Multi-view dynamic monitoring screen data leakage prevention alarm method, system and terminal

Abstract

This application relates to a multi-view dynamic monitoring screen anti-leakage alarm method, system, and terminal, belonging to the technical field of information security. The method includes: acquiring video frames in front of the screen, the brightness of the screen's reflective area, and the ambient background brightness; calculating the device shape matching degree of a suspicious device based on the video frames; determining whether the device shape matching degree is greater than a first threshold; if so, obtaining the center coordinates of the suspicious device and the center coordinates of the screen's reflective area; calculating the Euclidean distance based on the center coordinates of both; tracking the movement speed of the suspicious object in consecutive Z frames based on the video frames; calculating the reflectivity coefficient of consecutive Z frames based on the brightness of the screen's reflective area and the ambient background brightness, and calculating the motion reflectivity correlation coefficient; determining whether the Euclidean distance is less than a second threshold and the motion reflectivity correlation coefficient is greater than a third threshold; if so, outputting a first threat value and triggering a level one alarm. This application has the beneficial effect of ensuring that screen information is not illegally photographed or stolen.

Description

Technical Field

[0001] This application relates to the technical field of information security, and in particular to a screen anti-leakage alarm method, system and terminal for multi-view dynamic monitoring. Background Technology

[0002] The widespread use of mobile smart terminals has brought great convenience to people's lives, but drawbacks have also emerged: the security of sensitive information in the final presentation stage is difficult to be effectively guaranteed by prior protection measures.

[0003] With the increasing prevalence of portable devices such as mobile phones equipped with camera and video recording functions, using these devices to photograph or record computer screen content to steal information has become one of the most common and easily implemented methods of information leakage. In recent years, incidents of information leaks caused by mobile phone screen photography have occurred frequently.

[0004] Therefore, ensuring that screen information is not illegally photographed or stolen has become a critical issue that urgently needs to be addressed in the field of information security. Summary of the Invention

[0005] To ensure that screen information is not illegally photographed or stolen, this application provides a screen anti-leakage alarm method, system, and terminal with multi-view dynamic monitoring.

[0006] Firstly, this application provides a multi-view dynamic monitoring method for screen data leakage prevention alarms, employing the following technical solution:

[0007] A multi-view dynamic monitoring method for screen data leakage prevention alarms includes:

[0008] Capture video frames in front of the screen, brightness of the screen's reflective area, and ambient background brightness;

[0009] Calculate the device shape matching degree of the suspicious device based on the video frames;

[0010] Determine whether the device shape matching degree is greater than a first threshold;

[0011] If so, obtain the center coordinates of the suspicious device and the center coordinates of the reflective area on the screen;

[0012] Calculate the Euclidean distance based on the center coordinates of the suspected device and the center coordinates of the reflective area of ​​the screen;

[0013] Based on the video frames, track the movement speed of the suspicious object in consecutive Z frames;

[0014] The reflectivity coefficient of consecutive Z frames is calculated based on the brightness of the screen reflective area and the ambient background brightness.

[0015] Calculate the motion reflection correlation coefficient based on the moving speed and the reflectivity coefficient;

[0016] Determine whether the Euclidean distance is less than a second threshold and whether the action reflection correlation coefficient is greater than a third threshold;

[0017] If so, output the first threat value and trigger a level 1 alert.

[0018] By employing the aforementioned technical solution, video frames in front of the screen are captured and the shape matching degree of suspicious devices is calculated, enabling accurate identification of objects that match the preset shape characteristics of suspicious devices. After identifying a suspicious device, a comprehensive judgment is made by combining multiple features such as the Euclidean distance between the device's center coordinates and the center coordinates of the reflective area, the moving speed of the suspicious object, and the reflectivity coefficient. These features reflect the behavior and state of the suspicious device from different perspectives. A single feature cannot accurately determine whether there is a risk of information leakage, but multi-feature fusion can greatly improve the accuracy of the judgment. In real-world scenarios, information theft is often not instantaneous but has a certain time process and behavioral trajectory. Tracking suspicious objects in continuous Z-frame video allows for real-time monitoring of their dynamic changes. Through continuous frame tracking, information such as the moving speed and direction of movement of suspicious objects can be analyzed to determine whether they are engaging in purposeful theft. Real-time calculation of the reflectivity coefficient of continuous Z-frames allows for timely detection of reflectivity changes caused by the operation of suspicious devices. After completing the calculation and judgment of various features, a threat value can be quickly output, and a level one alarm can be triggered when conditions are met. Once a high risk of data leakage is detected, the system can immediately issue an alarm signal to notify relevant personnel to take measures to effectively prevent information leakage incidents from occurring.

[0019] Optionally, the steps following determining whether the Euclidean distance is less than a second threshold and whether the motion reflection correlation coefficient is greater than a third threshold further include:

[0020] If not, then it is further determined whether the reflectivity coefficient is greater than the fourth threshold.

[0021] If so, calculate the second threat value according to the distance compensation formula;

[0022] If not, calculate the motion trajectory mutation index based on the video frames;

[0023] Based on the device shape matching degree, the reflectivity coefficient, and the motion trajectory abruptness index, a third threat value is output;

[0024] Determine whether the second threat value or the third threat value is greater than the fifth threshold;

[0025] If so, a Level 1 alarm will be triggered;

[0026] If not, a level 2 alarm is triggered; the severity of the level 1 alarm is greater than the severity of the level 2 alarm.

[0027] By adopting the above technical solution, the risk assessment process is refined by progressively judging different thresholds. First, the correlation coefficient between Euclidean distance and motion reflection is determined. If the condition is not met, the reflection intensity coefficient is further judged, and second and third threat values ​​are calculated separately according to different situations. This phased assessment method allows for more detailed analysis of various scenarios, improving the accuracy of risk assessment. Tiered alarms avoid issuing the highest-level alarm in all situations, thereby reducing unnecessary resource waste.

[0028] Optionally, the steps following the triggering of the Level 1 alarm include:

[0029] Identify sensitive areas on the screen and generate a mask matrix;

[0030] Calculate the primary background color of the sensitive area;

[0031] Based on the background primary color, generate interference colors;

[0032] Determine whether the mask matrix is ​​1;

[0033] If so, then according to the interference color, the sensitive region is perturbed over time t.

[0034] By employing the above technical solution, after triggering a Level 1 alarm, sensitive areas on the screen are identified and a mask matrix is ​​generated. This allows for precise location of information areas requiring protection, enabling the system to target sensitive content and prevent leakage. Calculating the primary background color of the sensitive area and generating interference colors based on it is an effective method of information interference. The interference colors can blend with the primary background color, making the content of the sensitive area visually difficult to identify, thus increasing the difficulty of illegally obtaining sensitive information. Even if someone attempts to illegally obtain screen information, the presence of interference colors will make it difficult to distinguish the content, thereby improving information security. When the mask matrix is ​​1, it indicates that the sensitive area needs protection. Therefore, the interference colors can be superimposed on the sensitive area over time t, further enhancing the visual confusion effect. Over time, the superposition of interference colors will cause the content of the sensitive area to continuously change, making it difficult to continuously observe and record.

[0035] Optionally, the steps following the triggering of the Level 1 alarm may further include:

[0036] Capture the frequency bands of Wi-Fi or Bluetooth signals from suspicious devices;

[0037] Generate a noise signal with the same frequency as the Wi-Fi or Bluetooth signal of the suspected device;

[0038] The control screen periodically flashes LEDs and emits the noise signal.

[0039] By employing the aforementioned technical solution, the frequency bands of suspicious devices' Wi-Fi or Bluetooth signals are captured, and then a noise signal at the same frequency is generated, enabling targeted interference with the communication of suspicious devices. Because Wi-Fi and Bluetooth communication rely on specific frequency bands for data transmission, the noise signal at the same frequency disrupts their normal signal transmission, making it difficult for suspicious devices to stably receive or send information, thus effectively preventing potential illegal data theft, malicious control, and other such activities. Combined with the emitted noise signal and the flashing LED on the screen, a comprehensive warning effect is created; it conveys a sense of danger not only visually but also through electromagnetic interference. This multi-dimensional warning method can more effectively raise public awareness and improve the timeliness and effectiveness of threat response.

[0040] Optionally, the steps following the triggering of the Level 1 alarm may further include:

[0041] Floating text is overlaid in the sensitive area, and the font transparency is dynamically adjusted;

[0042] Generate and transmit ultrasonic watermark signals.

[0043] By employing the aforementioned technical solutions, superimposing floating text on sensitive areas and dynamically adjusting font transparency can further conceal and obscure sensitive information. The presence of floating text visually interferes with attempts to spy on sensitive information, making it difficult for others to clearly and accurately identify sensitive content. Dynamically adjusting transparency increases the complexity and randomness of this interference, making it impossible for someone to reliably obtain sensitive information even after prolonged screen observation, significantly improving the security of sensitive information. Ultrasonic watermark signals can carry specific security information, such as device identification and security levels. This watermark information can provide additional security for sensitive information without affecting normal business operations. In the event of an information leak, the source and propagation path of the information can be traced by detecting the ultrasonic watermark signal, helping to promptly discover and address security vulnerabilities, while also deterring potential information thieves.

[0044] Optionally, the screen anti-leakage alarm method further includes:

[0045] When a Level 1 alarm is triggered, determine whether the threat value remains below the minimum threat threshold for a set duration.

[0046] If so, it enters a 10-second countdown state and gradually reduces the interference density.

[0047] By employing the above technical solution, when the threat value remains consistently below the minimum threat threshold within a set time limit, it indicates that there is likely no significant risk of data leakage. At this point, entering a 10-second countdown state and gradually decreasing the interference density can prevent continuous high-intensity interference operations.

[0048] Optionally, the steps following the triggering of the secondary alarm include:

[0049] The control screen displays red triangle icons in the four corners and controls the red triangle icons to flash at a frequency of 2Hz.

[0050] Secondly, this application provides a multi-view dynamic monitoring screen anti-leakage alarm system, which adopts the following technical solution:

[0051] A multi-view dynamic monitoring screen leak prevention alarm system includes:

[0052] The data acquisition module is used to collect video frames in front of the screen, the brightness of the screen's reflective area, and the ambient background brightness;

[0053] The data processing module is used to calculate the device shape matching degree of the suspicious device based on the video frames;

[0054] The judgment module is used to determine whether the shape matching degree of the device is greater than a first threshold. If so, the data acquisition module is used to obtain the center coordinates of the suspicious device and the center coordinates of the reflective area. The data processing module is used to calculate the Euclidean distance based on the two center coordinates.

[0055] The data acquisition module is also used to track the movement speed of a suspicious object in consecutive Z frames based on the video frames. The data processing module is used to calculate the reflectivity coefficient of consecutive Z frames based on the brightness of the screen reflective area and the ambient background brightness, and then calculate the motion reflectivity correlation coefficient based on the movement speed and the reflectivity coefficient.

[0056] The judgment module is used to determine whether the Euclidean distance is less than the second threshold and whether the action reflection correlation coefficient is greater than the third threshold.

[0057] The alarm module is used to output a first threat value and trigger a level one alarm when the judgment module determines that it is a threat.

[0058] Thirdly, this application provides a terminal that adopts the following technical solution:

[0059] A terminal, comprising:

[0060] The memory stores screen anti-leakage alarm programs with multi-view dynamic monitoring.

[0061] A processor is used to execute a program stored in the memory to implement the steps of the above-described multi-view dynamic monitoring screen anti-leakage alarm method.

[0062] In summary, this application has at least the following beneficial effects:

[0063] By capturing video frames in front of the screen and calculating the shape matching degree of suspicious devices, objects matching the preset shape characteristics of suspicious devices can be accurately identified. After identifying a suspicious device, a comprehensive judgment is made by combining multiple features such as the Euclidean distance between the device's center coordinates and the center coordinates of the reflective area, the movement speed of the suspicious object, and the reflectivity coefficient. These features reflect the behavior and state of the suspicious device from different perspectives. A single feature cannot accurately determine whether there is a risk of information leakage, but multi-feature fusion can greatly improve the accuracy of the judgment. In real-world scenarios, the act of stealing information is often not instantaneous, but has a certain time process and behavioral trajectory. Tracking suspicious objects in continuous Z-frame video can monitor their dynamic changes in real time. Through continuous frame tracking, information such as the movement speed and direction of movement of suspicious objects can be analyzed to determine whether they are engaging in purposeful theft. Real-time calculation of the reflectivity coefficient of continuous Z-frames can promptly detect changes in reflectivity caused by the operation of suspicious devices. After completing the calculation and judgment of various features, a threat value can be quickly output, and a level one alarm can be triggered when conditions are met. Once a high risk of data leakage is detected, the system can immediately issue an alarm signal to notify relevant personnel to take measures to effectively prevent information leakage incidents from occurring. Attached Figure Description

[0064] Figure 1 This is a first flowchart of an embodiment of this application;

[0065] Figure 2 This is a second flowchart of an embodiment of this application;

[0066] Figure 3 This is a third flowchart of an embodiment of this application;

[0067] Figure 4 This is the fourth flowchart of an embodiment of this application. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the appendices in the embodiments of the present invention will be described below. Figure 1 -Appendix Figure 4 The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] The first embodiment of this application discloses a screen anti-leakage alarm method based on multi-view dynamic monitoring. (Refer to...) Figure 1 and Figure 2 As one embodiment of the screen leakage prevention alarm method, the screen leakage prevention alarm method includes S110-S200:

[0070] S110 captures video frames in front of the screen, brightness of the screen's reflective area, and ambient background brightness.

[0071] S120, Calculate the device shape matching degree of the suspicious device based on the video frames;

[0072] S130, determine whether the device shape matching degree is greater than the first threshold;

[0073] S140, if so, obtain the center coordinates of the suspicious device and the center coordinates of the screen reflection area;

[0074] S150, calculate the Euclidean distance based on the center coordinates of the suspicious device and the center coordinates of the screen's reflective area;

[0075] S160, based on video frames, tracks the movement speed of a suspicious object across consecutive Z frames;

[0076] S170 calculates the reflectivity coefficient of consecutive Z frames based on the brightness of the screen's reflective area and the ambient background brightness.

[0077] S180, calculate the motion reflection correlation coefficient based on the moving speed and reflectivity coefficient;

[0078] S190, determine whether the Euclidean distance is less than the second threshold and whether the motion reflection correlation coefficient is greater than the third threshold;

[0079] If S200 is selected, the first threat value will be output, and a level 1 alarm will be triggered.

[0080] Specifically, three high-definition cameras can be deployed around the screen bezel or perimeter to capture video from the screen. Light sensors can be used to measure the brightness of reflective areas on the screen and the ambient background brightness. After the cameras capture the current frame, mean-mode filtering is applied to each frame for noise reduction. Then, the Canny edge detection algorithm can be used to extract the object's contour, which is then matched against preset device templates in a template library. The similarity score is the device shape matching score. For example, if the extracted suspicious device contour has 46,000 matching pixels, and the preset mobile phone template has 50,000 pixels, then the device shape matching score is... If the first threshold is set to 0.8, then since 0.92 is greater than 0.8, it indicates that a suspicious device has been detected as a mobile phone. Therefore, it is necessary to further determine whether the suspicious device is strongly correlated with the reflectivity. The judgment conditions are as follows:

[0081] Obtain the center coordinates of the suspicious device and the center coordinates of the reflective area on the screen, and calculate the Euclidean distance based on the center coordinates of both.

[0082] The center coordinates of the suspicious device are obtained by sub-pixelating the edge points of the object's contour using polynomial interpolation (such as cubic splines) to obtain a sub-pixel edge point set. ;

[0083] Then, the centroid is calculated for the sub-pixel edge point set: .

[0084] The center coordinates of the screen's reflective area are obtained as follows: the center of the reflective area refers to the center of the peak brightness region of the screen's reflection in the image. The average image brightness is then calculated. Dynamically set the segmentation threshold:

[0085] Where k=1.2, The Blob detection algorithm is used to extract reflective connected components and filter the area. The noise region of a pixel is used to obtain the set of reflective connected pixels. and its grayscale value set The center of the reflective area is calculated using pixel brightness as the weight:

[0086] .

[0087] pixel coordinates Transform to the physical coordinate system with the image center as the origin (unit: mm):

[0088] The coordinates of the center pixel of the image. f is the physical pixel size, and f is the focal length.

[0089] European distance For example, if the center coordinates of the suspicious device are (300, 400) and the center coordinates of the reflective area on the screen are (320, 380), then the Euclidean distance is... .

[0090] Track the movement speed (pixels / frame) of a suspicious object across Z consecutive frames. Z can be set according to actual conditions or user requirements, for example, tracking the movement speed of a suspicious object across 5 consecutive frames. Reflectivity coefficient , Brightness of the reflective area of ​​the screen. For ambient background brightness, such as the reflectivity of 5 consecutive frames. .

[0091] Then the correlation coefficient of motion reflection ; The standard deviation of the motion trajectory series T is represented by... Indicates reflectivity standard deviation This represents the displacement of the suspicious object in the i-th frame. This represents the average displacement of the motion trajectory over 5 frames. This represents the brightness value of the reflective area of ​​the screen in the i-th frame. This represents the average reflectance brightness over 5 frames. For example, if the second threshold is 50 pixels and the third threshold is 0.7, then due to... Therefore, if the suspicious device is determined to be related to the reflectivity, the first threat value can be directly output. The first threat value is the highest threat value, thereby directly triggering the first alarm.

[0092] Reference Figure 2 and Figure 3 After determining whether the Euclidean distance is less than the second threshold and whether the motion reflection correlation coefficient is greater than the third threshold, the steps also include S310-S370:

[0093] S310, if not, then determine whether the reflectivity coefficient is greater than the fourth threshold;

[0094] If it is S320, then calculate the second threat value according to the distance compensation formula;

[0095] S330, if not, calculate the motion trajectory mutation index based on the video frames;

[0096] S340 outputs a third threat value based on the device shape matching degree, reflectivity coefficient, and motion trajectory mutation index;

[0097] S350, determine whether the second or third threat value is greater than the fifth threshold;

[0098] If S360 is the case, a Level 1 alarm will be triggered.

[0099] S370, if not, trigger a level 2 alarm; the severity of a level 1 alarm is greater than that of a level 2 alarm.

[0100] Specifically, if the fourth threshold is less than the third threshold, and the reflectivity coefficient is greater than the fourth threshold but less than the third threshold, the second threat value can be calculated according to the distance compensation formula. The distance compensation formula is as follows: When the reflectivity index is less than the fourth threshold, the motion trajectory abrupt change index needs to be calculated based on the video frames. Motion trajectory abrupt change index , This represents the movement speed in the i-th frame. This represents the average velocity of the five consecutive frames being tracked.

[0101] ; The experience weight can be modified and reset through the backend; .

[0102] Reference Figure 4 The steps following the triggering of a Level 1 alarm include S410-S460:

[0103] S410 identifies sensitive areas on the screen and generates a mask matrix;

[0104] S420 calculates the primary background color of the sensitive area;

[0105] S430 generates interference colors based on the background primary color;

[0106] S440, determine if the mask matrix is ​​1;

[0107] S450, if so, then according to the interference color, the perturbation is superimposed on the sensitive area according to time t;

[0108] S460, if not, will not respond.

[0109] Specifically, machine learning-based object detection algorithms, such as the YOLO series, can be used to detect objects in screen images and identify predefined sensitive regions, such as tables containing important data or areas with specific labels. When training the model, labeled screen image data is used to label the location and category of sensitive regions. Then, a mask matrix of the same size as the screen image is created based on the detection results. In the mask matrix, the positions corresponding to sensitive regions are assigned a value of 1, and non-sensitive regions are assigned a value of 0. The image of the sensitive regions is converted to the HSV color space, and the color value of each pixel is calculated. The K-means clustering algorithm is used to find the dominant color. K-means clustering divides the color space into different clusters, and the cluster center represents the average color within that cluster. The center of the largest cluster is selected as the dominant background color. If the dominant background color is... Then the interference color For pixels in the sensitive area The perturbation is superimposed over time t (seconds). Pixel perturbation formula:

[0110]

[0111] Where f = disturbance frequency (typically 0.5Hz), period is 2 seconds; This indicates an RGB channel overlay operation; The original pixels. These are the adjusted pixels.

[0112] It should be noted that when performing pixel perturbation, it can be determined whether the threat value remains below the minimum threat threshold for a set duration. If so, a 10-second countdown will begin, and the interference density will gradually decrease. Interference density refers to the degree of modification of the original pixel by the perturbation, such as the number of perturbation operations per unit time. This is the attenuation coefficient, which can be modified and reset in the backend.

[0113] Furthermore, after pixel perturbation, the frequency band of the suspicious device's Wi-Fi or Bluetooth signal can be captured, and a noise signal with the same frequency as the suspicious device's Wi-Fi or Bluetooth signal can be generated. Then, the screen can be controlled to periodically flash LEDs and emit noise signals.

[0114] Specifically, the frequency bands of Wi-Fi or Bluetooth signals from suspicious devices can be captured using radio frequency sensors; noise signals... A represents the signal strength (10mW-100mW, which meets safety standards). The noise is Gaussian white noise (mean 0, variance 1). k is the adjustment coefficient; The frequency band to be captured.

[0115] Furthermore, after emitting the noise signal, floating text can be superimposed on the sensitive area, and the font transparency can be dynamically adjusted before generating and emitting an ultrasonic watermark signal.

[0116] Specifically, the encoded string of the floating text can be "BPSK(Hash("ZX2037-V3.2"))", which is a binary sequence; Then, a single-sideband modulated signal is generated:

[0117] It is transmitted through a speaker (19kHz carrier wave). This is the carrier frequency (typically 19kHz). The frequency shift keying encoding of the decoy text can be the encoded string of the floating text.

[0118] It should also be noted that after the end of the day and before the start of the next day, the number of false alarms in the past 24 hours can be analyzed to determine if the false alarm rate is less than the false alarm rate threshold. If so, the historical threshold remains unchanged; otherwise, it is increased. Relevant thresholds, or manually reset relevant weights or coefficients.

[0119] Additionally, after triggering a level 2 alarm, you can control the display of red triangle icons in the four corners of the screen and control the red triangle icons to flash at a frequency of 2Hz; as well as start the buzzer and control the LED to flash; it should be noted that the buzzer and LED flashing triggered by a level 2 alarm are different from those of a level 1 alarm in terms of buzzer loudness and flashing brightness.

[0120] In addition, during the second-level alarm process, the threat value can be detected every second. When the threat value remains below the minimum threat threshold for a set duration, the alarm is turned off. If the threat value rises and exceeds the fifth threshold or the rate of increase exceeds the rate of increase threshold, a first-level alarm is triggered immediately.

[0121] One implementation scenario of this application is as follows:

[0122] The system collects video frames in front of the screen, brightness of the screen's reflective area, and ambient background brightness. Then, based on the video frames, it calculates the device shape matching degree of the suspicious device and determines whether the device shape matching degree is greater than the first threshold. If so, it obtains the center coordinates of the suspicious device and the center coordinates of the screen's reflective area, and calculates the Euclidean distance based on the two center coordinates.

[0123] Then, based on the video frames, the movement speed of the suspicious object in the consecutive frames is tracked, and the reflectivity coefficient of the consecutive Z frames is calculated based on the brightness of the screen reflective area and the ambient background brightness. Then, based on the movement speed and the reflectivity coefficient, the motion reflectivity correlation coefficient is calculated.

[0124] Determine if the Euclidean distance is less than the second threshold and if the motion reflection correlation coefficient is greater than the third threshold; if so, output the first threat value and trigger a level one alarm.

[0125] After triggering a Level 1 alarm, the system identifies sensitive areas on the screen and generates a mask matrix; it calculates the primary background color of the sensitive area, generates interference colors based on the primary background color, and determines whether the mask matrix is ​​1. If it is, it adds perturbation to the sensitive area according to the interference color over time t.

[0126] Then, the frequency band of the suspicious device's Wi-Fi or Bluetooth signal is captured, and a noise signal with the same frequency as the suspicious device's Wi-Fi or Bluetooth signal is generated. Then, the screen is controlled to periodically flash LEDs and emit noise signals.

[0127] Then, floating text is overlaid on the sensitive area, and the font transparency is dynamically adjusted. Finally, an ultrasonic watermark signal is generated and emitted.

[0128] Based on the above method embodiments, the second embodiment of this application discloses a screen leakage prevention alarm system with multi-view dynamic monitoring. The screen leakage prevention alarm system of this application embodiment can implement any of the above-mentioned multi-view dynamic monitoring screen leakage prevention alarm methods, and the specific working process of each module in the screen leakage prevention alarm system can refer to the corresponding process in the above method embodiments.

[0129] For ease of understanding, an example is as follows: A multi-view dynamic monitoring screen leak prevention alarm system includes:

[0130] The data acquisition module is used to collect video frames in front of the screen, the brightness of the screen's reflective area, and the ambient background brightness;

[0131] The data processing module is used to calculate the device shape matching degree of suspicious devices based on video frames;

[0132] The judgment module is used to determine whether the device shape matching degree is greater than the first threshold. If so, the data acquisition module is used to obtain the center coordinates of the suspicious device and the center coordinates of the reflective area. The data processing module is used to calculate the Euclidean distance based on the two center coordinates.

[0133] The data acquisition module is also used to track the movement speed of suspicious objects in consecutive Z frames based on video frames. The data processing module is used to calculate the reflectivity coefficient of consecutive Z frames based on the brightness of the screen reflective area and the ambient background brightness, and then calculate the motion reflectivity correlation coefficient based on the movement speed and reflectivity coefficient.

[0134] The judgment module is used to determine whether the Euclidean distance is less than the second threshold and whether the motion reflection correlation coefficient is greater than the third threshold.

[0135] The alarm module is used to output the first threat value and trigger a level 1 alarm when the judgment module determines that it is a threat.

[0136] A third embodiment of this application provides a terminal. As one implementation of this terminal, the terminal may include: a memory and a processor; wherein...

[0137] The memory is used to store the screen anti-leakage alarm program for multi-view dynamic monitoring;

[0138] The processor is used to execute the program stored in the memory to implement the steps of the above-described multi-view dynamic monitoring screen anti-leakage alarm method.

[0139] The memory can communicate with the processor via a communication bus, which can be an address bus, a data bus, a control bus, etc.

[0140] Additionally, the memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device.

[0141] Furthermore, the processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0142] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A screen anti-leakage alarm method for multi-view dynamic monitoring, characterized in that, The method comprises the following steps: Collecting video frames in front of the screen, brightness of the screen reflection area, and brightness of the environment background; According to the video frames, calculating the device shape matching degree of the suspicious device; Judging whether the device shape matching degree is greater than a first threshold value; If yes, obtaining the center coordinates of the suspicious device and the center coordinates of the screen reflection area; According to the center coordinates of the suspicious device and the center coordinates of the screen reflection area, calculating the Euclidean distance; According to the video frames, tracking the moving speed of the suspicious object in continuous Z frames; According to the brightness of the screen reflection area and the brightness of the environment background, calculating the reflection intensity coefficient of continuous Z frames; According to the moving speed and the reflection intensity coefficient, calculating the action reflection correlation coefficient; Judging whether the Euclidean distance is less than a second threshold value and whether the action reflection correlation coefficient is greater than a third threshold value; If yes, outputting a first threat value and triggering a first-level alarm; If no, judging whether the reflection intensity coefficient is greater than a fourth threshold value; If the reflection intensity coefficient is greater than the fourth threshold value, calculating a second threat value according to a distance compensation formula; If the reflection intensity coefficient is not greater than the fourth threshold value, calculating an action trajectory mutation index according to the video frames; According to the device shape matching degree, the reflection intensity coefficient, and the action trajectory mutation index, outputting a third threat value; Judging whether the second threat value or the third threat value is greater than a fifth threshold value; If yes, triggering a first-level alarm; If no, triggering a second-level alarm; the severity of the first-level alarm is greater than that of the second-level alarm; The action-reflection correlation coefficient is , The standard deviation of the action trajectory series T is represented by The standard deviation of the reflection intensity is represented by The displacement amount of the i-th frame suspicious object is represented by The average displacement amount of the action trajectory within the Z frames is represented by The brightness value of the i-th frame screen reflection region is represented by The average value of the reflection brightness within the Z frames is represented by 2. The screen anti-leakage alarm method of multi-view dynamic monitoring according to claim 1, characterized in that, The steps after triggering the first-level alarm comprise: Identifying a sensitive area in the screen and generating a mask matrix; Calculating the background main color of the sensitive area; According to the background main color, generating a disturbance color; Judging whether the mask matrix is 1; If yes, according to the disturbance color, superimposing disturbance on the sensitive area at time t.

3. The screen anti-leakage alarm method of multi-view dynamic monitoring according to claim 2, characterized in that, The steps after triggering the first-level alarm can further comprise: Capturing the frequency band of the Wi-Fi signal or the Bluetooth signal of the suspicious device; Generating a noise signal with the same frequency as the frequency band of the Wi-Fi signal or the Bluetooth signal of the suspicious device; Controlling the screen to periodically flash an LED and emit the noise signal.

4. The screen anti-leakage alarm method of multi-view dynamic monitoring according to claim 3, characterized in that, The steps after triggering the first-level alarm can further comprise: Superimposing floating text on the sensitive area and dynamically adjusting the font transparency; Generating and emitting an ultrasonic watermark signal.

5. The screen anti-leakage alarm method of multi-view dynamic monitoring according to claim 2, characterized in that, The screen anti-leakage alarm method further comprises: When in the first-level alarm, judging whether the threat value continuously falls below a threat minimum threshold value within a set time threshold value; If yes, entering a 10-second countdown state and gradually attenuating the disturbance density.

6. The screen anti-leakage alarm method of multi-view dynamic monitoring according to claim 1, characterized in that, The steps after triggering the second-level alarm comprise: Controlling the screen to display a red triangular icon at the four corners and controlling the red triangular icon to flash at a frequency of 2 Hz.

7. A multi-view dynamic monitoring screen anti-leakage alarm system, characterized in that, The screen anti-leakage alarm method for multi-view dynamic monitoring according to any one of claims 1-6 comprises: A data collection module for collecting video frames in front of the screen, brightness of the screen reflection area, and brightness of the environment background; A data processing module for calculating the device shape matching degree of the suspicious device according to the video frames; A judging module is configured to judge whether the device shape matching degree is greater than a first threshold value, and if yes, the data acquisition module is configured to acquire the center coordinates of the suspicious device and the center coordinates of the reflective area, and the data processing module is configured to calculate the Euclidean distance according to the two center coordinates. The data acquisition module is further configured to track the moving speed of the suspicious object in continuous Z frames according to the video frames, and the data processing module is configured to calculate the reflective intensity coefficient of the continuous Z frames according to the screen reflective area brightness and the environment background brightness, and then calculate the action reflective correlation coefficient according to the moving speed and the reflective intensity coefficient. The judging module is configured to judge whether the Euclidean distance is less than a second threshold value and whether the action reflective correlation coefficient is greater than a third threshold value, and if not, further judge whether the reflective intensity coefficient is greater than a fourth threshold value. An alarm module is configured to output a first threat value and trigger a first-level alarm when the judging module judges yes. The data processing module is configured to calculate a second threat value according to a distance compensation formula when the reflective intensity coefficient is greater than the fourth threshold value, calculate an action trajectory mutation index according to the video frames when the reflective intensity coefficient is not greater than the fourth threshold value, and output a third threat value according to the device shape matching degree, the reflective intensity coefficient and the action trajectory mutation index. The judging module is further configured to judge whether the second threat value or the third threat value is greater than a fifth threshold value, and if yes, the alarm module triggers a first-level alarm, and if not, the alarm module triggers a second-level alarm, and the severity of the first-level alarm is greater than that of the second-level alarm. The action-reflection correlation coefficient is , The standard deviation of the action trajectory series T is represented by The standard deviation of the reflection intensity is represented by The displacement amount of the i-th frame suspicious object is represented by The average displacement amount of the action trajectory within the Z frames is represented by The brightness value of the i-th frame screen reflection region is represented by The average value of the reflection brightness within the Z frames is represented by 8. A terminal, characterized by comprising: The application comprises: a memory storing a multi-view dynamic monitoring screen anti-leakage alarm program; a processor configured to execute the program stored on the memory to realize the steps of the multi-view dynamic monitoring screen anti-leakage alarm method according to any one of claims 1-6.

Images