Dynamic peep-proof method and device for transparent display screen, equipment and storage medium

Abstract

The invention relates to the technical field of transparent display screens, and provides a transparent display screen dynamic peep-proof method, device and equipment and a storage medium, and the method comprises the steps: carrying out the detection and identity preliminary screening of individuals in a scene based on an RGB image and a depth point cloud, and obtaining a preliminary screening user list; performing vision field coordinate positioning and tracking processing on the user through eyeball tracking and head posture analysis based on the primarily screened user list to obtain an updated vision field coordinate table; based on the vision field coordinate table and the obtained sensitive content area information, peeping risk assessment and security tunnel geometric modeling processing are carried out, and a dynamic peeping prevention instruction set is obtained; based on the dynamic peep-proof instruction set, brightness compensation driving is carried out on display screen pixels, and cooperative optical driving processing is carried out on the visual angle controllable optical layer; and carrying out real-time feedback and parameter optimization processing on the system state based on the change of the vision field coordinate table and the peep-proof efficiency self-checking signal. According to the method, the visual transparent characteristic of the screen is not damaged while the information privacy is guaranteed.

Description

Technical Field

[0001] This application relates to the field of transparent display technology, and in particular to a method, apparatus, device and storage medium for dynamic privacy protection of transparent display screens. Background Technology

[0002] Transparent display technology, as a cutting-edge field in human-computer interaction, is increasingly being applied in various scenarios such as commercial shop windows, museum display cases, intelligent cockpit display systems, and partition screens in open-plan office environments. Its core value lies in its ability to seamlessly integrate digital information with the underlying physical environment, thereby providing users with an enhanced visual experience that combines information acquisition with contextual awareness. However, while the inherent physical transparency of this technology creates unique value, it also introduces a significant drawback that urgently needs to be addressed: the electronic information presented on the display screen is equally visible to authorized users directly in front of it and to unauthorized observers to its side or even behind it.

[0003] Currently, the industry's common solution to this privacy leak risk is to attach a static anti-spy optical film to the screen surface. While this technical solution can narrow the viewing angle and play a role in preventing peeping to some extent, it is essentially a compromise that sacrifices the core functions of the display screen in exchange for security. This is because it irreversibly damages the screen's transparency and wide-angle viewing capabilities, completely eliminating the technological advantage of dynamic and transparent display effects. It is unable to achieve the intelligent functions of on-demand, dynamic display that protect privacy while also accommodating the display of public information. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for dynamic privacy protection of transparent displays, in order to solve the problems mentioned in the background art.

[0005] In a first aspect, this application provides a dynamic privacy protection method for a transparent display screen, comprising: Based on RGB images and depth point clouds, individuals in the scene are detected and their identities are initially screened to obtain a preliminary user list. Based on this preliminary user list, users are located and tracked using eye tracking and head posture analysis to obtain a real-time updated view coordinate table. Based on the view coordinate table and acquired sensitive content area information, a spying risk assessment and secure tunnel geometric modeling are performed to obtain a dynamic anti-spying instruction set. Based on the dynamic anti-spying instruction set, brightness compensation is applied to the display screen pixels, and collaborative optical driving is performed on the controllable viewing angle optical layer. Based on changes in the view coordinate table and anti-spying effectiveness self-check signals, the system status is fed back in real-time and parameters are adaptively optimized.

[0006] Secondly, this application provides a dynamic privacy screen device for transparent displays, comprising: The system comprises the following modules: an identity screening module, which detects and screens individuals in a scene based on RGB images and depth point clouds to obtain a preliminary user list; a view coordinate table update module, which, based on the preliminary user list, performs view coordinate positioning and tracking on users through eye tracking and head posture analysis to obtain a real-time updated view coordinate table; an anti-peeping instruction generation module, which, based on the view coordinate table and acquired sensitive content area information, performs peeping risk assessment and secure tunnel geometric modeling to obtain a dynamic anti-peeping instruction set; a parameter compensation driving module, which, based on the dynamic anti-peeping instruction set, performs brightness compensation driving on display pixels and collaborative optical driving processing on the controllable viewing angle optical layer; and a system optimization module, which, based on changes in the view coordinate table and anti-peeping effectiveness self-check signals, provides real-time feedback and adaptive parameter optimization of the system status.

[0007] Thirdly, this application provides a terminal device, the terminal device including a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the dynamic anti-peeping method for a transparent display screen as described in any of the preceding claims.

[0008] Fourthly, this application provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the dynamic privacy protection method for a transparent display screen as described in any of the preceding claims.

[0009] This application provides a method, apparatus, device, and storage medium for dynamic anti-peeping on transparent displays. The method includes: detecting and initially screening individuals in a scene based on RGB images and depth point clouds to obtain an initial user list; based on the initial user list, performing visual coordinate positioning and tracking on users through eye tracking and head posture analysis to obtain a real-time updated visual coordinate table; based on the visual coordinate table and acquired sensitive content area information, performing peeping risk assessment and secure tunnel geometric modeling to obtain a dynamic anti-peeping instruction set; based on the dynamic anti-peeping instruction set, performing brightness compensation driving on display pixels and collaborative optical driving processing on the controllable viewing angle optical layer; and based on changes in the visual coordinate table and anti-peeping effectiveness self-check signals, performing real-time feedback and parameter adaptive optimization processing on the system status. This method, on the one hand, dynamically delineates the privacy protection area on the transparent display screen and controls the direction of light propagation, thereby shielding sensitive content only when necessary and maintaining the visual transparency of the display screen in the non-privacy protection state; on the other hand, by tracking the user's field of view in real time and constructing a secure field of view tunnel, the privacy protection can automatically adjust with the user's movement, thereby reducing the user's operational burden; furthermore, through real-time feedback of system status and adaptive optimization of parameters, the robustness of field of view tracking and privacy control is improved, thus adapting to different usage environments and user behaviors. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart illustrating the dynamic privacy protection method for a transparent display screen provided in an embodiment of this application; Figure 2 A schematic block diagram of the structure of the transparent display screen dynamic privacy device provided in the embodiments of this application; Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of this application. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0013] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0014] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0015] It should also be further 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 related listed items and all possible combinations, and includes such combinations.

[0016] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features described herein can be combined with each other.

[0017] Please see Figure 1 , Figure 1 This is a flowchart illustrating the dynamic privacy protection method for a transparent display screen provided in an embodiment of this application, as shown below. Figure 1 As shown, the transparent display screen dynamic privacy protection method provided in this application embodiment includes steps S1 to S5.

[0018] S1. Based on RGB images and depth point clouds, individuals in the scene are detected and their identities are initially screened to obtain a preliminary user list. For example, a color camera captures a sequence of RGB images of the scene, while a depth sensor captures depth point cloud data aligned with the RGB images. Human detection analysis is performed on the RGB images, identifying the bounding boxes of all human targets in the images to form a set of human bounding boxes. Combining the depth point cloud data, the three-dimensional spatial position of each detected human target relative to the transparent display screen is calculated. The face region is located in the RGB images, facial features are extracted, and these features are compared with a pre-stored authorized user feature template. Based on the comparison results, individuals are classified as authorized users or unauthorized individuals. Finally, the human bounding box information, three-dimensional spatial position information, and preliminary identity screening results are correlated and fused to generate a structured preliminary user list, which records the identity category and initial spatial location information of each tracked target.

[0019] S2. Based on the initial screening user list, eye tracking and head posture analysis are used to locate and track the user's visual field coordinates, resulting in a real-time updated visual field coordinate table. For example, for authorized users in the initial screening user list, the pupil center and corneal reflection spot are precisely located from their facial image. Combined with the user's facial depth information, the high-precision coordinates of the authorized user's left and right eyes in three-dimensional space are calculated using stereoscopic vision principles. For unauthorized individuals in the initial screening user list, the average coordinates of their eyes in space are comprehensively estimated based on the orientation angle of their head in the image and the posture of their torso. All users' eye coordinates are uniformly converted to a coordinate system based on the transparent display screen, and these coordinate data are continuously updated over time, forming a dynamic visual field coordinate table reflecting the viewpoint positions of all users.

[0020] S3. Based on the view coordinate table and the acquired sensitive content area information, perform a spying risk assessment and secure tunnel geometric modeling to obtain a dynamic anti-spying instruction set. For example, from the metadata provided by the operating system or running applications, parse the boundaries of areas containing sensitive information in the currently displayed screen to obtain a list of sensitive content areas. Based on the authorized user's eye coordinates and the screen coordinates of the sensitive content areas in the view coordinate table, construct an authorized visual cone connecting the user's eyes to the boundaries of the sensitive areas; the space defined by this cone is the secure view tunnel. Compare the coordinates of unauthorized individuals in the view coordinate table with the geometric space of the secure view tunnel to determine whether an unauthorized individual is located in or near the tunnel, thereby generating a high-risk set. Based on the specific coordinates of unauthorized individuals in the high-risk set, calculate the optical control parameters required to block their view on the corresponding sensitive content areas, and finally summarize to generate a dynamic anti-spying instruction set containing the anti-spying area outline and optical driving parameters.

[0021] S4. Based on the dynamic privacy protection instruction set, brightness compensation driving is performed on the display pixels, and coordinated optical driving processing is performed on the controllable viewing angle optical layer. For example, the dynamic privacy protection instruction set is parsed to identify the screen areas where the privacy protection function needs to be activated. For each pixel in these areas, based on the known light attenuation characteristics of the controllable viewing angle optical layer in narrow viewing angle mode, the original driving signal strength is correspondingly increased to perform brightness and color pre-compensation, generating a compensated pixel driving signal. Simultaneously, based on the optical driving parameters in the dynamic privacy protection instruction set, a corresponding voltage control sequence is generated, which is applied to the controllable viewing angle optical layer unit corresponding to the privacy protection area. Through strict timing synchronization control, the compensated pixel driving signal and the driving voltage sequence of the controllable viewing angle optical layer take effect at the same time, achieving coordinated operation of display and optical control.

[0022] S5. Based on the changes in the visual field coordinate table and the anti-peeping effectiveness self-check signal, the system status is fed back in real time and parameters are adaptively optimized. For example, the updated data of the visual field coordinate table is continuously monitored, and behavioral patterns such as the user's movement speed and gaze duration are analyzed to assess the reliability of the current tracking results and generate a tracking status feedback signal. The anti-peeping effectiveness self-check process is periodically initiated, and the actual anti-peeping effect is evaluated by analyzing the light signals collected from the perspective of unauthorized users, generating an anti-peeping effectiveness feedback signal. The tracking status feedback signal and the anti-peeping effectiveness feedback signal are combined to fine-tune and calibrate the sensitivity parameters of the eye-tracking algorithm, the judgment threshold for risk assessment, and the calculation model of the optical drive parameters, forming a closed-loop optimization process.

[0023] The method provided in this embodiment, on the one hand, dynamically delineates the privacy protection area on the transparent display screen and controls the direction of light propagation, thereby shielding sensitive content only when necessary, thus maintaining the visual transparency of the display screen in the non-privacy protection state; on the other hand, by tracking the user's field of view in real time and constructing a secure field of view tunnel, the privacy protection can automatically adjust with the user's movement, thereby reducing the user's operational burden; furthermore, through real-time feedback of system status and adaptive optimization of parameters, the robustness of field of view tracking and privacy control is improved, thereby adapting to different usage environments and user behaviors.

[0024] In some embodiments, the detection and initial identity screening of individuals in the scene includes: S11, performing human target detection processing based on the RGB image to obtain a set of human bounding boxes. For example, the acquired RGB color image is preprocessed, and then an analysis method based on predefined features or a neural network model is used to scan the entire image to identify all regions that conform to human morphological features. For each identified human region, the position and size parameters of the smallest rectangle that can completely enclose the region are calculated in the image coordinate system. The position parameter is typically defined by the coordinate values ​​of the rectangle's vertices, and the size parameter is defined by the width and height values ​​of the rectangle. All detected human bounding box information is collected to form a set of human bounding boxes containing multiple bounding box data. This set provides the target range for subsequent spatial localization and identity analysis.

[0025] S12. Based on the depth point cloud, perform 3D spatial position estimation processing to obtain the approximate spatial coordinates of each individual. For example, acquire depth point cloud data that is aligned temporally and spatially with the RGB image. This data records the distance information from the object corresponding to each pixel in the scene to the sensor. Map the human bounding box set obtained in S11 onto this depth point cloud to extract the 3D spatial point set corresponding to all pixels within each bounding box. Calculate the coordinates of the geometric center point of each point set in 3D space, or calculate the average coordinates of all points in the point set. Use this calculated 3D coordinate point as the spatial position representative of the corresponding individual. This coordinate is described in a 3D coordinate system with the sensor as the origin, providing approximate position information of each individual in the physical world.

[0026] S13. Based on the face region in the RGB image, perform face feature extraction and comparison processing to obtain the initial screening result of individual identity. For example, within each human body bounding box determined in S11, a detection method specifically targeting facial morphological features is used again to locate the specific region where the face is located. From the located face region image, feature data that characterizes the unique features of the face is extracted, and these feature data constitute a feature vector. This feature vector is compared one by one with all template feature vectors in a pre-established and stored authorized user face feature template library to calculate their similarity degree. According to a preset similarity threshold judgment rule, the individual is classified as an identified authorized user or an unidentified unauthorized individual, thereby obtaining the initial screening result of individual identity.

[0027] S14. Based on the human bounding box set, coarse spatial coordinates, and initial screening results of individual identities, perform data fusion processing to generate the initial screening user list. For example, associate and bind the human bounding box information, coarse spatial coordinate information, and initial identity screening results generated for the same individual in S11, S12, and S13 respectively. Create an independent data record item for each successfully tracked individual, which fully includes the individual's target identifier, bounding box parameters, 3D spatial coordinates, and identity category. Summarize all individuals' data record items into a structured list, which is the initial screening user list. This list provides a clear target set and initial state for subsequent view coordinate localization and tracking processing.

[0028] The method provided in this embodiment, on the one hand, obtains the appearance and location information of the target by processing RGB images and depth point clouds in parallel, thereby constructing a multi-dimensional perception of individuals in the scene and providing rich initial data for subsequent processing; on the other hand, by extracting features in the face region and comparing them with pre-stored templates, it realizes the automatic differentiation between authorized users and unauthorized individuals, thereby laying the identity foundation for dynamic privacy protection; furthermore, by fusing and binding appearance, location and identity information, it generates a structured initial screening user list, thereby providing a clear and consistent processing target for the entire system.

[0029] In some embodiments, the visual coordinate localization and tracking processing for the user includes: S21, performing binocular eye tracking processing based on the authorized user information in the initial screening user list to obtain high-precision three-dimensional coordinates of the authorized user's left and right eyes. For example, the facial image region and depth information corresponding to the individual marked as an authorized user are obtained from the initial screening user list. Within this facial image region, the pupil center positions of the left and right eyes are precisely located, and simultaneously, the positions of the reflected light spots generated on the cornea by the infrared illumination source are located. Combining the depth point cloud data of the authorized user's face, and using stereo vision and the pupil-corneal reflection vector principle, the high-precision coordinates of the center of the authorized user's left and right eyeballs in three-dimensional space are calculated respectively. These coordinates are described in a unified world coordinate system with a transparent display screen as a reference.

[0030] S22. Based on the information of unauthorized individuals in the initial screening user list, a comprehensive analysis of head orientation and body posture is performed to obtain the estimated visual field coordinates of the unauthorized individuals. For example, target information marked as unauthorized individuals is obtained from the initial screening user list. Based on the individual's head image region, the geometric distribution of facial feature points (such as the corners of the eyes, the tip of the nose, and the corners of the mouth) is analyzed to calculate the horizontal deflection angle and vertical pitch angle of the head. Simultaneously, the gaze direction is further constrained by combining the individual's torso contour and posture in the image. Combining the analysis results of head orientation and body posture, the midpoint of the line connecting the unauthorized individual's eyes is estimated, and its coordinates in three-dimensional space are used as the estimated visual field coordinates of the unauthorized individual.

[0031] S23. Based on the coordinate data of all users, perform spatiotemporal synchronization and coordinate system registration processing to obtain the view coordinate table unified with the display screen's reference coordinate system. For example, the high-precision binocular coordinates of authorized users calculated in S21 and the view coordinates of unauthorized individuals estimated in S22 are all transformed into the same three-dimensional world coordinate system with the center of the transparent display screen as the origin and the display screen plane as the reference plane. This ensures that all coordinate data are synchronized in timestamps to reflect the scene state at the same moment. All user coordinate data after transformation and synchronization are organized according to a preset data structure to form a real-time updated view coordinate table. This view coordinate table dynamically reflects the spatial distribution of the viewpoints of all tracked users at the current moment.

[0032] The method provided in this embodiment, on the one hand, obtains the precise position of the authorized user's eyes in space by performing high-precision binocular eye tracking, thereby providing a key geometric endpoint for constructing a directional optical tunnel; on the other hand, it estimates the approximate direction of the gaze of unauthorized individuals with low computational cost by performing head and body posture analysis, thereby achieving broad coverage of potential spying risks; furthermore, by unifying the coordinates of all users to the display screen's reference coordinate system, a common reference describing the spatial relationship between the user and the display screen is established, thereby ensuring the consistency of subsequent risk assessment and optical control calculations.

[0033] In some embodiments, the binocular eye-tracking processing to obtain high-precision three-dimensional coordinates of the authorized user's left and right eyes includes: S211, based on the facial region image of the authorized user in the initial screening user list, performing pupil center and corneal reflection spot localization processing to obtain two-dimensional image coordinates of eye feature points. For example, a region of interest containing a single eye is extracted from the authorized user's facial region image. This eye region image is preprocessed to enhance contrast and reduce noise interference. The region is analyzed using algorithms based on grayscale gradients or ellipse fitting to find the darkest circular or elliptical region in the image, which is identified as the pupil, and its center point is calculated in the image pixel coordinate system. Simultaneously, within the same eye region, a tiny spot with the highest brightness generated by near-infrared light source illumination is searched; this spot is identified as a corneal reflection point, and its center point is calculated in the image. The above operations were performed on the left and right eyes respectively, resulting in a set of two-dimensional image coordinates characterizing the physical properties of the eyeball, including the center of the left pupil, the left corneal reflection point, the center of the right pupil, and the right corneal reflection point.

[0034] S212. Based on the data corresponding to the facial region in the depth point cloud, perform three-dimensional reconstruction processing of the facial surface to obtain the depth information of the face in space. For example, acquire depth point cloud data precisely registered with the authorized user's facial region image. This depth point cloud data is directly provided by a depth sensor, where each point contains three-dimensional spatial coordinates relative to the sensor. Extract all three-dimensional points corresponding to the authorized user's facial region from the complete scene depth point cloud. Integrate these three-dimensional spatial points and construct a geometric model describing the continuous distribution of the authorized user's facial surface in three-dimensional space using surface fitting or point cloud registration algorithms. This three-dimensional facial surface model provides the necessary depth reference for calculating the spatial coordinates of the eyeballs, and its accuracy directly determines the accuracy of subsequent triangulation calculations.

[0035] S213. Based on the two-dimensional image coordinates of the eye feature points and the depth information, perform stereoscopic vision triangulation to obtain the high-precision three-dimensional coordinates of the authorized user's left and right eyes. For example, the two-dimensional image coordinates of the pupil center and corneal reflection spot obtained in S211 are combined with the three-dimensional facial surface model established in S212. Using a pinhole model of camera imaging, each two-dimensional image coordinate point is back-projected into a line of sight in three-dimensional space. Utilizing the relative positional relationship between the pupil center and the corneal reflection spot—that is, the characteristic that the vector of the pupil center relative to the corneal reflection spot changes with eye movement—and combined with the three-dimensional geometric constraints of the facial surface, a stereoscopic vision triangulation model is constructed. By solving this model through iterative optimization or analytical geometric methods, the high-precision three-dimensional coordinates of the authorized user's left and right eye centers in a unified world coordinate system are finally calculated.

[0036] The method provided in this embodiment, on the one hand, provides stable and reliable two-dimensional input data for three-dimensional spatial calculation by locating two key eye feature points: the pupil center and the corneal reflective spot; on the other hand, it establishes an accurate depth reference system for coordinate transformation from two-dimensional to three-dimensional by using the depth point cloud data of the facial region for three-dimensional reconstruction; furthermore, it fuses and calculates two-dimensional image coordinates with three-dimensional facial geometry through a stereo vision triangulation model, thereby mapping image features to precise eye coordinates in physical space, thus providing indispensable basic data for the subsequent construction of precise optical control.

[0037] In some embodiments, the process of performing spying risk assessment and secure tunnel geometric modeling includes: S31, performing sensitive content area identification processing based on metadata obtained from the operating system or application programming interface to obtain a list of sensitive content areas. For example, through the application programming interface provided by the operating system, the attribute information and content description of currently active windows are monitored or actively queried. This metadata includes, but is not limited to, the window's title, class name, control type, and content security level label marked by the application itself. This metadata is parsed to identify windows or controls explicitly marked as containing sensitive information, such as password input boxes, private message dialog boxes, specific file browser windows, or financial application interfaces. The position and size parameters of these windows or controls in the screen coordinate system are obtained, typically defined as rectangular bounding boxes in pixel coordinates. The screen coordinate boundary information of all identified sensitive content areas is collected and organized to form a list of sensitive content areas.

[0038] S32. Based on the authorized user coordinates in the view coordinate table and the sensitive content region list, perform authorized visual cone construction processing to define the geometric space of the safe view tunnel. For example, read the three-dimensional spatial coordinates of the authorized user's left and right eyes from the real-time updated view coordinate table. For each region in the sensitive content region list, connect each boundary pixel on the screen to the authorized user's left and right eye coordinates respectively, forming two sets of ray clusters in three-dimensional space. A visual cone is formed by the ray connecting the left eye to the region boundary, and similarly, another visual cone is formed by the ray connecting the right eye to the region boundary. These two visual cones together define a three-dimensional, bi-frustum-shaped geometric space, which is the safe view tunnel. Any object located within this tunnel can emit light that reaches the authorized user's eyes, thus being clearly visually perceived.

[0039] S33. Based on the coordinates of unauthorized individuals in the viewport coordinate table and the geometric space of the secure viewport tunnel, perform intrusion risk assessment to generate a high-risk set. For example, read the three-dimensional spatial coordinates of all unauthorized individuals from the viewport coordinate table. For each unauthorized individual coordinate, calculate its relative positional relationship with the geometric space of each secure viewport tunnel constructed in S32. The assessment logic includes checking whether the unauthorized individual coordinate point is located inside any secure viewport tunnel, or whether the distance to the tunnel boundary is less than a preset tolerance threshold. If one of the above conditions is met, it is determined that the unauthorized individual poses a spying risk to the sensitive content area protected by the secure viewport tunnel. This combination of <unauthorized individual, sensitive content area> is recorded as a risk item, and all recorded risk items together constitute a high-risk set.

[0040] S34. Based on the high-risk set, perform anti-peeping optical parameter solving to obtain the dynamic anti-peeping instruction set. For example, traverse each risk item in the high-risk set. For each item, based on the specific spatial coordinates of the unauthorized individual in the risk item and the position of its associated sensitive content area on the screen, use a geometric optics calculation model to solve for the light control parameters required to block the view of the sensitive content area from the unauthorized individual's perspective. These parameters specifically manifest as: the driving signal (e.g., voltage distribution) that needs to be applied to the controllable viewing angle optical layer on the corresponding pixel of the sensitive content area. This driving signal will configure the optical layer to generate a narrow beam in that area, with the main lobe precisely pointing towards the authorized user and forming a radiation null point in the direction of the unauthorized individual. Integrate the anti-peeping area contour information corresponding to all risk items with the solved optical driving parameters, and package them to generate the dynamic anti-peeping instruction set.

[0041] The method provided in this embodiment, on the one hand, automatically identifies sensitive content by parsing the metadata of the operating system and applications, thereby reducing the reliance on manual user intervention and thus automating privacy protection; on the other hand, it defines a secure field-of-view tunnel by constructing a visual cone for authorized users, thereby transforming the abstract concept of privacy into a specific geometric space problem, thus providing a mathematical basis for accurate risk assessment; furthermore, by performing intrusion judgment on the coordinates of unauthorized individuals and the secure tunnel and solving for optical parameters, the risk assessment results are directly mapped into executable hardware control instructions, thereby completing a closed loop from environmental perception to protective actions.

[0042] In some embodiments, the step of performing brightness compensation driving on the display screen pixels and performing cooperative optical driving processing on the controllable viewing angle optical layer includes: S41, based on the privacy area information contained in the dynamic privacy instruction set, performing brightness and color gain compensation calculation processing on the original pixel driving value of the area to obtain the compensated pixel driving signal.

[0043] For example, the dynamic privacy instruction set is parsed to extract the defined privacy area contour, which clarifies the set of pixels on the screen that require the activation of the narrow viewing angle optical mode. The original pixel drive values, which would normally be applied to these pixels in normal display mode, are obtained. Based on the known optical characteristics of the controllable viewing angle optical layer in narrow viewing angle operation mode, particularly its light transmittance curves at different viewing angles, the gain coefficient required to compensate for the brightness attenuation caused by the narrow viewing angle mode is calculated. This gain coefficient is applied to the original pixel drive values, performing multiplication or lookup table operations to synchronously adjust color components to maintain color accuracy, thereby generating a set of pre-compensated, higher-intensity pixel drive signals. This ensures that the image brightness and color saturation viewed by authorized users in narrow viewing angle mode are consistent with the surrounding non-privacy areas.

[0044] S42. Based on the optical driving parameters contained in the dynamic anti-spy instruction set, generate a controllable viewing angle optical layer driving voltage sequence corresponding to the anti-spy area.

[0045] For example, the optical driving parameters contained in the dynamic privacy control instruction set are parsed. These parameters define the electrical control signals that need to be applied to the controllable viewing angle optical layer unit corresponding to the privacy area to achieve specific light direction control. The abstract optical driving parameters are mapped to specific digital instructions that can be recognized by a digital-to-analog converter or driver chip. Based on the shape and location of the privacy area and the specific addressing mode of the optical layer driving circuit, these digital instructions are organized into a temporally and spatially ordered sequence of driving voltages. This voltage sequence is designed to precisely and synchronously change the state of the controllable viewing angle optical layer unit within the privacy area, switching it from a wide-angle scattering mode to a preset narrow-angle directional light emission mode.

[0046] S43. Based on the synchronization timing signal, perform timing alignment output processing on the compensated pixel driving signal and the driving voltage sequence.

[0047] For example, a synchronization timing signal generated from the system master clock or display controller is received, which serves as the timing reference for the entire driving process. Triggered by this signal, the compensated pixel driving signal generated in step S41 is applied to the corresponding pixel of the transparent display screen through the display driving circuit. At a precise moment within the exact same frame period, the driving voltage sequence generated in step S42 is applied to the unit on the controllable viewing angle optical layer corresponding to the privacy protection area through a dedicated optical layer driving circuit. This strict timing alignment control ensures that the pixel's emission state and the optical layer's viewing angle modulation state change simultaneously, achieving seamless coordination between pixel emission and light direction control, and avoiding visual asynchrony or image tearing.

[0048] The method provided in this embodiment, on the one hand, compensates for the light loss introduced by the narrow-view optical mode by performing pre-compensation calculations on the brightness and color of the pixels in the privacy area, thereby maintaining the consistency of the visual experience under the authorized user's viewpoint; on the other hand, it converts the optical driving parameters into specific driving voltage sequences, thereby converting the calculation instructions into physically executable optical layer control actions, thereby achieving precise programming of the light propagation direction; furthermore, through strict timing alignment output control, it ensures a high degree of synchronization between image display and optical modulation in time, thereby avoiding privacy function failure or visual defects caused by timing deviations.

[0049] In some embodiments, the real-time feedback and parameter adaptive optimization of the system status includes: S51, performing user behavior intent analysis and tracking confidence assessment based on the dynamic changes of the field coordinate table to obtain a tracking status feedback signal.

[0050] For example, continuously monitor the changes in all user coordinates in the view coordinate table over time. Analyze this data to calculate the user's movement velocity vector, the smoothness of the movement trajectory, and the dwell time in front of a specific screen area. Based on a pre-defined behavior pattern library, infer the user's current behavior intent, such as distinguishing whether the user is actively interacting with the screen or simply passing by. Simultaneously, evaluate the stability and reliability of the current eye-tracking or head pose analysis algorithm outputs, for example, by generating a tracking confidence score based on the continuity of feature point tracking or the noise level of the coordinate data. Combine the behavior intent inference results with the tracking confidence score to form a tracking status feedback signal.

[0051] S52. Based on the preset anti-spy performance self-test mechanism, perform directional light leakage assessment processing to obtain anti-spy performance feedback signal.

[0052] For example, a built-in privacy shutter effectiveness self-test process is initiated periodically or during system idle periods. This process may include controlling the display screen to show a specific test pattern within the privacy shutter area. Using photosensitive sensors positioned at the screen edge with their light-sensing direction simulating the typical viewing angle of an unauthorized user, the intensity signal of light leaking from the current privacy shutter area is collected. The collected light signal intensity is compared to a preset threshold representing an acceptable leakage level. Based on the comparison result, it is determined whether the actual optical isolation effect of the current privacy shutter meets the expected standard, and a quantified or binary privacy shutter effectiveness feedback signal is generated to indicate whether the privacy shutter effectiveness is normal or suboptimal.

[0053] S53. Based on the tracking status feedback signal and the anti-peeping effectiveness feedback signal, perform dynamic calibration processing on the sensitivity parameters and optical drive parameters of the field of view coordinate positioning.

[0054] For example, a tracking status feedback signal from S51 and a privacy protection effectiveness feedback signal from S52 are received. If the tracking status feedback signal indicates that the user is in a fast-moving state or the tracking confidence is low, the filtering parameters or prediction parameters in the field-of-view coordinate positioning algorithm are fine-tuned accordingly to improve the system's ability to track dynamic targets or reduce output coordinate jitter. If the privacy protection effectiveness feedback signal indicates that light leakage exceeds expectations, the calculation model of the optical drive parameters in the dynamic privacy protection command set is fine-tuned according to the relative orientation of the leakage sensor, for example, slightly increasing the beam suppression depth in a specific direction. These calibration operations are incremental and small-scale, designed to adapt the system parameters to the current environment and usage conditions, forming a continuous optimization loop.

[0055] The method provided in this embodiment, on the one hand, evaluates behavioral intent and tracking quality by analyzing the dynamic changes of user coordinates, thereby providing the system with the ability to understand user scenarios and judge its own perception reliability; on the other hand, it establishes an objective verification mechanism for the actual effect of anti-peeping by periodically evaluating directional light leakage; furthermore, it dynamically calibrates the core parameters of the system by using the feedback signals of tracking status and anti-peeping effectiveness, thereby forming a self-optimizing control loop and improving the system's adaptability and robustness under different operating conditions.

[0056] Please see Figure 2 , Figure 2 A schematic block diagram of the structure of the transparent display screen dynamic privacy device 100 provided in the embodiments of this application is shown below. Figure 2 As shown, the transparent display screen dynamic privacy device 100 provided in this application embodiment includes: The identity screening module 110 is used to detect and screen individuals in a scene based on RGB images and depth point clouds to obtain a preliminary user list. The view coordinate table update module 120 is used to perform view coordinate positioning and tracking on users based on the preliminary user list through eye tracking and head posture analysis to obtain a real-time updated view coordinate table. The anti-peeping instruction generation module 130 is used to perform peeping risk assessment and secure tunnel geometric modeling based on the view coordinate table and acquired sensitive content area information to obtain a dynamic anti-peeping instruction set. The parameter compensation driving module 140 is used to perform brightness compensation driving on display pixels and collaborative optical driving processing on the controllable viewing angle optical layer based on the dynamic anti-peeping instruction set. The system optimization module 150 is used to provide real-time feedback and adaptive parameter optimization processing on the system status based on changes in the view coordinate table and anti-peeping effectiveness self-check signals.

[0057] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the device and each module described above can be referred to the process in the aforementioned embodiment of the dynamic anti-peeping method for transparent displays, and will not be repeated here.

[0058] The transparent display screen dynamic privacy device 100 provided in the above embodiments can be implemented as a computer program, which can be used in, for example... Figure 3 The terminal device 200 shown is running on it.

[0059] Please see Figure 3 , Figure 3 The present invention provides a schematic block diagram of the structure of a terminal device 200. The terminal device 200 includes a processor 201 and a memory 202, which are connected via a device bus 203. The memory 202 may include a non-volatile storage medium and internal memory.

[0060] The non-volatile storage medium can store a computer program. The computer program includes program instructions that, when executed by the processor 201, cause the processor 201 to perform any of the aforementioned dynamic privacy protection methods for transparent displays.

[0061] The processor 201 provides computing and control capabilities to support the operation of the entire terminal device 200.

[0062] The internal memory provides an environment for the execution of computer programs in non-volatile storage media. When the computer program is executed by the processor 201, the processor 201 can execute any of the above-mentioned dynamic anti-peeping methods for transparent displays.

[0063] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal device 200 involved in the present application. The specific terminal device 200 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0064] It should be understood that processor 201 can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0065] In some embodiments, the processor 201 is configured to run a computer program stored in memory to perform the following steps: Based on RGB images and depth point clouds, individuals in the scene are detected and their identities are initially screened to obtain a preliminary user list. Based on this preliminary user list, users are located and tracked using eye tracking and head posture analysis to obtain a real-time updated view coordinate table. Based on the view coordinate table and acquired sensitive content area information, a spying risk assessment and secure tunnel geometric modeling are performed to obtain a dynamic anti-spying instruction set. Based on the dynamic anti-spying instruction set, brightness compensation is applied to the display screen pixels, and collaborative optical driving is performed on the controllable viewing angle optical layer. Based on changes in the view coordinate table and anti-spying effectiveness self-check signals, the system status is fed back in real-time and parameters are adaptively optimized.

[0066] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the terminal device 200 described above can be referred to the process of the aforementioned dynamic anti-peeping method for transparent displays, and will not be repeated here.

[0067] This application also provides a computer-readable storage medium storing a computer program that, when executed by one or more processors, causes the one or more processors to implement the dynamic privacy protection method for a transparent display screen as provided in this application.

[0068] The computer-readable storage medium can be an internal storage unit of the terminal device 200 described in the foregoing embodiments, such as a hard disk or memory of the terminal device 200. The computer-readable storage medium can also be an external storage device of the terminal device 200, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., provided with the terminal device 200.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A dynamic privacy protection method for a transparent display screen, characterized in that, include: Based on RGB images and depth point clouds, individuals in the scene are detected and their identities are initially screened to obtain a preliminary list of users. Based on the initial user list, eye tracking and head posture analysis are used to locate and track users' visual coordinates, resulting in a real-time updated visual coordinate table. Based on the visual coordinate table and the acquired sensitive content area information, a spying risk assessment and secure tunnel geometric modeling are performed to obtain a dynamic anti-spying instruction set. Based on the dynamic anti-spying instruction set, brightness compensation is applied to the display screen pixels, and collaborative optical driving is applied to the controllable viewing angle optical layer. Based on the changes in the field of view coordinate table and the anti-peeping effectiveness self-test signal, the system status is fed back in real time and the parameters are adaptively optimized.

2. The method according to claim 1, characterized in that, The detection and initial identity screening of individuals in the scene includes: performing human target detection processing based on the RGB image to obtain a set of human bounding boxes; performing three-dimensional spatial position estimation processing based on the depth point cloud to obtain the coarse spatial coordinates of each individual; performing facial feature extraction and comparison processing based on the face region in the RGB image to obtain the initial identity screening result of the individual; and performing data fusion processing based on the set of human bounding boxes, the coarse spatial coordinates, and the initial identity screening result of the individual to generate the initial screening user list.

3. The method according to claim 1, characterized in that, The process of locating and tracking user visual coordinates includes: performing binocular eye tracking based on the authorized user information in the initial screening user list to obtain high-precision three-dimensional coordinates of the left and right eyes of the authorized users; performing comprehensive analysis of head orientation and body posture based on the information of unauthorized individuals in the initial screening user list to obtain estimated visual coordinates of the unauthorized individuals; and performing spatiotemporal synchronization and coordinate system registration based on the coordinate data of all users to obtain the visual coordinate table unified with the display screen reference coordinate system.

4. The method according to claim 3, characterized in that, The process of performing binocular eye tracking to obtain high-precision three-dimensional coordinates of the authorized user's left and right eyes includes: performing pupil center and corneal reflection spot localization processing based on the facial region image of the authorized user in the initial screening user list to obtain two-dimensional image coordinates of eye feature points; performing three-dimensional reconstruction processing of the facial surface based on the data corresponding to the facial region in the depth point cloud to obtain the depth information of the face in space; and performing stereo vision triangulation processing based on the two-dimensional image coordinates of the eye feature points and the depth information to obtain high-precision three-dimensional coordinates of the authorized user's left and right eyes.

5. The method according to claim 1, characterized in that, The process of assessing the risk of peeping and geometrically modeling the secure tunnel includes: identifying sensitive content areas based on metadata obtained from the operating system or application programming interface (API) to obtain a list of sensitive content areas; constructing authorized visual cones based on the coordinates of authorized users in the view coordinate table and the list of sensitive content areas to define the geometric space of the secure view tunnel; determining intrusion risk based on the coordinates of unauthorized individuals in the view coordinate table and the geometric space of the secure view tunnel to generate a high-risk set; and solving for anti-peeping optical parameters based on the high-risk set to obtain the dynamic anti-peeping instruction set.

6. The method according to claim 1, characterized in that, The step of performing brightness compensation driving on the display screen pixels and coordinating optical driving processing on the controllable viewing angle optical layer includes: performing brightness and color gain compensation calculation processing on the original pixel driving value of the region based on the privacy region information contained in the dynamic privacy instruction set to obtain the compensated pixel driving signal; generating a controllable viewing angle optical layer driving voltage sequence corresponding to the privacy region based on the optical driving parameters contained in the dynamic privacy instruction set; and performing time-aligned output processing on the compensated pixel driving signal and the driving voltage sequence based on a synchronization timing signal.

7. The method according to claim 1, characterized in that, The real-time feedback and adaptive parameter optimization of the system status includes: performing user behavior intent analysis and tracking confidence assessment based on the dynamic changes of the field of view coordinate table to obtain a tracking status feedback signal; performing directional light leakage assessment based on a preset anti-peeping effectiveness self-checking mechanism to obtain an anti-peeping effectiveness feedback signal; and performing dynamic calibration of the sensitivity parameters and optical drive parameters of the field of view coordinate positioning based on the tracking status feedback signal and the anti-peeping effectiveness feedback signal.

8. A dynamic privacy screen device for a transparent display, characterized in that, include: The identity screening module is used to detect and screen individuals in the scene based on RGB images and depth point clouds to obtain a preliminary list of users. The field of view coordinate table update module is used to perform field of view coordinate positioning and tracking processing on users based on the initial screening user list through eye tracking and head posture analysis, so as to obtain a field of view coordinate table that is updated in real time. The anti-peeping instruction generation module is used to perform peeping risk assessment and secure tunnel geometric modeling based on the field coordinate table and the acquired sensitive content area information to obtain a dynamic anti-peeping instruction set; The parameter compensation driving module is used to perform brightness compensation driving on the display screen pixels based on the dynamic anti-peeping instruction set, and to perform cooperative optical driving processing on the controllable viewing angle optical layer. The system optimization module is used to perform real-time feedback and adaptive parameter optimization of the system status based on the changes in the field of view coordinate table and the anti-peeping effectiveness self-test signal.

9. A terminal device, characterized in that, The terminal device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the dynamic anti-peeping method for a transparent display screen as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the dynamic privacy protection method for a transparent display screen as described in any one of claims 1 to 7.

Images