Method for monitoring security checkpoint personnel, server and storage medium

Abstract

The security inspection machine guard monitoring method, the server and the storage medium provided by the embodiment of the application, by acquiring the monitoring video collected by the monitoring camera installed in the monitoring place, inputting the first target frame of the monitoring video into the device detection model obtained by pre-training, obtaining the device detection result, when the first device detection result includes the security inspection machine and the display device, judging whether the working state of the security inspection machine is normal according to the on-off state of the display device, when the working state of the security inspection machine is normal, demarcating the seat range of the guard in the first target frame to obtain a range demarcated image, inputting the range demarcated image into the human body target detection model, judging whether there is a guard in the seat range, when there is no guard in the seat range, outputting the first alarm information to prompt the management personnel that the security inspection machine is unattended, realizing the automatic monitoring of the working state of the guard, and improving the efficiency of the supervision work.

Description

TECHNICAL FIELD

[0001] The present application relates to the intelligent video analysis technology, and particularly to a security inspection machine guard monitoring method, a server and a storage medium. BACKGROUND

[0002] The intelligent video monitoring technology is to filter out useless or interfering information in the video picture by means of the powerful data processing capability of the computer, automatically analyze and extract the key useful information in the video source, and perform alarm or other actions.

[0003] The intelligent video monitoring technology is also widely used in the scene of deploying security inspection machines, and is mainly used for monitoring the guards of the security inspection machines. The existing security inspection machine guard monitoring method can determine the working state of the guard, such as whether to be off-duty or lazy at work, only by manually monitoring or collecting the X-ray information generated by the X-ray imaging device in the security inspection machine and the related operation information of the guard obtained by the image shooting device.

[0004] The manual monitoring method is not only labor-intensive and inefficient for large work places such as express sorting centers where the guards work for a long time, but also needs to modify the hardware of the security inspection machine to monitor the guards by collecting the related information of the security inspection machine, which is high in cost and difficult to implement. SUMMARY

[0005] The present application provides a security inspection machine guard monitoring method, a server and a storage medium to solve the problem of high labor and cost in monitoring the guards of the security inspection machine.

[0006] In a first aspect, the present application provides a security inspection machine guard monitoring method, comprising:

[0007] Real-time acquisition of a monitoring video collected by a monitoring camera installed in a monitoring place;

[0008] Inputting a first target frame of the monitoring video into a pre-trained device detection model to obtain a first device detection result;

[0009] When the first device detection result includes a security inspection machine and a display device, judging whether the working state of the security inspection machine is normal according to the on-off state of the display device, and when the working state of the security inspection machine is normal, marking a seat range of the guard in the first target frame to obtain a range marked image, the display device including a monitoring screen and / or a signal light;

[0010] input the range calibration image into a human target detection model to obtain a human detection result, and when the human detection result indicates that no guard personnel is detected in the seat range, output first alarm information, the first alarm information being used to prompt a manager that the security inspection machine is unattended.

[0011] Optionally, when the working state of the security inspection machine is abnormal and an article is detected at the entrance of the security inspection machine, input a second target frame of the monitoring video into the equipment detection model to obtain a second equipment detection result, the second target frame being after the first target frame.

[0012] When the second equipment detection result includes the article, the display equipment and the security inspection machine, and the article is at the exit of the security inspection machine, it is determined that the article enters or exits the security inspection machine.

[0013] Output second alarm information, the second alarm information being used to prompt the guard personnel to power on the security inspection machine.

[0014] Optionally, the method further comprises:

[0015] When the human detection result indicates that the guard personnel is detected in the seat range and the display equipment includes the monitoring video, calculate an actual distance between the guard personnel and the monitoring screen.

[0016] When the actual distance is greater than a preset distance, output prompt information, the prompt information being used to prompt the guard personnel that the distance between the guard personnel and the monitoring screen is far and the work is not standardized.

[0017] Optionally, the first equipment detection result includes coordinates of the monitoring screen, and the calculation of the actual distance between the guard personnel and the monitoring screen comprises:

[0018] input the range calibration image into a pre-trained key point detection model to obtain human key point coordinates of the guard personnel;

[0019] According to the human key point coordinates, the coordinates of the monitoring screen, a preset upper body height of the guard personnel, a preset width of the monitoring screen and an imaging relationship between an image physical coordinate system and a camera coordinate system, the actual distance between the guard personnel and the monitoring screen is calculated.

[0020] Optionally, the method further comprises:

[0021] Obtain training data, the training data including equipment images of a security inspection machine and equipment labels, the equipment labels being used to describe types corresponding to the equipment images;

[0022] According to the training data, a preset model is trained to obtain the equipment detection model.

[0023] In a second aspect, the present application provides a server, comprising:

[0024] an acquisition module, configured to acquire, in real time, a monitoring video collected by a monitoring camera installed in a monitoring place;

[0025] a first processing module, configured to input a first target frame of the monitoring video into a device detection model pre-trained to obtain a first device detection result;

[0026] a second processing module, configured to, when the first device detection result includes a security inspection machine and a display device, determine whether the security inspection machine is in a normal working state according to a switching state of the display device, and when the security inspection machine is in the normal working state, demarcate a seat range of a guard in the first target frame to obtain a range demarcated image, the display device including a monitoring screen and / or a signal lamp;

[0027] a third processing module, configured to input the range demarcated image into a human body target detection model to obtain a human body detection result, and when the human body detection result indicates that no guard is detected in the seat range, output a first alarm information, the first alarm information being used to prompt a management personnel that the security inspection machine is unattended.

[0028] Optionally, the second processing module is specifically configured to:

[0029] when the security inspection machine is not in the normal working state and an article is detected at an entrance of the security inspection machine, input a second target frame of the monitoring video into the device detection model to obtain a second device detection result, the second target frame being after the first target frame;

[0030] when the second device detection result includes the article, the display device and the security inspection machine, and the article is at an exit of the security inspection machine, it is determined that the article enters or exits the security inspection machine;

[0031] output a second alarm information, the second alarm information being used to prompt the guard to power on the security inspection machine.

[0032] Optionally, the third processing module is specifically configured to:

[0033] when the human body detection result indicates that a guard is detected in the seat range and the display device includes the monitoring video, calculate an actual distance between the guard and the monitoring screen;

[0034] when the actual distance is greater than a preset distance, output a prompt information, the prompt information being used to prompt that the guard is far away from the monitoring screen and is not in a standard working state.

[0035] In a third aspect, the present application provides a server, comprising a processor and a memory connected with the processor;

[0036] The memory stores computer-executable instructions.

[0037] The processor executes the computer-executable instructions stored in the memory to implement the method according to the first aspect.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method for monitoring a security-checking machine guard according to the first aspect.

[0039] In a fifth aspect, the present application provides a computer program product, comprising a computer program, and the computer program is executed by a processor to implement the method according to the first aspect.

[0040] The method for monitoring a security-checking machine guard, the server and the storage medium provided by the present application can realize automatic monitoring of the working state of the guard by inputting a first target frame of a monitoring video collected by a monitoring camera installed in a monitoring place into a device detection model pre-trained to obtain a first device detection result, judging whether the working state of the security-checking machine is normal according to the on-off state of a display device when the security-checking machine and the display device are included in the first device detection result, calibrating a seat range of the guard in the first target frame to obtain a range calibration image when the working state of the security-checking machine is normal, inputting the range calibration image into a human target detection model, judging whether there is a guard in the seat range, determining that the security-checking machine is unattended when there is no guard in the seat range, and outputting a first alarm information to prompt a management personnel that the security-checking machine is unattended, thereby improving the efficiency of supervision work. BRIEF DESCRIPTION OF DRAWINGS

[0041] The accompanying drawings, which are incorporated herein and constitute part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0042] Figure 1 A schematic diagram of an application scenario applicable to the present application;

[0043] Figure 2 A flowchart of a method for monitoring a security-checking machine guard provided by Embodiment One of the present application;

[0044] Figure 3 A schematic diagram of a monitoring screen coordinate point and a human key point provided by Embodiment One of the present application;

[0045] Figure 4A structural schematic diagram of a server provided for Embodiment Two of the present application;

[0046] Figure 5 A structural schematic diagram of a server provided for Embodiment Three of the present application.

[0047] The specific embodiments of the present application have been shown in the above-described drawings, and will be described in more detail hereinafter. These drawings and the written description are not intended to restrict the scope of the present application in any way, but to illustrate the concept of the present application for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0048] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The same numbers are used in different drawings to represent the same or similar elements. The embodiments described in the following exemplary embodiments are not meant to represent all implementations consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as detailed in the appended claims.

[0049] The existing monitoring method of the guard personnel of the security inspection machine is to determine the working state of the guard personnel, such as whether to be off-duty or lazy at work, by manually monitoring or collecting the X-ray information generated by the X-ray imaging device in the security inspection machine and the relevant operation information of the guard personnel obtained by the image shooting device. The manual monitoring method is not only labor-intensive, but also inefficient for large workplaces such as express sorting centers where the guard personnel work for a long time. The monitoring of the guard personnel by collecting the relevant information from the imaging device of the security inspection machine requires hardware modification of the security inspection machine, which is high in cost and difficult to implement.

[0050] In view of the above technical problems of the prior art, the present application provides a security inspection machine guard personnel monitoring method, a server and a storage medium. By analyzing the monitoring video collected by the monitoring camera installed in the monitoring place, the present application does not need to modify the device of the security inspection machine itself, realizes automatic monitoring of the working state of the guard personnel, improves the efficiency of supervision work and reduces the manpower and cost of supervision work.

[0051] Figure 1 A schematic diagram of an application scenario applicable to the present application. As Figure 1As shown, the camera 101 sends the collected monitoring video to the server 102 through the Internet, the server 102 analyzes the security inspection machine 103, the monitoring screen 104 and the guard 105 in the monitoring video to supervise the working state of the guard 105, the camera 101 is installed in the monitoring place 106 to monitor the monitoring place 106, the application does not limit the installation angle of the camera 101, and the monitoring screen 104 can display the working state of the security inspection machine 103. It can be understood that the number of the camera 101, the server 102, the security inspection machine 103, the monitoring screen 104 and the guard 105 can be multiple, which are not shown in the figure.

[0052] The technical solutions of the application and how the technical solutions of the application solve the above technical problems will be described in detail below with specific examples. The following specific examples can exist independently or can be combined with each other, and the same or similar concepts or processes can not be described in detail in some examples. The embodiments of the application will be described below with reference to the drawings.

[0053] Reference Figure 2 , Figure 2 A flowchart of a guard monitoring method of a security inspection machine provided by Embodiment One of the application, the method can be executed by a server, and the method comprises the following steps.

[0054] S201, real-time acquisition of monitoring video collected by a monitoring camera installed in a monitoring place.

[0055] The monitoring camera installed in the monitoring place can collect the monitoring video of the monitoring place in real time, and the monitoring video acquired by the server is composed of a large number of image frames, for example, 25 frames of images for 1 second of video.

[0056] S202, inputting a first target frame of the monitoring video into a pre-trained device detection model to obtain a first device detection result.

[0057] The device detection model is obtained by inputting training data into a preset model for multiple times of training by the server, the training data comprises a large number of device images and device labels of the security inspection machine, the device label is used to describe the type corresponding to the device image, for example, the label of the security inspection machine image can be "0", the label of the monitoring screen can be "1", the label of the signal lamp can be "2", and the label of the article can be "3", that is, "0" represents the security inspection machine, "1" represents the monitoring screen, "2" represents the signal lamp, and "3" represents the article. It should be noted that the article can be a parcel or other article that needs to be checked for safety.

[0058] The server inputs the first target frame of the obtained monitoring video into the pre-trained device detection model to obtain a first device detection result, and the first device detection result includes labels corresponding to the detected multiple targets, and the labels represent the categories of the targets. For example, when the label is "1", the detected target is a monitoring screen.

[0059] In S203, when the first device detection result includes the security inspection machine and the display device, it is determined whether the working state of the security inspection machine is normal according to the on-off state of the display device. When the working state of the security inspection machine is normal, the seat range of the guard in the first target frame is calibrated to obtain a range calibration image.

[0060] The first device detection result at least includes the security inspection machine and the display device of the security inspection machine. The display device can include a monitoring screen and / or a signal lamp, that is, the display device can be a monitoring screen or a signal lamp, or a monitoring screen and a signal lamp. The on-off state of the two devices generally represents the working state of the security inspection machine.

[0061] According to the on-off state of the display device, it is determined whether the working state of the security inspection machine is normal. The on-off state is that the display device is off or on. When the display device is off, that is, the monitoring screen or the signal lamp is off, it indicates that the working state of the security inspection machine is not normal. When the display device is on, that is, the monitoring screen or the signal lamp is not off, it indicates that the working state of the security inspection machine is normal.

[0062] When the working state of the security inspection machine is normal, the seat range of the guard in the first target frame is calibrated, that is, the seat range of the guard in the first target frame is framed to obtain a range calibration image. The seat range can be in front of the monitoring screen or on the side of the security inspection machine. The seat range can be calibrated according to the actual situation of the security inspection machine in the monitoring place, and the seat range is used to determine whether there is a guard.

[0063] Optionally, when the working state of the security inspection machine is not normal, and an article is detected at the entrance of the security inspection machine, the server inputs a second target frame of the monitoring video into the device detection model to obtain a second device detection result, and the second target frame is after the first target frame.

[0064] When the second device detection result includes the article, the display device, and the security inspection machine, it is determined that the article enters or exits the security inspection machine, and the article is at the outlet of the security inspection machine. That is, the distance between the first target frame and the second target frame corresponds to the frame number of the time when the article enters or exits the security inspection machine. For example, 1 second of video includes 26 frames of images, and the time when the article enters or exits the security inspection machine is 4 seconds, so the second target frame and the first target frame are different by 104 frames.

[0065] When the security inspection machine state is abnormal, but there are objects in and out of the security inspection machine, it is determined that the security inspection machine is not powered on, i.e., the guard personnel does not power on the security inspection machine, the server outputs a second alarm information, which can be in the form of a short message sent to the management personnel, or in the form of a voice to inform the management personnel, for prompting the guard personnel to power on the security inspection machine, and reminding the guard personnel, and the form of the server informing the management personnel of the second alarm information is not limited in the present application.

[0066] S204, input the range calibration image into the human target detection model to obtain a human detection result, and output a first alarm information when the human detection result indicates that no guard personnel is detected in the seat range.

[0067] The server inputs the range calibration image into the human target detection model, and since the existing human detection model has high accuracy, the human detection of the range calibration image can be detected by using the model to obtain a human detection result in the present embodiment. When the human detection result indicates that no guard personnel is detected in the seat range, i.e., the label corresponding to the guard personnel is not included in the detection result, the server outputs a first alarm information. The form of the server informing the management personnel of the first alarm information can refer to the second alarm information, prompting the management personnel that the security inspection machine is unattended, and the guard personnel needs to be reminded in time to avoid that if the dangerous object needs to be inspected, it is missed due to unattended when passing through the security inspection machine.

[0068] When the human detection result indicates that the guard personnel is detected in the seat range, and the display device includes a monitoring screen, the first device detection result further includes the coordinates of the monitoring screen, which are specifically the coordinates of two opposite corners (the upper left corner and the lower right corner) of the monitoring screen. It is necessary to determine whether the work of the guard personnel is standardized. The server calculates the actual distance between the guard personnel and the monitoring screen, and the specific process is as follows:

[0069] I. input the range calibration image into the pre-trained key point detection model to obtain the human key point coordinates of the guard personnel, which include the key points of the top of the head and the middle of the body of the guard personnel.

[0070] II. Calculate the upper body height of the guard personnel in the second output image according to the human key point coordinates of the guard personnel.

[0071] III. Calculate the actual distance between the guard personnel and the monitoring screen according to the human key point coordinates, the coordinates of the security inspection machine in the first output image, the preset upper body height of the guard personnel, the preset width of the security inspection machine, and the imaging relationship between the image physical coordinate system and the camera coordinate system.

[0072] For example, the actual distance between the guard personnel and the monitoring screen can be calculated by the following formula:

[0073]

[0074] wherein, ΔX is the actual distance between the guard and the monitoring screen in the horizontal direction, and ΔZ is the actual distance between the guard and the monitoring screen in the depth direction.

[0075]

[0076]

[0077] H is the preset upper body height of the guard, and W is the preset width of the monitoring screen, as shown in Figure 3 x a and x b are the horizontal coordinates of points A and B of the monitoring screen in the first output image, x c is the horizontal coordinate of the head top key point C in the human body key points, y c and y d are the vertical coordinates of the head top key point C and the human center key point D in the human body key points, respectively, and f is the focal length of the camera, generally in mm (millimeter).

[0078] The horizontal coordinate of the above-mentioned point A can be obtained according to the horizontal coordinate of point E in Figure 3 , the preset upper body height of the guard refers to the empirical value of the actual upper body height of the human body, the preset width of the monitoring screen refers to the empirical value of the actual width of the monitoring screen, for example, the preset actual upper body height of the guard can be 90 cm (centimeter), i.e. 900 mm, W is the preset actual width of the monitoring screen can be 80 cm, i.e. 800 mm, x a is 15 mm, x b is 5 mm, x c is 3 mm, y c is 25 mm, y d is 5 mm, and f is 20 mm, the calculated actual distance ΔX between the guard and the monitoring screen in the horizontal direction is 230 mm, the actual distance ΔZ between the guard and the monitoring screen in the depth direction is 100 mm, and the actual distance D between the guard and the monitoring screen is 251 mm, i.e. 25.1 cm.

[0079] When the actual distance between the guard and the monitoring screen is greater than the preset distance, the server outputs a prompt information, and the server can notify the management personnel in the form of the second alarm information, prompting that the distance between the guard and the monitoring screen is far, and the work is not standardized, on the contrary, when the actual distance between the guard and the monitoring screen is less than or equal to the preset distance, it indicates that the distance between the guard and the monitoring screen is within the preset range.

[0080] For example, when the preset distance is 1 meter, i.e. 100 cm, and the actual distance between the detected guard and the monitoring screen is 25.1 cm, it indicates that the distance between the guard and the monitoring screen is within the preset range, and the guard does not appear to be in an irregular working condition.

[0081] In the embodiment, the monitoring video collected by the monitoring camera installed in the monitoring place is acquired in real time, a first target frame of the monitoring video is input into the device detection model pre-trained to obtain a first device detection result, when the first device detection result includes the security inspection machine and the display device, whether the working state of the security inspection machine is normal is determined according to the on-off state of the display device, when the working state of the security inspection machine is normal, the seat range of the guard in the first target frame is calibrated to obtain a range calibration image, the range calibration image is input into the human body target detection model to determine whether there is a guard in the seat range, when there is no guard in the seat range, it is determined that the security inspection machine is unattended, and the first alarm information is output to prompt the management personnel that the security inspection machine is unattended, thereby automatically monitoring the working state of the guard and improving the efficiency of supervision.

[0082] Reference Figure 4 , Figure 4 A structure schematic diagram of a server provided in Embodiment Two of the present application is shown in FIG. 4. As shown in FIG. 4, the server 40 includes an acquisition module 401, a first processing module 402, a second processing module 403 and a third processing module 404. Figure 4

[0083] The acquisition module 401 is configured to acquire in real time the monitoring video collected by the monitoring camera installed in the monitoring place.

[0084] The first processing module 402 is configured to input a first target frame of the monitoring video into the device detection model pre-trained to obtain a first device detection result.

[0085] The second processing module 403 is configured to, when the first device detection result includes the security inspection machine and the display device, determine whether the working state of the security inspection machine is normal according to the on-off state of the display device, when the working state of the security inspection machine is normal, calibrate the seat range of the guard in the first target frame to obtain a range calibration image, and the display device includes the monitoring screen and / or the signal lamp.

[0086] The third processing module 404 is configured to input the range calibration image into the human body target detection model to obtain a human body detection result, and output the first alarm information when the human body detection result indicates that no guard is detected in the seat range, and the first alarm information is used to prompt the management personnel that the security inspection machine is unattended.

[0087] Optionally, the second processing module 403 is specifically configured to:

[0088] ​When the working state of the security inspection machine is abnormal, and the security inspection machine detects the article at the entrance, the second target frame of the monitoring video is input into the device detection model to obtain a second device detection result, and the second target frame is after the first target frame.

[0089] When the second device detection result includes the article, the display device and the security inspection machine, and the article is at the outlet of the security inspection machine, it is determined that the article enters and exits the security inspection machine.

[0090] The second alarm information is output, and the second alarm information is used to prompt the guard personnel to power on the security inspection machine.

[0091] Optionally, the third processing module 404 is specifically configured to:

[0092] When the human body detection result indicates that the guard personnel is detected in the seat range, and the display device includes the monitoring video, the actual distance between the guard personnel and the monitoring screen is calculated.

[0093] When the actual distance is greater than the preset distance, the output prompt information is used to prompt the guard personnel that the distance from the monitoring screen is far and the work is not standardized.

[0094] Optionally, the first device detection result includes the coordinates of the monitoring screen, and the actual distance between the guard personnel and the monitoring screen is calculated, including:

[0095] The range calibration image is input into the key point detection model to obtain the human body key point coordinates of the guard personnel.

[0096] According to the human body key point coordinates, the coordinates of the monitoring screen, the preset upper body height of the guard personnel, the preset width of the monitoring screen and the imaging relationship between the image physical coordinate system and the camera coordinate system, the actual distance between the guard personnel and the monitoring screen is calculated.

[0097] Optionally, it further includes:

[0098] Obtain training data, the training data including device images and device labels of the security inspection machine, and the device labels are used to describe the types corresponding to the device images.

[0099] According to the training data, the preset model is trained to obtain the device detection model.

[0100] The server of the embodiment can be used to execute the security inspection machine guard personnel monitoring method of the first embodiment, and the specific implementation manner and technical effects are similar, which will not be described here.

[0101] Reference Figure 5 , Figure 5 A structural schematic diagram of a server provided by the third embodiment of the application is shown in FIG. 4. Figure 5As shown, the server 50 comprises a processor 501, a memory 502, and a transceiver 503. The processor 501 executes computer-executable instructions stored in the memory 502 and controls receiving and transmitting actions of the transceiver 503, so that the at least one processor performs steps of a method for monitoring a guard of a security inspection machine according to an embodiment. The specific implementation and technical effects are similar, and will not be repeated here.

[0102] Embodiment four of the present application provides a computer-readable storage medium, which stores computer-executable instructions. The computer-executable instructions, when executed by a processor, are used to implement steps of a method for monitoring a guard of a security inspection machine according to embodiment one. The specific implementation and technical effects are similar, and will not be repeated here.

[0103] Embodiment five of the present application provides a computer program product, which comprises a computer program. The computer program, when executed by a processor, implements steps of a method for monitoring a guard of a security inspection machine according to embodiment one. The specific implementation and technical effects are similar, and will not be repeated here.

[0104] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. The specification and examples given are exemplary only and the true scope and spirit of the application are indicated by the following claims. It will be appreciated by those skilled in the art that changes could be made to the application described above, without departing from the broad discipline concept thereof. It is intended to be covered by the specification and claims.

[0105] It should be understood that the application is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the appended claims.

Claims

1. A method for monitoring personnel on duty at a security inspection machine, characterized in that, Applied to servers, including: Real-time acquisition of surveillance video collected by surveillance cameras installed in the monitored area; The first target frame of the surveillance video is input into the pre-trained device detection model to obtain the first device detection result; When the first device detection result includes a security inspection machine and a display device, the working status of the security inspection machine is determined according to the on / off status of the display device. When the working status of the security inspection machine is normal, the seating range of the on-duty personnel is marked in the first target frame to obtain a range marking image. The display device includes a monitoring screen and / or signal lights. The range calibration image is input into the human target detection model to obtain the human detection result. When the human detection result indicates that no staff member is detected within the seat area, a first alarm message is output. The first alarm message is used to remind the management personnel that the security inspection machine is unattended. When the security inspection machine is not working properly and an item is detected at the entrance of the security inspection machine, the second target frame of the monitoring video is input into the device detection model to obtain the second device detection result. The second target frame is after the first target frame. When the detection result of the second device includes the item, the display device, and the security inspection machine, and the item is at the exit of the security inspection machine, it is determined that an item has entered or exited the security inspection machine; Output a second alarm message, which is used to remind the on-duty personnel that the security inspection machine has not been powered on.

2. The method according to claim 1, characterized in that, Also includes: When the human detection result indicates that a staff member is detected within the seating area, and the display device includes the monitoring screen, the actual distance between the staff member and the monitoring screen is calculated. When the actual distance is greater than the preset distance, a prompt message is output. The prompt message is used to remind the on-duty personnel that the distance between them and the monitoring screen is too far and that their work is not in accordance with regulations.

3. The method according to claim 2, characterized in that, The first device detection result also includes the coordinates of the monitoring screen, and the calculation of the actual distance between the on-duty personnel and the monitoring screen includes: The range calibration image is input into the key point detection model to obtain the coordinates of the human body key points of the guard. Based on the coordinates of the key points of the human body, the coordinates of the monitoring screen, the preset upper body height of the on-duty personnel, the preset width of the monitoring screen, and the imaging relationship between the physical coordinate system of the image and the camera coordinate system, the actual distance between the on-duty personnel and the monitoring screen is calculated.

4. The method according to claim 1, characterized in that, Also includes: Acquire training data, which includes equipment images and equipment tags of the security inspection machine, wherein the equipment tags are used to describe the type corresponding to the equipment images; The device detection model is obtained by training the preset model based on the training data.

5. A monitoring device for personnel on duty at a security inspection machine, characterized in that, include: The acquisition module is used to acquire surveillance video collected by surveillance cameras installed in the monitored area in real time. The first processing module is used to input the first target frame of the surveillance video into the pre-trained device detection model to obtain the first device detection result; The second processing module is used to determine whether the working status of the security inspection machine is normal based on the on / off status of the display device when the detection result of the first device includes a security inspection machine and a display device. When the working status of the security inspection machine is normal, the module marks the seating range of the on-duty personnel in the first target frame to obtain a range marking image. The display device includes a monitoring screen and / or signal lights. The third processing module inputs the range calibration image into the human target detection model to obtain the human detection result. When the human detection result indicates that no staff member is detected within the seat range, the first alarm message is output. The first alarm message is used to remind the management personnel that the security inspection machine is unattended. The second processing module is specifically used for: When the security inspection machine is not working properly and an item is detected at the entrance of the security inspection machine, the second target frame of the monitoring video is input into the device detection model to obtain the second device detection result. The second target frame is after the first target frame. When the detection result of the second device includes the item, the display device, and the security inspection machine, and the item is at the exit of the security inspection machine, it is determined that an item has entered or exited the security inspection machine; Output a second alarm message, which is used to remind the on-duty personnel that the security inspection machine has not been powered on.

6. The apparatus according to claim 5, characterized in that, The third processing module is specifically used for: When the human detection result indicates that a staff member is detected within the seating area, and the display device includes the surveillance video, the actual distance between the staff member and the monitoring screen is calculated; When the actual distance is greater than the preset distance, a prompt message is output. The prompt message is used to remind the duty personnel that the distance between them and the monitoring screen is too far and that their work is not in accordance with regulations.

7. A server, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-4.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement a method for monitoring personnel on duty at a security inspection machine as described in any one of claims 1 to 4.

Images