Mobile equipment uniqueness authentication method

Abstract

The invention is applicable to the technical field of computers and information security, and provides a mobile equipment uniqueness authentication method, which comprises the following steps that: an application program controls a screen of mobile equipment to execute an illumination sequence; in the process of executing the illumination sequence by the screen, acquiring image data by using a front image acquisition device, and sending the image data to a server; the application program performs physical living body detection on the mobile equipment based on the image data, generates an application program side local light response non-uniform fingerprint based on the currently generated random number seed and the image data when the mobile equipment passes the physical living body detection, and sends the application program side local light response non-uniform fingerprint to the server side; and the server side generates a local light response non-uniform fingerprint of the server side based on the random number seed and the received image data, and performs uniqueness authentication on the mobile device based on the local light response non-uniform fingerprint of the server side and the local light response non-uniform fingerprint of the application side. The reliability of uniqueness authentication of the mobile equipment can be improved.

Description

Technical Field

[0001] This application belongs to the field of computer and information security technology, and in particular relates to a method for unique authentication of mobile devices. Background Technology

[0002] Currently, mobile internet services (such as bank transfers and game equipment transactions) have extremely high requirements for verifying the authenticity of devices (also known as uniqueness authentication). Existing technologies mainly rely on software identity identifiers (IDs) (such as International Mobile Equipment Identity (IMEI) / Anonymous Device Identifier (OAID)) or passive full-image optical response non-uniformity (PRNU) forensics.

[0003] However, existing unique authentication technologies suffer from low reliability due to the following drawbacks:

[0004] 1. Lack of physical liveness verification: Existing solutions passively receive data from cameras. Attackers can easily deceive the server at the data layer by injecting a pre-recorded real mobile video stream using cloud phones or hooking tools. The system cannot determine whether the video was captured in real-time or replayed from a recording.

[0005] 2. Static full-image comparison is inefficient and poses privacy risks: Existing PRNU solutions typically extract and upload full-image (e.g., 12-megapixel) fingerprints, which is not only computationally time-consuming (in seconds), but also makes it easy for hackers to intercept and reuse fixed fingerprint data. Summary of the Invention

[0006] This application provides a method for unique authentication of mobile devices, which can solve the problem of low reliability in unique authentication of mobile devices.

[0007] This application provides a method for unique authentication of mobile devices, including:

[0008] The application controls the mobile device's screen to execute a lighting sequence; the lighting sequence is generated by the server corresponding to the application when it needs to uniquely authenticate the mobile device.

[0009] During the execution of the lighting sequence on the screen, the application uses the front-facing image acquisition device of the mobile device to acquire image data and sends the image data to the server.

[0010] The application performs physical liveness detection on the mobile device based on image data. When the mobile device passes the physical liveness detection, it generates an application-side local light response non-uniform fingerprint based on the currently generated random number seed and image data, and sends the application-side local light response non-uniform fingerprint to the server.

[0011] The server generates a server-side local optical response non-uniform fingerprint based on a random number seed and the received image data, and performs unique authentication of the mobile device based on the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint.

[0012] Optionally, physical liveness detection of mobile devices based on image data includes:

[0013] Calculate the cross-correlation between the brightness signal of the video frame corresponding to the image data and the illumination sequence;

[0014] If the cross-correlation is greater than the preset threshold and the acquisition delay is within the preset range, then the mobile device is determined to have passed the physical liveness detection; otherwise, the mobile device is determined to have failed the physical liveness detection.

[0015] Optionally, calculate the cross-correlation between the video frame luminance signal corresponding to the image data and the illumination sequence, including:

[0016] The NCC algorithm is used to calculate the cross-correlation between the brightness signal of the video frame corresponding to the image data and the illumination sequence.

[0017] Optionally, an application-side local optical response non-uniform fingerprint is generated based on the currently generated random number seed and image data, including:

[0018] Generate pseudo-random coordinates based on the currently generated random number seed;

[0019] Based on pseudo-random coordinates and the cropping size sent by the server, the target region is determined, and the target image data corresponding to the target region is obtained from the image data.

[0020] The target image data is denoised and PRNU extracted to generate a local light response non-uniform fingerprint on the application side.

[0021] Optionally, the target area can be determined based on pseudo-random coordinates and the cropping size sent by the server, including:

[0022] The target region is the area that starts at pseudo-random coordinates and has a size equal to the cropping size sent by the server.

[0023] Optionally, the target image data is denoised and PRNU extracted to generate an application-side local light response non-uniform fingerprint, including:

[0024] Denoising is achieved by performing wavelet transform on the target image data to obtain denoised image data;

[0025] The noise residual between the target image data and the denoised image data is calculated, and Wiener filtering is applied to the noise residual to obtain the local optical response non-uniform fingerprint on the application side.

[0026] Optionally, unique authentication of mobile devices is performed based on server-side local optical response non-uniform fingerprints and application-side local optical response non-uniform fingerprints, including:

[0027] Calculate the PCE value between the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint;

[0028] If the PCE value is greater than the preset PCE value, the mobile device is determined to have passed the uniqueness authentication; otherwise, the mobile device is determined to have failed the uniqueness authentication.

[0029] The above-mentioned solution in this application has the following beneficial effects:

[0030] In the embodiments of this application, the screen of a mobile device is controlled using a temporarily generated illumination sequence, and image data is acquired during the execution of the illumination sequence on the screen. Physical liveness detection of the mobile device is then performed based on the correspondence between the image data and the illumination sequence. Since pre-recorded image data cannot be matched with the temporarily generated illumination sequence, this application can effectively identify whether the current image data is captured in real-time or pre-recorded. Simultaneously, when the mobile device passes physical liveness detection, the application generates a local light response non-uniform fingerprint on the application side based on a random number seed, and the server generates a server-side local light response non-uniform fingerprint based on the same random number seed. Unique authentication is then performed based on these two local light response non-uniform fingerprints. Because the fingerprint data uploaded to the server by the application is completely different each time, hackers cannot attack subsequent authentications by recording a successful fingerprint once. Therefore, this application achieves a unique authentication method resistant to injection and reuse, improving the reliability of unique authentication.

[0031] Other beneficial effects of this application will be described in detail in the following detailed description section. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 A flowchart of a mobile device uniqueness authentication method provided in an embodiment of this application. Detailed Implementation

[0034] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0035] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0036] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0037] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0038] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0039] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0040] To address the low reliability of current mobile device uniqueness authentication methods, this application provides a mobile device uniqueness authentication method. This method utilizes a temporarily generated illumination sequence to control the mobile device's screen and collects image data during the execution of the illumination sequence. Physical liveness detection of the mobile device is then performed based on the correspondence between the image data and the illumination sequence. Since pre-recorded image data cannot be matched with the temporarily generated illumination sequence, this application can effectively identify whether the current image data is captured instantly or pre-recorded. Simultaneously, when the mobile device passes physical liveness detection, the application generates a local light response non-uniform fingerprint on the application side based on a random number seed, and the server generates a server-side local light response non-uniform fingerprint based on the same random number seed. Uniqueness authentication is then performed based on these two local light response non-uniform fingerprints. Because the fingerprint data uploaded by the application to the server is completely different each time, hackers cannot attack subsequent authentications by recording a successful fingerprint once. Therefore, this application achieves a uniqueness authentication method resistant to injection and reuse, improving the reliability of uniqueness authentication.

[0041] The mobile device uniqueness authentication method provided in this application will be illustrated below with reference to specific embodiments.

[0042] like Figure 1 As shown, the mobile device uniqueness authentication method provided in this application includes the following steps:

[0043] Step 11: The application controls the mobile device's screen to execute a lighting sequence; the lighting sequence is generated by the server corresponding to the application when it needs to perform unique authentication of the mobile device.

[0044] The aforementioned application (App) has high requirements for the uniqueness of mobile devices. This can be understood as: when logging into an account using this application, the uniqueness of the mobile device to which the application belongs needs to be verified. For example, this application could be a game app (such as Fantasy Westward Journey) or TikTok, to prevent fraudulent logins and account manipulation.

[0045] The aforementioned mobile devices refer to terminal devices that can install the aforementioned applications, such as tablets and smartphones.

[0046] In some embodiments of this application, the aforementioned illumination sequence is randomly generated during each authentication. Specifically, it can be generated by the server or by the application. For example, the illumination sequence can be: completely white at time T0, completely black at time T1, and flashing at time T2; correspondingly, the mobile device's screen is completely white at time T0, completely black at time T1, and flashing at time T2.

[0047] It is understood that the mobile device unique authentication method in this application embodiment is generally triggered when a user needs to log in to an account on a mobile device application. For example, when a user clicks the login button, the server corresponding to the application sends a request for unique authentication to the application (the actual executor can be understood as the mobile device), and then executes the mobile device unique authentication method provided in this application embodiment. The application (the actual executor can be understood as the mobile device) and the server communicate via a network.

[0048] Step 12: During the execution of the lighting sequence on the screen, the application uses the front-facing image acquisition device of the mobile device to acquire image data and sends the image data to the server.

[0049] The aforementioned front-facing image acquisition device can be a front-facing camera. During the screen's illumination sequence, screen light shines on the user's face or an obstruction (such as a finger), causing diffuse reflection, and photons enter the front-facing camera's complementary metal-oxide-semiconductor (CMOS) sensor. This process relies on a physical screen and physical optical components, which cloud phones lack.

[0050] Step 13: The application performs physical liveness detection on the mobile device based on the image data. When the mobile device passes the physical liveness detection, it generates an application-side local light response non-uniform fingerprint based on the currently generated random number seed and the image data, and sends the application-side local light response non-uniform fingerprint to the server.

[0051] In some embodiments of this application, the purpose of physical liveness detection on the mobile device is to confirm that "the light currently seen by the camera is the light currently emitted by the screen", thereby excluding cloud phones and video injection.

[0052] In some embodiments of this application, the specific implementation of the application performing physical liveness detection on the mobile device based on image data is as follows: calculate the cross-correlation between the brightness signal of the video frame corresponding to the image data and the illumination sequence; if the cross-correlation is greater than a preset threshold and the acquisition delay is within a preset range, then determine that the mobile device has passed the physical liveness detection; otherwise, determine that the mobile device has not passed the physical liveness detection.

[0053] In some preferred embodiments of this application, the normalized cross-correlation (NCC) algorithm can be used to calculate the cross-correlation between the video frame luminance signal corresponding to the image data and the illumination sequence. Here, the video frame luminance signal can be understood as the luminance signal captured by the camera, and the illumination sequence can be understood as the luminance signal commanded by the screen.

[0054] Specifically, the cross-correlation mentioned above can be calculated using the following formula. :

[0055] ;

[0056] In the above formula, Indicates the illumination sequence at the 1st The screen brightness value corresponding to the screen flicker control signal (e.g., 0 represents full black, 1 represents full white) at a given moment. This represents the average screen brightness value corresponding to the illumination sequence within the entire time window (i.e., the duration corresponding to the illumination sequence). Indicates the camera at the Brightness of video frames captured at each moment This represents the average brightness of the video frames captured by the camera within the entire time window.

[0057] The aforementioned acquisition delay refers to the time difference between the moment when the application (App) issues a screen display command (i.e., a command to control the screen to execute a lighting sequence, such as commanding the screen to turn white) and the moment when the application reads the image data acquired by the front-end image acquisition device.

[0058] It should be noted that the preset threshold and preset range can be set according to the actual situation. For example, the preset threshold can be set to 0.85, and the preset range can be from 100ms to 300ms.

[0059] When a mobile device undergoes physical liveness detection, the application generates an application-side local light response non-uniform fingerprint based on a currently generated random number seed and image data. In some embodiments of this application, the application-side local light response non-uniform fingerprint can be generated through the following steps 13.1 to 13.3:

[0060] Step 13.1: Generate pseudo-random coordinates based on the currently generated random number seed.

[0061] The aforementioned random number seed is generated by the server and sent to the application after the mobile device passes physical liveness detection. Along with the random number seed, the server also sends the cropped image data size (e.g., 512). 512).

[0062] In some embodiments of this application, a linear congruent generator (LCG) random algorithm can be used to calculate the random number seed to obtain pseudo-random coordinates.

[0063] Step 13.2: Based on the pseudo-random coordinates and the cropping size sent by the server, determine the target region and obtain the target image data corresponding to the target region from the image data.

[0064] In some embodiments of this application, a region with pseudo-random coordinates as the starting point and a size equal to the cropping size sent by the server can be used as the target region.

[0065] Specifically, you can filter out all image frames in the image data corresponding to a completely white screen, and then for each image frame, crop out an area with pseudo-random coordinates as the starting point and a size equal to the cropping dimension (e.g., 512). The image regions of 512) are then processed, and finally, pixel alignment is performed on the image regions corresponding to all image frames to obtain the target image data.

[0066] Step 13.3: Denoise the target image data and extract PRNU to generate a local light response non-uniform fingerprint on the application side.

[0067] In some embodiments of this application, application-side local light response non-uniform fingerprints can be generated through the following steps 13.31 to 13.32:

[0068] Step 13.31: Perform wavelet transform on the target image data to denoise it, and obtain denoised image data.

[0069] Step 13.32: Calculate the noise residual between the target image data and the denoised image data, and perform Wiener filtering on the noise residual to obtain the local light response non-uniform fingerprint on the application side.

[0070] Specifically, the noise residual between the target image data and the denoised image data is calculated using the following formula. :

[0071] ;

[0072] In the above formula, Represents target image data. This represents the denoised image data.

[0073] In some embodiments of this application, after calculating the noise residual, the noise residual can be processed by Wiener Filter to remove random shot noise and obtain the final local light response non-uniform fingerprint (i.e., application-side local light response non-uniform fingerprint). The application-side local light response non-uniform fingerprint can be understood as the PRNU feature of the target image data obtained by the application side. The essence of the PRNU feature is the "pixel sensitivity difference map" of the CMOS sensor.

[0074] It should be noted that the above applications can implement the relevant steps based on the hardware architecture and computing power of the mobile device.

[0075] Step 14: The server generates a server-side local light response non-uniform fingerprint based on the random number seed and the received image data, and performs unique authentication of the mobile device based on the server-side local light response non-uniform fingerprint and the application-side local light response non-uniform fingerprint.

[0076] In some embodiments of this application, the specific implementation process of the server generating the server-side local optical response non-uniform fingerprint is the same as the process of the application generating the application-side local optical response non-uniform fingerprint described above. The difference is that the server processes the received image data, and the final server-side local optical response non-uniform fingerprint is: the PRNU feature of the target image data obtained by cropping the received image by the server.

[0077] In some embodiments of this application, after the server obtains the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint, uniqueness authentication can be performed as follows: The PCE value between the server-side and application-side local optical response non-uniform fingerprints is calculated; if the PCE value is greater than a preset PCE value, the mobile device is determined to have passed uniqueness authentication; otherwise, the mobile device is determined to have failed uniqueness authentication. It is understood that if the mobile device passes uniqueness authentication, the server allows the account to log in on the current mobile device; if the mobile device fails uniqueness authentication, the server refuses the account to log in on the current mobile device.

[0078] In related technologies, PCE (Peak to Correlation Energy) is a key metric used in signal processing and image processing to measure the similarity between two signals or images, and is often used to evaluate the accuracy of matching.

[0079] In some embodiments of this application, the PCE value between the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint can be calculated using the conventional calculation formula of PCE. The calculation process will not be described in detail here.

[0080] It's worth noting that this application uses PRNU feature comparison based on a random number seed to confirm that "this device is the same device used during registration," thus preventing replay attacks. This also prevents hackers from attacking subsequent authentication attempts by recording a successful fingerprint once.

[0081] In summary, the mobile device uniqueness authentication method provided in this application has the following advantages:

[0082] 1) Absolute physical anti-counterfeiting (closed-loop verification): It utilizes the physical causal chain of screen-light-lens. Cloud phones only have pure digital signals and no physical screen or optical reflection, so they cannot achieve synchronous detection of light illumination, and the simulator cannot respond to millisecond-level screen flicker commands in real time.

[0083] 2) Replay-resistant dynamic fingerprint: A random ROI mechanism is introduced, and the fingerprint data uploaded each time is completely different. Hackers cannot attack the next authentication by recording a successful packet (i.e. fingerprint data).

[0084] 3) Compared to full-image extraction, random ROIs (such as 512) The 512) solution reduces mobile device computation time by 92% (from 1.5s to 0.12s) and data consumption by 95%.

[0085] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for unique authentication of mobile devices, characterized in that, include: The application controls the mobile device's screen to execute lighting sequences; The illumination sequence is generated by the server corresponding to the application controller when it needs to perform unique authentication on the mobile device; During the execution of the illumination sequence on the screen, the application uses the front-facing image acquisition device of the mobile device to acquire image data and sends the image data to the server. The application performs physical liveness detection on the mobile device based on the image data, and when the mobile device passes the physical liveness detection, it generates an application-side local light response non-uniform fingerprint based on the currently generated random number seed and the image data, and sends the application-side local light response non-uniform fingerprint to the server. The server generates a server-side local light response non-uniform fingerprint based on the random number seed and the received image data, and performs unique authentication on the mobile device based on the server-side local light response non-uniform fingerprint and the application-side local light response non-uniform fingerprint.

2. The mobile device uniqueness authentication method according to claim 1, characterized in that, The physical liveness detection of the mobile device based on the image data includes: Calculate the cross-correlation between the video frame brightness signal corresponding to the image data and the illumination sequence; If the cross-correlation is greater than a preset threshold and the acquisition delay is within a preset range, then the mobile device is determined to have passed physical liveness detection; otherwise, the mobile device is determined to have failed physical liveness detection.

3. The mobile device uniqueness authentication method according to claim 2, characterized in that, The calculation of the cross-correlation between the video frame brightness signal corresponding to the image data and the illumination sequence includes: The cross-correlation between the video frame brightness signal corresponding to the image data and the illumination sequence is calculated using the NCC algorithm.

4. The mobile device uniqueness authentication method according to claim 1, characterized in that, The generation of application-side local light response non-uniform fingerprints based on the currently generated random number seed and the image data includes: Generate pseudo-random coordinates based on the currently generated random number seed; Based on the pseudo-random coordinates and the cropping size sent by the server, the target region is determined, and the target image data corresponding to the target region is obtained from the image data; The target image data is denoised and PRNU extracted to generate a local light response non-uniform fingerprint on the application side.

5. The mobile device uniqueness authentication method according to claim 4, characterized in that, The process of determining the target region based on the pseudo-random coordinates and the cropping size sent by the server includes: The region starting from the pseudo-random coordinates and having a size equal to the cropping size sent by the server is taken as the target region.

6. The mobile device uniqueness authentication method according to claim 4, characterized in that, The step of denoising and extracting PRNU from the target image data to generate a local light response non-uniform fingerprint on the application side includes: The target image data is denoised by wavelet transform to obtain denoised image data; The noise residual between the target image data and the denoised image data is calculated, and the noise residual is subjected to Wiener filtering to obtain the local light response non-uniform fingerprint on the application side.

7. The mobile device uniqueness authentication method according to claim 1, characterized in that, The unique authentication of the mobile device based on the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint includes: Calculate the PCE value between the server-side local optical response non-uniform fingerprint and the application-side local optical response non-uniform fingerprint; If the PCE value is greater than the preset PCE value, the mobile device is determined to have passed the uniqueness authentication; otherwise, the mobile device is determined to have failed the uniqueness authentication.

Images