers, and transmit the completed answer sheets to the server terminal for processing.

[0027] The server terminal is configured to receive the data uploaded by the image acquisition module and the examination terminal, and process the uploaded data. The server terminal monitors the examinees through the image acquisition module, identifies whether the examinees cheat through processing the acquired human body images, and processes the paper information uploaded by the examination terminal to obtain the examination scores.

[0028] The display terminal is configured to be used by the invigilators, and the server terminal and the display terminal perform data transmission through a wireless network. The data processed by the server terminal is transmitted to the display terminal through the wireless network for display, and the invigilators make corresponding processing according to the displayed content.

[0029] Further, the examination terminal comprises an identity verification module, which is configured to verify the identity of the examinee. Only the examinee who passes the verification can enter the system and use the examination terminal. The identity verification module can be a second image acquisition module, which is configured to acquire the facial information of the examinee and compare the acquired facial information with the information stored in the server terminal. Only the examinee who passes the comparison can enter the examination terminal. The identity verification module can also be a user login module,

[0030] For the security of information data, the examinee of the examination terminal must enter the system through the account and the corresponding password assigned by the system administrator. The examinee who does not perform the login operation cannot enter the system and perform any system operation. The examinee who performs the login operation and can enter the system can perform various operations and can only perform the operations within the examinee's authority.

[0031] Further, the server terminal comprises a first image processing module, which is configured to process the human body image by the following method, as shown in FIG. 2:

[0032] S1: performing video frame extraction processing on the examination monitoring video;

[0033] S2: constructing an examination student irregular behavior recognition prediction model based on a deep learning algorithm;

[0034] S3: inputting the video frame extracted in step S1 into the recognition prediction model for recognition processing;

[0035] S4: performing examinee irregular behavior recognition prediction on the input video frame in time sequence. If the prediction model detects that the current video frame has examinee irregular behavior, the examinee is marked in the video frame, the image is saved to the local, and then the next video frame is repeated step S3. If the prediction model does not detect that the current video frame has examinee irregular behavior, the next video frame is directly repeated step S3.

[0036] The deep learning method in step S2 specifically comprises the following steps:

[0037] S21: sorting all frames according to the confidence score, wherein the frame is a portrait acquisition frame;

[0038] S22: calculating the IOU of the frame with the largest confidence score and the adjacent frame, wherein the calculation formula of the IOU is as follows:

[0039] wherein a and b respectively represent two adjacent frames, IOU represents the overlapping ratio of the two frames, and S() is an area function;

[0040] S23: comparing the calculated IOU with a threshold value, if the IOU is less than the threshold value, the confidence score is subjected to a linear operation, and is used as a new confidence score to participate in the next round of competition until all target regions are found, the linear decay non-maximum suppression algorithm can be expressed by the formula:

[0041] Wherein, F conf is the confidence score, is the confidence score after joining the linear function for smoothing, t is the class label, Bt is the candidate box to be compared, A is the selected candidate box with the largest current confidence score, IOU(A, B t ) is the intersection over union of A and Bt, and M represents the selected threshold value.

[0042] Further, the server terminal further comprises a test paper processing module, which is configured to compare the test paper uploaded by the examinee with a standard test paper with answers, mark errors in the test paper, and then obtain the score of the test paper.

[0043] The display terminal comprises an administrator login module, which is configured to enable the invigilator to enter the display terminal through secret key authentication, the secret key authentication adopts a U disk certificate and authentication password encryption mode, when the invigilator enters the display terminal, if no secret key is inserted into the display terminal, the system refuses to enter; only after the secret key is inserted and the administrator password is input at the same time, the invigilator can enter the display terminal to perform various operations.

[0044] In addition, in order to ensure the safety of system and database information, the system sets different system permissions for different users, the more the contents of the permissions, the more functions of the system can be used, the system creates various user permission templates for various users, such as information query personnel and ordinary administrators. The ordinary administrator only has the functions of data entry, saving data, printing score sheet, modifying own password, etc. The information query personnel only has the functions of querying data, printing queried data, and modifying own password, etc. The difference between the ordinary administrator and the administrator is that the administrator cannot create, modify, and delete user information.

[0045] In summary, the monitoring system comprises a first image acquisition module, an examination terminal, and a server terminal, the examinee is monitored through the first image acquisition module, and image information is transmitted to the server terminal for processing, the server terminal performs identification processing through an identification prediction model, can accurately and timely discover disciplinary violations, and marks the disciplinary violations to the display terminal for display, the invigilator processes the disciplinary personnel according to the relevant information.

Claims

1. An examination room monitoring system based on artificial intelligence, characterized in that... include: The first image acquisition module is used to acquire human images of candidates in the examination room and upload the acquired image data to the server terminal for data processing. The examination terminal is used to verify the identity of examinees and distribute examination papers, and transmit the completed examination papers to the server terminal for processing. The server terminal is used to receive data uploaded by the image acquisition module and the examination terminal, and to process the uploaded data accordingly. The image acquisition module monitors the examinees, processes the acquired human images to identify whether the examinees are cheating, and processes the test paper information uploaded by the examination terminal to obtain the test score. The display terminal is used by invigilators. It transmits data to the server terminal via a wireless network. The data processed by the server terminal is transmitted to the display terminal via the wireless network for display. Invigilators then take appropriate actions based on the displayed content.

2. The examination room monitoring system based on artificial intelligence as described in claim 1, characterized in that: The examination terminal includes an identity verification module, which verifies the identity of the examinee. Only examinees who pass the verification can enter the system and use the examination terminal.

3. The examination room monitoring system based on artificial intelligence as described in claim 1, characterized in that: The server terminal includes a first image processing module, which processes human body images using the following method: S1: Perform frame extraction processing on the exam monitoring video; S2: Construct a prediction model for identifying student violations during exams based on deep learning algorithms; S3: Input the video frames extracted in step S1 into the recognition and prediction model for recognition processing; S4: Predict and identify student misconduct in the input video frames sequentially. If the prediction model detects misconduct in the current video frame, it marks the student in that frame and saves the image locally. Then, it repeats step S3 for the next video frame. If the prediction model does not detect misconduct in the current video frame, it directly proceeds to the next video frame. Repeat step S3.

4. The examination room monitoring system based on artificial intelligence as described in claim 1, characterized in that: The server terminal also includes a test paper processing module, which is used to compare the test paper uploaded by the examinee with a standard test paper with answers, mark the errors in the test paper, and then obtain the score of the test paper.

5. The examination room monitoring system based on artificial intelligence as described in claim 1, characterized in that: The display terminal includes an administrator login module, which allows invigilators to access the display terminal via key authentication. Key authentication uses a USB flash drive certificate plus authentication password encryption method. When an invigilator enters the display terminal, if no key is inserted on the display terminal, the system will refuse entry; only after inserting the key and entering the administrator password can the invigilator enter the display terminal to perform various operations.

Images