EA044006B1 - METHOD OF ADDING DIGITAL MARKS INTO A DIGITAL IMAGE AND DEVICE FOR IMPLEMENTING THE METHOD - Google Patents

Description

Field of technology

This technical solution generally relates to the field of data encoding and decoding, and in particular to a method and device for adding digital marks to a digital image.

State of the art

The problem of protecting confidential information (hereinafter - CI) from its unauthorized disclosure and dissemination is not only important and relevant in commercial and government organizations due to the value of information as a digital asset, but also a direct requirement of the current legislation of the Russian Federation on the protection of personal data, banking, commercial and other types of secrets.

Unauthorized/wrongful disclosure of confidential information can lead to both reputational damage for organizations (lawsuits, claims, negative experiences and customer outflow, severance of relationships with contractors and partners) and direct financial losses (fines from regulators, compensation to clients and counterparties, loss of market share, loss of profit as a result of suspension/termination of activities due to revocation of licenses, etc.). Despite the development of modern digital technologies and the improvement of methods/means of protecting information from unauthorized access and disclosure, the existence of a specific information leakage channel has been and remains the most risky, effective and difficult to prevent channel for compromising confidential information. At the same time, it is extremely labor-intensive and resource-intensive to create reliable procedures for monitoring leaks through the view channel, which would make it possible to unambiguously identify an attacker at an arbitrary point in time who performed unauthorized actions while taking photos and videos of the monitor screen during CI processing.

In modern business conditions, the importance of information is constantly growing as one of the most valuable assets of an organization, which dictates the need to protect both IT systems and the confidential data themselves processed in such systems, not only from external but also from internal attackers [1]. In particular, insiders can compromise sensitive information simply by opening an electronic document, presentation or spreadsheet file, database, website, or application and taking screenshots of the computer screen using a digital camera or cell phone. This method of copying confidential information will allow an attacker to avoid recording his actions by standard or additional information security tools used in CI processing systems, unlike, for example, sending a document by email. Carrying out such actions greatly complicates the detection and identification of an attacker. Even the presence of a security policy adopted by the organization that prohibits photo and video copying of confidential information from the screen of a personal computer (hereinafter referred to as PC) cannot completely eliminate the threat of illegal copying and subsequent unauthorized disclosure of CI. In order to minimize the described risk, a new method is proposed for the formation and implementation of digital watermarking, invisible to the human eye, embedded in the image displayed on the computer screen. Information encoded in digital tags contains information about the date, time, computer name and login of the user who processed the information, and can be extracted from copies of images obtained as a result of illegal photographing of such information from a computer screen.

The evolution of watermarking technologies has developed from embedding visible marks into electronic documents prepared both directly on a PC and printed on paper copies of documents using means built into multifunctional printing devices, to embedding affine transformations invisible to the human eye into arbitrary digital content converted from basic, commonly used ones. electronic document formats (Microsoft Office, Adobe PDF, etc.) into one of the graphic formats (JPEG, PNG, TIFF, etc.). Earlier approaches to latent image labeling, mainly [2, 11, 13], used methods of placing watermark information by embedding it in the low-order bits responsible for the color of individual pixels in the image. The resulting small changes in the color of the image are invisible to the human eye, but the described approaches to covert water marking do not involve cases of image conversion, for example, from color to black and white, which would be quite acceptable for an attacker stealing CI contained in the text of a document displayed on a PC screen . In more promising approaches [9], when electronic documents are directly subject to protection, they are converted from commonly used formats into graphic ones and, at the same time, invisible distortions are introduced directly into the image (affine transformations), a photocopy of such a protected document is converted into another format, or a color image is converted to black and white will be meaningless, since a photocopy of a graphic document protected in this way will inherit all the implemented affine transformations. However, the conversion from one format to another requires not only the time and resource expenditure of PC power, the storage of initially labeled electronic documents in conjunction with artifacts of the affine transformations made in connection with a specific recipient of a unique copy, the existence of a content indexing mechanism to search for samples of compromised content , but also labor-intensive tools

- 1 044006 conducting investigations into the compromise of documents containing confidential information primarily in a manual or semi-automated mode, which makes the use of this type of technology limited in large enterprises, with the number of employees numbering in the thousands, and where electronic document management involves the exchange of a significant number of confidential documents with a large number of the number of recipients of unique copies of electronic documents protected using invisible affine tags embedded in the content. At the same time, the processes of preparing secure copies of electronic documents, storing artifacts of generated transformations, conducting investigations into leaks that have become known to the security services of organizations, without the relative costs of identifying the leaks themselves, can be extremely labor and time-consuming. The technologies used for introducing invisible affine transformations also have restrictions on the number of unique, distinguishable copies of the original confidential document prepared for each recipient.

Works such as Caronni [5], which embed a digital label by changing the brightness of several adjacent pixels, require the source image to extract data from the label, while the proposed solution does not require this. The methods used in the modern multimedia industry to protect digital content by introducing digital marks into the frequency spectrum of the image [7, 18, 19] provide acceptable secrecy for color images, but usually manifest themselves in a revealing way in text documents [12, 14]. Thus, the above methods for introducing digital tags are not suitable for covertly marking heterogeneous content. The syntactic and semantic approaches used to change the text itself in electronic documents, by, for example, dividing the text into blocks of words or letters and then moving or replacing them, are also unsuitable for the purposes of CI protection since it is impossible to cover the entire variety of text processing programs for different operating systems , require large computational costs, which casts doubt on their suitability for processing information in real time, and in any case, do not cover cases of hidden marking of CI processed in non-text editors (for example, in web forms of sites or in databases).

The proposed solution makes it possible to make modifications to the brightness of the video image played on the PC screen unnoticed by the user. The change in the brightness of the video image is carried out taking into account a predefined pixel scheme - a digital mark symbol, the change in the brightness of some pixels of which in relation to the brightness of the base video image depends on what data is to be encoded (hereinafter we will call the described process digital marking). The set of data used to uniquely identify the electronic account of the user who processed the CI on a PC, with an accuracy of date and time, will be further called a digital tag. The invention is intended to protect any confidential information processed by a PC user, regardless of the form of its presentation (text, graphics, video) and applications in which processing is carried out, from unauthorized copying from the PC screen for the purpose of subsequent unlawful distribution or disclosure of such information, compromise by others by any means. Digital marking of a video image is carried out secretly, using specialized software (hereinafter referred to as software), visually invisible to the eyes of the PC user. The secrecy of encoded digital marks embedded in a video image displayed on a PC screen is achieved due to the fact that the human eye, especially in areas of light colors [3, 6], is less sensitive to small changes in brightness than digital cameras or mobile phones.

The reverse process associated with the analysis of the pixel brightness differentials of a copy of a video image received from a PC screen, as a result of which the size of the digital mark symbol is determined and the sequential (character-by-character) determination of the data embedded in the video image, will be called data extraction.

The fundamental technological result of the invention is to increase the protection of confidential information from unauthorized copying from the PC monitor screen and their unlawful distribution by making it possible to establish the exact date, name of the PC and identify the user who leaked protected information by extracting such information from digital marks hidden in the image. This application discloses a method for modifying a video image, introducing digital marks invisible to the human eye, as well as methods for processing digital images made by photographing and/or copying a screen (screen shot) containing protected information, extracting embedded data, restoring digital marks in order to determine the source and identification of the culprit of leakage of confidential information.

The proposed invention does not require the creation of a separate database for storing information about the unique characteristics (artifacts) embedded in the video image of digital tags, since it is assumed that during software implementation all the necessary data about the name of the PC, the login of the user who successfully logged into the PC, as well as date and time will be directly received on the PC and embedded as invisible digital marks in the video image displayed on the PC screen. In cases of investigations into data leakage, a security officer will only need to obtain a copy (screenshot) or photo of the PC screen taken by the attacker in order to, after processing the image, programmatically restore the above data embedded in the form of hidden marks and thereby uniquely identify the user, and the PC on which the CI was compromised at a specific point in time.

The essence of the invention

The technical problem or task posed in this technical solution is to create a simple, invisible to users and reliable method and device for adding digital marks to a digital image containing information about the date, time and circumstances of CI processing (for example, the name of the device and the login of the person authorized to username).

The technical result achieved by solving the above technical problem is to ensure the possibility of automated generation and implementation of digital marks in a digital image, grouped from pixels of variable brightness into a sequence of symbols containing encoded information, for example, necessary for conducting an investigation and establishing the circumstances in the event of a CI leak by unauthorized copying, incl. through the viewing channel (for example, by photographing the screen).

This technical result is achieved by implementing a method for adding digital marks to a digital image, performed by at least one computing device, comprising the stages of obtaining a tuple of data to be added in the form of digital marks to a digital image displayed on the screen of the user's device;

determine the shape (in the form of a pixel-by-pixel representation) and the maximum permissible dimension of the digital mark symbol based on the device screen resolution used by the user;

based on information about the shape and dimension of the digital mark symbol, the shape and dimension of the internal matrix of pixels of the base brightness of the digital mark symbol are determined;

based on the data contained in the tuple, the idea of the shape and maximum allowable dimension of the digital mark symbol and the shape and dimension of the internal matrix of pixels of the basic brightness of the digital mark symbol, a sequence of digital mark symbols is formed;

based on the device screen resolution used by the user, a template matrix is formed that determines the line-by-line arrangement of the sequence of generated digital label symbols for display on the screen;

superimpose the generated template-matrix onto a digital image intended for display on the screen;

displaying a digital image with the above-mentioned template-matrix superimposed, containing a sequence of generated digital mark symbols, onto the screen of the user's device.

In one of the particular examples of implementation of the method, noise-resistant coding of the received data tuple is additionally performed. In another particular example of the method, the steps are additionally performed in which: a checksum is calculated for the data tuple to check the integrity of the data; add the resulting checksum to the data tuple.

In another particular example of the method, the current screen parameters are additionally determined, and the maximum allowable dimension of the digital mark symbol is determined taking into account the current screen parameters and the size of the data tuple.

In another particular example of the method, data about the user's device is additionally collected, including: the date set on the user's device; time set on the user's device; user device ID; username (LOGIN) who has authorized user device 10; wherein the data tuple is formed from data about the user's device.

In another preferred embodiment of the claimed solution, a device for adding digital marks to a digital image is provided, comprising at least one computing device and at least one memory device containing machine-readable instructions, which, when executed by at least one computing device, perform the above method.

Brief description of drawings

The features and advantages of the present invention will become apparent from the following detailed description of the invention and the accompanying drawings.

In fig. 1 presents possible scenarios for compromising confidential information.

In fig. Figure 2 shows an example of a data display system.

In fig. Figure 3 shows examples of digital mark symbols.

In fig. Figure 4 shows examples of the original screen image with and without overlaid digital mark symbols (the brightness of the symbols is increased).

In fig. Figure 5 shows an example of a general view of a computing device.

- 3 044006

Detailed Description of the Invention

Below we will describe the concepts and terms necessary to understand this technical solution.

In this technical solution, a system means, including a computer system, a computer (electronic computer), CNC (computer numerical control), PLC (programmable logic controller), computerized control systems and any other devices capable of performing a given, clearly defined sequence of operations (actions, instructions).

A command processing device is an electronic unit or an integrated circuit (microprocessor) that executes machine instructions (programs).

A command processing device reads and executes machine instructions (programs) from one or more storage devices. Storage devices can include, but are not limited to, hard drives (HDD), flash memory, ROM (read-only memory), solid-state drives (SSD), and optical drives.

A program is a sequence of instructions intended to be executed by a computer control device or a command processing device.

A data tuple is an ordered set of data of a fixed length. In the vast majority of cases, the administrative, organizational and technical measures to protect private information implemented in modern organizations, as a rule, cannot prevent insider attackers [1] from copying confidential information by taking photographs from the PC screen. Such unauthorized actions do not leave any digital traces in the PC logs, which makes it impossible to identify the attacker, as well as to establish the very fact of photographing the CI.

The approach proposed in the invention covers the following two scenarios, depicted in FIG. 1.

Scenario 1: Malefactor Insider 1 takes a photo of sensitive information displayed on PC 2's screen, for example, using a mobile phone or camera, then forwards and publishes the file on the Internet 3. Scenario 2: Malicious Insider 1 makes a photocopy, e.g. , from the screen of PC 2 (for example, through the standard print screen function), then sends and publishes the file on the Internet 3. In both cases, an attacker can compromise the CI by transmitting the file both in its original and converted form, using special photo effects ( blur, noise, sharpen, etc.) [4] or image compression through conversion to an image of another format (for example, from BMP to PNG or from RAW to JPEG), and thus the digital copy of the screen image will acquire artifacts (halo or spots characteristic of the JPEG format). The situation with scenario 1 can be complicated by the fact that photographing the screen can be done in poor lighting, at an angle (not perpendicular to the PC screen 2), with a predetermined poor quality of shooting. In addition, a photocopy can be noisy due to the peculiarities of the hardware and software implementation of the camera itself and result in both color noise and luminosity noise (grain). The easiest option for investigating CI leaks involves obtaining access to a file with a copy of the screen according to scenario 2, i.e. to the original unaltered image or, in some cases, to files stored on a mobile device, for example, seized from an attacker during an arrest by law enforcement agencies. But such cases cannot be considered frequent or typical. In view of the above, it is necessary to solve the problem of identifying attacker 1 who compromised the CI, location (for example, the name of PC 2 on which processing was carried out) and time only based on the analysis of the image that was obtained directly from the source of illegal placement. To solve the problem described above, an approach is proposed for introducing hidden (invisible to the human eye) digital marks (signs) into the image displayed on the PC screen. A digital tag may contain, for example, such encoded information as: date, time, PC name (name of the workstation from the screen of which a copy of the CI was made), as well as the name of the user (LOGIN) who authorized authorization on this PC. If they gain access to a (altered) photocopy of the screen, computer forensics or organizational security personnel will be able to decode the above information from the digital tags and identify the attacker and the workplace where the CI was processed with date and time accuracy. The proposed approach is permanent in its implementation, since an attacker is able to take a screenshot at any time, not all PCs can have or operate CI protection measures, there are no organizational or administrative control procedures, therefore a digital tag must be present when processing any CI in any form of representation regardless of the applications used. A digital mark embedded in an image displayed on a monitor screen must be invisible to the user’s eye both on text documents and not add visible artifacts to graphics that could become unmasking elements of the image itself. The described approach involves the introduction of digital tags that are robust to distortion of a copy of the image, allowing, at the same time, the blind extraction of information encoded in them, that is, the extraction of data about the date, time, PC name and user without the original

- 4 044006 images. In accordance with the diagram shown in Fig. 2, Data display system includes:

a user device 10 and a device 100 for adding digital marks to a digital image. The user device 10 may be a computing device equipped with digital image output capabilities, and may be, for example, a laptop or desktop computer, tablet, mobile phone, smartphone, or other device.

The device 100 for adding digital marks to a digital image can be implemented on the basis of the software and hardware of the user device 10, or represent an independent device connected to the said device 10 via wired and/or wireless communication and containing at least one computing device. The digital marking device 100 may include: a data acquisition module 101, a data processing module 102, a data encoding module 103, and a data output module 104. These modules may be implemented in firmware of device 100, configured to perform the functions assigned to these modules herein. According to the software algorithm incorporated by the developer into the data collection module 101, the module 101 accesses the user device 10 to collect information about the current screen parameters, for example, screen resolution, and data about the user device 10, which may include the date set on the user device 10;

the time set on the user device 10;

an identifier of the user's device 10, for example, the name of said device;

username (LOGIN) who has authorized user device 10.

The collected data about the user device 10 is combined by the data collection module 101 into a tuple of data to be entered in the form of digital marks in a digital image displayed on the screen of the user device, for example, like:

||rrMCflfl44MHFQDN|UserLogin, where:

- | - a special character indicating the beginning of the data tuple and acting as a separator for the FQDN and UserLogin fields, 0x7C (character code | in hexadecimal). It is assumed that the specified special character cannot be used in the FQDN and UserLogin fields;

- yymsdddhchmn - 10 bytes of fixed length data about the current year, month, day, hour and minutes, received from the operating system of the user device 10 on which CI is processed, where in turn:

yy - the last 2 digits of the current year, for example, for 2019 the value 19 will be written;

ms - month;

dd-day;

hh - hour;

mn - minutes;

FQDN - Fully Qualified Domain Name or user's full device name 10 [8], has a limit of 255 bytes, for example: WS-CA1 11.company.com;

UserLogin is the name of the device 10 user who successfully authenticated on the workstation; this field has a limit of 20 bytes if the sAMAccountName attribute is used in the Microsoft Active Directory domain [10]. If the userPrincipalName attribute is used and 10 64 bytes are allocated in the domain for the device user name, it is proposed to trim the user name to 20 bytes, since the main identifying feature of CI processing will be the FQDN.

Thus, the total length of the data tuple to be encoded (payload), in the above example, will not exceed 288 bytes in the limit. Next, information about the current screen parameters and the data tuple is transmitted by the data collection module 101 to the data processing module 102, which, based on the generated data tuple, calculates a checksum to check data integrity - a cyclic redundancy check (CRC), for example, using an algorithm CRC32 [17]. The received checksum is added by the mentioned module 102 to the data tuple, after which the resulting 292-byte data tuple is transferred to the data encoding module 103, which performs noise-resistant encoding of the received data tuple using, for example, an adapted convolutional encoder and transmits the encoded data tuple to the output module 104 data. The parameters necessary for the operation of the module 103, in particular, the noise-correcting coding algorithm used and the corresponding settings, can be preset by the developer of the digital marking device 100 based on the ratio of the encoder's power to correct errors that may arise in the process of extracting data from a copy of the image, to the complexity and labor-intensiveness of the software implementation of the selected noise-resistant codec (encoder-decoder), as well as the availability of the corresponding algorithms for implementation, taking into account compliance with their patent protection.

Additionally, the data processing module 102 is configured to determine the shape (in the form of a pixel-by-pixel representation) and the maximum allowable dimension of a digital mark symbol (examples of which are shown in Fig. 3) - i.e. calculating the number of pixels vertically and horizontally,

- 5 044006 forming a symbol, for example, based on information about the current screen parameters and the size of the data tuple received from the module 101. For this purpose, the mentioned module 102, based on information about the current screen parameters, in particular based on the screen resolution of the device 10 used by the user, determines the total number of pixels that the screen will accommodate, then determines the number of bits of data based on the size of the data tuple to be encoded as digital label characters. Next, module 102 determines the shape and maximum possible dimension of the digital mark symbol, using which all data to be encoded (in the form of a sequence of bits of a data tuple) can be completely placed on the screen in the form of sequentially displayed digital mark symbols with line-by-line transfer upon reaching the limit of the horizontal screen resolution , based on the relationship: one bit of data is encoded by one symbol of a digital mark.

Examples of digital mark symbols are shown in FIG. 3, in particular for dimensions 3x3 and 6x6 pixels. In gray color in Fig. Figure 3 shows the pixels of the digital mark symbol with the basic brightness of the original image, black - pixels with changed (distinctive) brightness. Information about the screen resolutions of the user device 10, the corresponding shapes of digital marks, and the shapes of the internal pixel matrix of the base brightness of the digital mark symbol can be predetermined by the designer of the digital marking device 100.

According to FIG. 3, based on the shape and the calculated maximum possible dimension of the digital mark symbol, said module 102 determines the shape and dimension of the internal base luminance pixel matrix of the digital mark symbol, which characterizes the number of base luminance pixels that will be contained in the digital mark symbol. Information about the digital mark symbol dimension and the dimension of the internal base luminance pixel matrix corresponding to the digital mark symbol dimension can be preset in the data processing unit 102 by the designer of the digital mark insertion device 100 . So, for example, for a resolution of 2560x1440 pixels, the screen will accommodate 3686400 pixels. For a data tuple of length, for example, 289 bytes = 2312 bits (without noise-resistant coding, but with CRC32), each bit encoded character-by-character into the image displayed on the screen of the user device 10 can be displayed by a matrix with a maximum dimension of: V3686400 pixels g 2312 bit « 39 pixels. Based on the obtained maximum permissible dimension of the digital mark symbol in module 102, a decision is made on the dimension of the internal matrix of base brightness pixels, the brightness of which will differ from the perimeter pixels of the same symbol, based on the following calculations and restrictions. As defined in FIG. 3 square shape of the digital mark symbol (pixel matrix), the size of the internal matrix of base brightness pixels must be three times less than the maximum allowable dimension of the symbol. So, for the example above, based on the calculated maximum character size of 39 pixels, the size of the internal matrix of base brightness pixels will be equal to 39/3=13 pixels. In cases where the calculated maximum character dimension is not a multiple of three, when calculating the size of the internal matrix of base brightness pixels, it is necessary to round down to the nearest integer value.

In order to introduce information redundancy into the image, for subsequent recovery of data from digital marks, the proposed approach involves an adaptive reduction of the pixel dimension of the symbol from the maximum possible downwards, which increases the secrecy of digital marking, potentially allows you to recover data from a fragment of a copy of the screen image, but reduces the robustness digital marks to transformational transformations of the image. Accordingly, information about the shape and calculated dimension of the digital mark symbol and the shape and dimension of the internal matrix of base brightness pixels by the data processing unit 102 is transmitted to the data output unit 104.

Data output module 104, having received from module 103 a tuple of encoded data protected by an adaptive noise-resistant encoder, and digital mark symbol parameters characterizing the shape and dimension of the digital mark symbol and the shape and dimension of the internal matrix of base brightness pixels, generates a sequence of digital mark symbols to be embedded (luminance overlay) into a video image displayed on the screen of the user device 10 in accordance with the following principles.

When it is necessary to encode the zero bit of data, the pixels shown in FIG. 3 in black are made less intense in brightness than the brightness of the pixels in the base image. To encode a unit bit, the pixels shown in FIG. 3 in black are made more intense in brightness than the brightness of the pixels in the base image. In the work of M. Ramasubramanian [15], a model of human image perception was described, taking into account the different sensitivity of the eye to changes in brightness depending on the brightness of the background and homogeneity. The calculation of the amount by which the brightness of the perimeter pixels of a digital mark symbol must differ from the background so that a person does not notice the difference can be calculated, for example, based on the proposed Ramasubramanian model. After a sequence of digital mark symbols has been generated in module 104, said module 104, based on information about the current screen parameters of the user device 10, in particular

- 6 044006 based on the device screen resolution used by the user, it generates a template-matrix that determines the row-by-line arrangement of the sequence of prepared digital mark symbols for display on the screen of the user's device 10. Thus, the template matrix generated in module 104 consists of pixels of variable brightness with a dimension corresponding to the current parameters of the screen of the user’s device 10. Next, the data output module 104, using the mentioned template-matrix containing a sequence of generated symbols of digital marks, superimposes it with the original digital image, after which the final (with the changed brightness of certain pixels according to the mentioned template-matrix) generated digital image is sent to the user device 10, which performs, using methods known from the prior art, the display of an image on the screen of the device 10. Thus, the achievement of the specified technical result is ensured, which consists in providing the possibility of automated generation and implementation of digital labels in a digital image.

In fig. Figure 4 shows examples of enlarged copies of the original image of a screen fragment with a superimposed digital mark and without a digital mark (the brightness of the characters is increased). Accordingly, in FIG. 4 presented:

A. a fragment of a copy of an image with superimposed symbols of digital marks of increased brightness;

B. fragment of a copy of the original image without overlaying digital mark symbols.

Due to the preliminary formation of a tuple of encoded data with a frequency of once per minute (time is encoded with an accuracy of a minute), resource savings of the digital marking device 100 are achieved. In this case, the module 101 for collecting data from the operating system of the user's device 10 must monitor a number of system events associated with the user's login/exit from the system. If module 101 is configured to monitor screen lock/unlock events, user device 10 going into sleep mode or hibernation mode, it will also be possible to save resources of user device 10, since applying the above-mentioned matrix template (overlay mask) will be meaningless due to blocking PC screen and the inability for an attacker to make a photocopy of the image. Thus, only the process of modifying the video image displayed on the screen in accordance with the preformed brightness pixel-by-pixel template will be critical from the point of view of implementation and resource-intensive.

Data extraction from digital tags is carried out as follows. During the reverse conversion of digital mark symbols into a binary data sequence, a pixel-by-pixel analysis of the brightness differentials of the image of the user's device 10 screen, for example, a PC screen, obtained during the investigation takes place. In this case, the dimension of the digital mark symbol is determined, for example, by analyzing the brightness of pixels in the margins of a copy of an electronic document that does not contain text. This approach is based on the simple fact that white and background luminance are clearly dominant in typical text documents, allowing the background to be used to determine local luminance reference values for subsequent decoding of digital mark symbols. By setting the size of the digital mark symbol, the brightness of the central pixels of the symbol is compared with their surroundings. If the brightness of the perimeter pixels is less intense than the brightness of the central pixels of the digital mark symbol, a zero bit is registered, otherwise a unit bit is recorded. The binary data sequence reconstructed from the image is subjected to noise-resistant decoding, for example, using the Viterbi algorithm [16], which makes it possible to find and correct errors in the reconstructed data array that can arise both during the conversion of digital mark symbols into a binary sequence, and as a result of quality deterioration copies of a PC screen image, application of special photo effects or conversion to an image of a different format. The power of noise-resistant decoding for correcting single and block errors directly depends on the efficiency of the algorithm used and the redundancy introduced into the original binary data array protected from distortion. As a result of noise-resistant decoding, a data tuple of up to 292 bytes should be obtained, of which the last 4 bytes are a checksum calculated using the CRC32 algorithm. If the calculated checksum for the initial 288 bytes coincides with the last 4 check bytes, a decision is made to completely and accurately restore data about the date, time, user and name of the PC on which the CI was processed.

Thus, the technical problem to which the claimed invention is aimed is to provide additional protection of any confidential information processed on a PC and displayed on the PC screen, regardless of the form of presentation and the software used, from unauthorized disclosure and distribution by embedding in the image digital tags invisible to the human eye that allow one to uniquely identify both the user and the PC on which the CI was compromised at a specific point in time. The solution to the stated technical problem is achieved by the fact that a method for modifying the brightness of a video image transmitted to the PC screen, robust to transformations of photocopies of a PC screen, is implemented in software, without visible distortion for the user, whose embedded digital tags contain encoded using effective noise-resistant

- 7 044006 algorithms information about the date, time, name of the PC and login of the user who logged into the operating system of the PC on which the CI was processed. Protecting confidential information processed on a PC using the approach described above will make it possible to implement a control point for illegal photocopying of information from a PC screen, ensuring the ability to establish the location (name of the PC) and the person who processed the CI, accurate to date and time, during the investigation incidents related to CI leakage. The technical effect of the proposed invention is the ability to obtain information about the facts and channel of leakage of confidential information processed on a PC, directly by extracting data from a copy of a compromised image obtained by photographing the PC screen. In general (see Fig. 5), the process data processing device 10 contains one or more processors (201), memory devices such as RAM (202) and ROM (203), and input/output interfaces (204) united by a common information exchange bus. ), input/output devices (205), and a network communication device (206).

The processor (201) (or multiple processors, multi-core processor, etc.) may be selected from a variety of devices commonly used today, for example, from manufacturers such as: Intel™, AMD™, Apple™, Samsung Exynos™, MediaTEK ™, Qualcomm Snapdragon™, etc. The processor or one of the processors used in the device (200) must also include a graphics processor, such as an NVIDIA or Graphcore GPU, the type of which is also suitable for carrying out the method in whole or in part, and can also be used for training and applying machine learning models in various information systems. RAM (202) is a random access memory and is designed to store machine-readable instructions executed by the processor (201) to perform the necessary operations for logical data processing. The RAM (202) typically contains executable operating system instructions and associated software components (applications, program modules, etc.). In this case, the available memory capacity of the graphics card or graphics processor can act as RAM (202).

ROM (203) is one or more permanent storage devices, such as a hard disk drive (HDD), a solid state drive (SSD), flash memory (EEPROM, NAND, etc.), optical storage media (CD-R) /RW, DVD-R/RW, BlueRay Disc, MD), etc. To organize the operation of device components (200) and organize the operation of external connected devices, various types of I/O interfaces (204) are used. The choice of appropriate interfaces depends on the specific design of the computing device, which can be, but is not limited to: PCI, AGP, PS/2, IrDa, FireWire, LPT, COM, SATA, IDE, Lightning, USB (2.0, 3.0, 3.1, micro, mini, type C), TRS/Audio jack (2.5, 3.5, 6.35), HDMI, DVI, VGA, Display Port, RJ45, RS232, etc. To ensure user interaction with the device (200), various means (205) of I/O information are used, for example, a keyboard, a display (monitor), a touch display, a touch pad, a joystick, a mouse, a light pen, a stylus, a touch pad, a trackball. , speakers, microphone, augmented reality tools, optical sensors, tablet, light indicators, projector, camera, biometric identification tools (retina scanner, fingerprint scanner, voice recognition module), etc.

The network communication means (206) provides data transmission via an internal or external computer network, for example, Intranet, Internet, LAN, etc. One or more means (206) may be used, but not limited to: Ethernet card, GSM modem, GPRS modem, LTE modem, 5G modem, satellite communication module, NFC module, Bluetooth and/or BLE module, Wi-Fi module and etc.

Additionally, satellite navigation tools can also be used as part of the system (200), for example, GPS, GLONASS, BeiDou, Galileo. The specific selection of device elements (200) for implementing various software and hardware architectural solutions can vary while maintaining the required functionality provided.

Modifications and improvements to the above-described embodiments of the present technical solution will be apparent to those skilled in the art. The foregoing description is provided by way of example only and is not intended to be limiting. Thus, the scope of the present technical solution is limited only by the scope of the attached claims.

-

Claims (18)

1. Ссылки на литературу.

   [1] https://encvclopedia.kasperskv.com/knowledqe/recoqnizinq-different-typesof-insiders/
2. [2] W. Bender, D. Gruhl, N. Morimoto, and A. Lu. Techniques for data hiding. IBM Syst. J., 35(3-4):313-336, Sept. 1996.
3. [3] Tom N. Cornsweet. Visual Perception. Academic Press, 1970.
4. [4] https://helpx.adobe.com/ru/photoshop/usinq/filter-effects-reference.html
5. [5] G. Caronni. Assuring ownership rights for digital images. In Verlaessliche ITSysteme, DUD-Fachbeitraege, pages 251-263. Vieweg+Teubner Verlag, 1995.
6. [6] I. Cox, M. Miller, J. Bloom, J. Fridrich, and T. Kalker. Digital Watermarking and Steganography. Morgan Kaufmann, 2007.
7. [7] I. J. Cox, J. Kilian, F. Leighton, and T. Shamoon. Secure spread spectrum watermarking for multimedia. IEEE Trans, on Image Processing, 6(12):16731687, Dec 1997.
8. [8] RFC #1035 DOMAIN NAMES - IMPLEMENTATION AND SPECIFICATION, November 1987, https://tools.ietf.orq/html/rfc1035
9. [9] http://www1.fips.ru/reqisters-doc-view/fips servlet?DB=RUPAT&DocNumber =2646341 &TypeFile=htm I
10. [10] https://docs.microsoft.com/en-us/dotnet/api/system.web.security.activedirec torvmembershipprovider.createuser?redirectedfrom=MSDN&view=netframewor k-4.8#System Web Security ActiveDirectoryMembershipProvider CreateUser System String System String System String System String Sv stem String System Boolean System Object System Web Security Membe rshipCreateStatus

    [11 ] R. Van Schyndel, A. Tirkel, and C. Osborne. A digital watermark. In IEEE Int. Conf, on Image Processing (ICIP), volume 2, pages 86-90 vol.2, Nov 1994.
11. [12] Y. Liu, J. Mant, E. Wong, and S. H. Low. Marking and detection of text documents using transformdomain techniques. Proceedings of SPIE - Volume 3657, Electronic Imaging Conference on Security and Watermarking of Multimedia Contents, pages 317-328, 1999.
12. [13] N. Nikolaidis and I. Pitas. Robust image watermarking in the spatial domain. Signal Processing, 66(3):385-403, 1998.
13. [14] A. M. Alattar and О. M. Alattar. Watermarking electronic text documents containing justified paragraphs and irregular line spacing. Proceedings of SPIE Volume 5306, Security, Steganography, and Watermarking of Multimedia Contents VI, pages 685-695, Jan 2004.
14. [15] Mahesh Ramasubramanian, Sumanta Pattanaik, Donald Greenberg A Perceptually Based Physical Error Metric for Realistic Image Sythesis, SIGGRAPH-99 Proceedings
15. [16] A. Viterbi. Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. IEEE Trans, on Information Theory, 13(2):260-269, 1967
16. [17] Lin, Shu & D. Costello, Error Control Coding, Prentice-Hall, 1983
17. [18] C.-S. Shieh, H.-C. Huang, F.-H. Wang, and J.-S. Pan. Genetic watermarking based on transform-domain techniques. Pattern Recognition, 37(3):555 - 565, 2004.
18. [19] T. K. Tsui, X.-P. Zhang, and D. Androutsos. Color image watermarking using multidimensional fourier transforms. IEEE Trans, on Information Forensics and Security, 3(1):16-28, March 2008.

    ФОРМУЛА ИЗОБРЕТЕНИЯ

    1. Способ автоматизированного внесения цифровых водяных знаков (ЦВЗ) в цифровое изображение, предназначенное для вывода на экран, выполняемый по меньшей мере одним вычислительным устройством, содержащий этапы, на которых получают кортеж данных, подлежащих внесению в виде ЦВЗ в цифровое изображение, предназначенное для вывода на экран устройства пользователя;

    определяют форму в виде попиксельного представления и максимально допустимую размерность символа ЦВЗ исходя из используемого пользователем разрешения экрана устройства;

    на основе информации о форме и размерности символа ЦВЗ определяют форму и размерность внутренней матрицы пикселей базовой яркости символа ЦВЗ, яркость которых будет отличаться от яркости периметровых пикселей того же символа;

    на основе данных, содержащихся в кортеже, представлении о форме и максимально допустимой

    -