HUD display adjusting method and system

Abstract

The invention provides an HUD display adjusting method and system. The method comprises the following steps: predicting a shielding area of a windshield wiper in an HUD display area at the next preset time, and determining shielding duration; if the shielding duration exceeds a first time threshold, determining a sight line focus area in the HUD display area, and calculating an area overlapping ratio between the shielding area and the sight line focus area; if the area overlapping rate is greater than a preset area overlapping rate threshold value, performing image compensation on HUD display information of the overlapping area; otherwise, carrying out lightweight compensation on the HUD display information of the overlapped region; and if the shielding duration is less than the first time threshold, shifting the HUD display information to be shielded by the windshield wiper. According to the method, through threshold judgment and dynamic compensation / offset, safe, efficient and user-friendly information management is realized, the driving safety is improved, the resource use is optimized, and the use experience is enhanced.

Description

Technical Field

[0001] This application relates to the field of HUD display technology, and in particular to a HUD display adjustment method and system. Background Technology

[0002] With the rapid development of automotive intelligence, Head-Up Display (HUD) technology has become a key feature of modern vehicles. By projecting crucial information (such as vehicle speed and navigation prompts) in front of the driver's line of sight, it effectively reduces eye shifts and improves driving safety and convenience. HUDs typically project information onto the windshield, ensuring the driver can simultaneously focus on road conditions and vehicle data. However, during rain, the windshield wipers repeatedly sweep across the windshield, causing dynamic obstruction of the HUD display area. Specifically, this obstructs parts of the external view, such as the lane, leading to visual fragmentation for the driver; or it covers the HUD display information, making it blurry and difficult for the driver to see. This obstruction not only disrupts the continuity of information but may also cause the driver to miss crucial data, increasing the risk of accidents. Existing HUD occlusion handling solutions mainly focus on dynamically adjusting the HUD display position to avoid the wipers or performing information compensation. However, this approach fails to comprehensively consider dynamic factors such as the duration of occlusion, the driver's visual focus, and the area overlap rate, resulting in coarse and inaccurate information compensation. Furthermore, if the compensated area or the adjusted display position is not within the driver's current field of vision, it may still lead to missing critical information and causing driving risks. Summary of the Invention

[0003] To address the aforementioned technical problems, this application provides a method and system for adjusting HUD displays.

[0004] A HUD display adjustment method includes: predicting the area obstructed by the wiper in the HUD display area at a preset time, and determining the duration of obstruction; if the duration of obstruction exceeds a first time threshold, determining the focal point area of ​​the line of sight in the HUD display area, and calculating the area overlap rate between the obstructed area and the focal point area; if the area overlap rate is greater than a preset area overlap rate threshold, performing image compensation on the HUD display information in the overlapping area; otherwise, performing minor compensation on the HUD display information in the overlapping area; if the duration of obstruction is less than the first time threshold, shifting the HUD display information obstructed by the wiper.

[0005] This solution predicts the area and duration of wiper obstruction and proactively compensates for or shifts the displayed information before obstruction occurs, achieving a smooth transition and ensuring that critical driving data, such as navigation or alerts in the focus area, remains visible. Combining real-time prediction with proactive adjustments, it compensates for the most critical areas during prolonged obstruction by analyzing overlap rates based on the viewpoint. For short-term obstruction, an offset strategy is employed, adapting to different scenarios to avoid over-processing and ensuring that resources, such as processing power, are used only where necessary. Information offset is used to smoothly move information during brief obstructions, avoiding the frustration of flickering or disappearing information and maintaining a natural and smooth interface. Focus area compensation prioritizes areas of user interest when overlap rates are high, improving comfort. This method, through threshold judgment and dynamic compensation / offset, achieves safe, efficient, and user-friendly information management, improving driving safety, optimizing resource utilization, and enhancing the user experience.

[0006] Furthermore, predicting the occlusion area of ​​the wiper in the HUD display area at the next preset time includes: calculating the predicted position angle of the wiper at the next preset time, and determining the occlusion area based on the predicted position angle.

[0007] In this solution, the obstruction area is determined by calculating and predicting the wiper position angle, and the future obstruction area of ​​the wiper on the windshield is predicted, ensuring that critical driving information is not interrupted, which can significantly improve driving safety, information display continuity and system response efficiency.

[0008] Further, the step of calculating the predicted position angle of the wiper at the next preset time includes: acquiring wiper angle data, the angle data including: the wiper position angle, angular velocity and angular acceleration at the current moment; performing data calibration and fusion on the angle data to obtain calibrated angle data; and calculating the predicted position angle of the wiper at the next preset time based on the calibrated angle data.

[0009] In this solution, wiper angle data calibration fusion calculation refers to integrating raw sensor data and combining it with dynamic calibration algorithms to eliminate errors such as mechanical clearance and temperature drift, ultimately outputting a high-precision real-time wiper angle and future position prediction. Its core lies in overcoming the performance limitations of a single sensor through multi-source data cross-validation and compensation. Wiper angle data calibration fusion calculation, through multi-source data collaborative optimization, can significantly improve the accuracy and robustness of the predicted position angle, providing reliable data support for HUD occlusion compensation.

[0010] Furthermore, the step of calibrating and fusing the angle data to obtain calibrated angle data includes: dynamically generating a wiper occlusion mask; determining the visual angle of the wiper using the occlusion mask; and calibrating and fusing the angle data based on the visual angle to obtain calibrated angle data.

[0011] In this solution, the occlusion of the windshield wipers on the windshield is determined by visual technology, thereby outlining the contour of the wipers and determining their visual angle. Based on the visual angle of the wipers, the angle data of the wipers obtained by the sensors is fused and calibrated to improve the accuracy of the data.

[0012] Furthermore, determining the focal point area of ​​the gaze within the HUD display area includes: determining the focal coordinates of the gaze within the HUD display area, and expanding the area around the focal coordinates to obtain the focal point area of ​​the gaze.

[0013] In this solution, the coordinates of the driver's gaze focus can be captured in real time through dynamic tracking technology and expanded into a dynamic area, namely the gaze focus area.

[0014] Further, calculating the region overlap rate between the occluded region and the viewing focus region includes: determining the number of overlapping pixels in the overlapping portion between the occluded region and the viewing focus region, and calculating the region overlap rate between the occluded region and the viewing focus region based on the number of overlapping pixels and the number of pixels in the viewing focus region.

[0015] In this scheme, the coverage of key visual areas by occlusions is directly quantified by statistically analyzing the ratio of overlapping pixels to the total number of pixels in the focal region. Compared to traditional coarse estimations based on bounding boxes or region centers, this method more accurately reflects the actual intensity of visual interference.

[0016] Furthermore, the step of performing image compensation on the HUD display information in the overlapping area includes: acquiring the obscured external real-world image information, and preprocessing and enhancing the external real-world image information to obtain an information mask; and based on the information mask, compensating and reconstructing the HUD display information in the overlapping area.

[0017] In this solution, the information displayed in the HUD's focus area is compensated and reconstructed using the ADAS camera image, which can significantly improve the security of information interaction, the accuracy of environmental perception, and the robustness of the intelligent driving system. It can generate a high-precision information mask by combining the environmental data collected in real time by the ADAS (Advanced Driver Assistance System) camera with deep learning algorithms, and overlay key information such as navigation arrows and collision warnings with the real road conditions to dynamically compensate and reconstruct the HUD display content.

[0018] Furthermore, the step of performing lightweight compensation on the HUD display information in the overlapping area includes: extracting core abstract information from the HUD display information in the overlapping area, performing lightweight processing on the core abstract information, and compensating and reconstructing the HUD display information in the overlapping area based on the processed core abstract information.

[0019] In this solution, the core abstract information in the HUD display information of the occluded area is extracted. The core abstract information can be simplified and brightened, and then the HUD display information of the occluded area can be adapted based on the processed core abstract information.

[0020] Furthermore, the offsetting of the HUD display information that will be blocked by the wiper includes: determining the time when the wiper reaches the blocked area, and, based on the focal point of the line of sight, shifting the HUD display information that will be blocked by the wiper towards the focal point of the line of sight by a preset distance within a preset offset time.

[0021] This solution significantly improves driving safety, information continuity, and system adaptability by predicting wiper trajectories in real time and dynamically adjusting information positions.

[0022] Based on the same concept, a HUD display adjustment system is also proposed, including: The system includes an information acquisition module for collecting wiper angle data; a HUD module for projecting and displaying information; and a processing module for predicting the wiper's obstruction area in the HUD display area at the next preset time and determining the obstruction duration. If the obstruction duration exceeds a first time threshold, the system determines the focal point area in the HUD display area and calculates the overlap rate between the obstructed area and the focal point area. If the overlap rate is greater than a preset overlap rate threshold, the system performs image compensation on the HUD display information in the overlapping area; otherwise, it performs minor compensation on the HUD display information in the overlapping area. If the obstruction duration is less than the first time threshold, the system shifts the HUD display information obstructed by the wiper.

[0023] Compared with the prior art, the beneficial effects of this application are as follows: This solution predicts the area and duration of wiper obstruction and proactively compensates for or shifts the displayed information before obstruction occurs, achieving a smooth transition and ensuring that critical driving data, such as navigation or alerts in the focus area, remains visible. Combining real-time prediction with proactive adjustments, it compensates for the most critical areas during prolonged obstruction by analyzing overlap rates based on the viewpoint. For short-term obstruction, an offset strategy is employed, adapting to different scenarios to avoid over-processing and ensuring that resources, such as processing power, are used only where necessary. Information offset is used to smoothly move information during brief obstructions, avoiding the frustration of flickering or disappearing information and maintaining a natural and smooth interface. Focus area compensation prioritizes areas of user interest when overlap rates are high, improving comfort. This method, through threshold judgment and dynamic compensation / offset, achieves safe, efficient, and user-friendly information management, improving driving safety, optimizing resource utilization, and enhancing the user experience. Attached Figure Description

[0024] Figure 1 This is a flowchart of the HUD display adjustment method described in this application.

[0025] Figure 2 This is a flowchart illustrating the calculation of the predicted position angle of the wiper at the next preset time, as described in this application.

[0026] Figure 3 This is a flowchart illustrating the data calibration and fusion of angle data as described in this application.

[0027] Figure 4 This is a schematic diagram of the mapping table of wiper angle, reference point coordinates, and occlusion size (pixels) described in this application.

[0028] Figure 5 This is a simplified structural diagram of the HUD display adjustment system described in this application. Detailed Implementation The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application. Example 1

[0029] like Figure 1As shown, this embodiment provides a HUD display adjustment method, including: predicting the area obstructed by the wiper in the HUD display area at the next preset time, and determining the duration of obstruction; if the duration of obstruction exceeds a first time threshold, determining the focal point area of ​​the line of sight in the HUD display area, and calculating the area overlap rate between the obstructed area and the focal point area; if the area overlap rate is greater than a preset area overlap rate threshold, performing image compensation on the HUD display information in the overlapping area; otherwise, performing minor compensation on the HUD display information in the overlapping area; if the duration of obstruction is less than the first time threshold, shifting the HUD display information obstructed by the wiper.

[0030] It should be noted that image compensation or minor compensation for the HUD display information in the overlapping area, as well as offsetting the HUD display information that will be obscured by the wiper, can all be performed before the wiper enters the obscured area or during the process of the wiper entering the obscured area. In this embodiment, the operation is performed before the wiper enters the obscured area. Specifically, after determining the obscured area, the time of entry into the obscured area can be estimated, and the operation can be performed before this time.

[0031] It should be noted that predicting the area obstructed by the wiper in the HUD display area at the next preset time includes: calculating the predicted position angle of the wiper at the next preset time, and determining the obstruction area based on the predicted position angle. In this embodiment, by calculating the predicted wiper position angle to determine the obstruction area, the future obstruction area of ​​the wiper on the windshield is predicted, ensuring that critical driving information is not interrupted, which can significantly improve driving safety, information display continuity, and system response efficiency.

[0032] Specifically, the calculation of the predicted position angle of the wiper at the next preset time includes the following steps S10 to S30.

[0033] Step S10: Obtain the wiper angle data, which includes the wiper position angle, angular velocity, and angular acceleration at the current moment.

[0034] Step S20: Perform data calibration and fusion on the angle data to obtain calibrated angle data.

[0035] Step S30: Calculate the predicted position angle of the wiper at the next preset time based on the calibrated angle data.

[0036] It should be noted that in step S10 above, the wiper position angle can be acquired by a Hall sensor. Specifically, the Hall sensor detects the change in the rotating magnetic field of the wiper motor and converts it into an angle signal. The Hall sensor can output the wiper position angle (θ) once every 10ms. t ); angular velocity of the wiper (v) tThe angular acceleration of the windshield wiper can be calculated by dividing the angle difference between two consecutive frames collected by the Hall sensor by the time interval, with the unit being ° / s; the angular acceleration of the wiper can be obtained by dividing the angle difference between two consecutive frames collected by the Hall sensor by the time interval. t The data is calculated, that is, from v t -v t-1 Dividing by Δt, the result is expressed in ° / s² and is used to describe the rate of change of the wiper angular velocity. Here, Δt represents the prediction time step, set to 0.1 s. This step size matches the Hall sensor's sampling frequency of 10 ms / sample, enabling continuous segmented prediction.

[0037] In step S20 above, the occlusion mask of the windshield wiper on the windshield can be determined using visual technology, thereby outlining the wiper's contour and determining its visual angle. Based on this visual angle, the wiper angle data acquired by the sensor is fused and calibrated to improve data accuracy. The wiper angle data fusion calibration calculation integrates raw sensor data and combines it with a dynamic calibration algorithm to eliminate errors such as mechanical clearance and temperature drift, ultimately outputting a high-precision real-time wiper angle. Its core lies in overcoming the performance limitations of a single sensor through multi-source data cross-validation and compensation. The wiper angle data calibration fusion calculation, through multi-source data collaborative optimization, can significantly improve the accuracy and robustness of the predicted position angle, providing reliable data support for HUD occlusion compensation. Specifically, it includes the following steps S201 to S203.

[0038] Step S201: Dynamically generate the shading mask for the windshield wipers.

[0039] Step S202: Determine the visual angle of the windshield wiper by blocking the mask.

[0040] Step S203: Based on the visual angle, perform data calibration and fusion on the angle data to obtain calibrated angle data.

[0041] In step S201, the occlusion mask is a binarized image (pixel-level precision) generated by the YOLOv8 visual recognition algorithm based on the windshield image captured by the forward-facing camera, with the same resolution as the camera's captured image. In this embodiment, the camera's captured image resolution can be set to 1280×720. This image contains only two types of information: occluded pixels and unoccluded pixels. Pixels corresponding to the area where the wiper is projected onto the windshield are marked as 1 (representing that this area will obscure the HUD display content), while pixels in other areas are marked as 0. For example, when the wiper is positioned in the upper middle part of the windshield, the pixel matrix corresponding to the upper middle part in the mask will show a continuous distribution of 1 values, clearly outlining the wiper's occlusion contour.

[0042] It should be noted that in this embodiment, the camera frame rate is set to 30fps, corresponding to outputting one frame of image every 33ms, and correspondingly, one frame of occlusion mask every 33ms. This frequency is highly matched with the motion characteristics of the wiper and the system processing capability, and has the following core functions: Real-time performance assurance: The sway speed of the wiper is typically 0.5-3m / s, and the 33ms frame interval can completely capture every tiny displacement of the wiper, avoiding missed detection of occluded areas due to excessively long frame intervals; Data calibration benchmark: Although the Hall sensor has a higher sampling frequency (10ms / time), it is susceptible to angle errors caused by electromagnetic interference from the motor. The occlusion mask can serve as a visual benchmark to correct the sensor data; Fusion basis: Each frame of mask can correspond one-to-one with the wiper angle data collected by the Hall sensor at the same time, providing paired samples for Kalman filtering data fusion and improving the accuracy of subsequent trajectory prediction.

[0043] In step S202, the visual angle can be determined by searching and matching using the previously calibrated database. Specifically, an angle-mask mapping database can be constructed, which is completed during vehicle manufacturing or system initialization, with the aim of creating a complete and high-precision reference dataset. Features are extracted from the occlusion mask determined in step S201, and similarity is compared in the angle-mask mapping database to determine the visual angle corresponding to the occlusion mask.

[0044] In step S203, a Kalman filter can be used to calibrate and fuse Hall sensor data based on visual angles. High-frequency but potentially noisy Hall sensor data (angle data) is fused with low-frequency but high-precision visual angles to output calibrated and optimized position angle, angular velocity, and angular acceleration. Specifically, the wiper angle, angular velocity, and angular acceleration are constructed as an interrelated state vector. High-frequency but potentially inaccurate Hall sensor data is responsible for tracking the wiper's movement trend in real time, while the low-frequency but high-precision visual angle serves as an absolute reference. Whenever the visual angle is calculated, the filter calculates its residual with the current predicted state and, based on a preset confidence level (i.e., low noise in visual data and high noise in Hall data), intelligently allocates the angle residual to the angular velocity and angular acceleration states through the Kalman gain matrix, thereby indirectly correcting them. Finally, a deeply fused and optimized state vector is output, which retains the high-frequency response characteristics of the Hall sensor while possessing the accuracy of visual data, thus obtaining high-precision position angle, angular velocity, and angular acceleration data.

[0045] In step S30, the predicted position angle of the wiper at the next preset time is accurately predicted by a second-order kinematic model. The preset time is set to 0.2-0.5s. In this embodiment, the preset time is 0.3s. In other possible embodiments, the preset time may also be other values.

[0046] The movement of a windshield wiper is not uniform; it accelerates when starting from its initial position and decelerates before reaching its final position. A first-order kinematic model that only considers velocity cannot accurately describe this process. However, a second-order kinematic model, which introduces acceleration parameters, can completely fit the start-stop and speed-changing trajectories of the wiper, and its prediction accuracy is significantly better than that of the first-order model. In this embodiment, 0.3s is chosen as the preset time, which is the optimal value derived by considering system latency, driving safety requirements, and prediction errors. Firstly, it matches the system processing latency. The preprocessing and rendering of the HUD image take approximately 150-200ms in total. A prediction duration of 0.3s allows sufficient time for compensation preparation, ensuring that the compensation image is displayed synchronously when the wiper reaches the obstruction position. Secondly, it conforms to the wiper movement cycle. The single oscillation cycle of a typical car wiper is approximately 0.8-1.2s. 0.3s is only 1 / 4-1 / 3 of the cycle, during which the wiper's movement (acceleration / deceleration) is relatively stable, and the prediction error can be controlled within 20mm. If the prediction duration exceeds 0.5s, the wiper may enter a reverse oscillation phase, and sudden changes in motion parameters will lead to a significant increase in error. Thirdly, it adapts to the driver's visual requirements. There is approximately a 0.2s visual delay in the driver's visual capture of HUD information. A prediction duration of 0.3s ensures that the compensation information is synchronized with the driver's visual perception, avoiding information lag affecting judgment.

[0047] The other second-order kinematic model is as follows: θ t+1 =θ t +v t •Δt+0.5•a t •Δt². Wherein, θ t The wiper position angle at the current time t is the angle between the wiper and its initial position (usually the lowest point where the wiper stops), as collected by the Hall sensor at time t. The unit is degrees (°), with an accuracy of ±0.5°. θ t+1 : The predicted wiper position angle at the next moment t+1, where the time interval between t+1 and t is Δt (0.1s). v t The angular velocity of the wiper at the current time t is calculated by dividing the angle difference between two consecutive frames collected by the Hall sensor by the time interval, and the unit is ° / s; Δt: Prediction time step, set to 0.1s. This step size matches the sampling frequency of the Hall sensor at 10ms / time, enabling continuous segmented prediction. a t The wiper angular acceleration at the current time t is calculated from three consecutive sets of θt data, i.e., obtained by dividing vt-vt-1 by Δt, with units of ° / s², and is used to describe the rate of change of the wiper angular velocity. v t• Δt: The increment of the angle at which the wiper moves only at the current angular velocity vt within the time interval Δt; 0.5・a t • Δt²: The additional angular increment of the wiper blade during the time interval Δt due to angular acceleration at.

[0048] To facilitate understanding, the following calculation example is used. Under a certain operating condition, the initial state data of the windshield wipers (at t=0) are as follows: θ0=10°, v0=30° / s, a0=50° / s². With a step size of Δt=0.1s, the wiper angle in the next 0.3s is predicted piecewise: (1) Predict the angle at t=0.1s: θ 0.1 =10+30×0.1+0.5×50×(0.1)²=10+3+0.25=13.25°; (2) Calculate the angular velocity at t=0.1s: v 0.1 =v0 + a0 × Δt = 30 + 50 × 0.1 = 35° / s, assuming the acceleration remains constant (a 0.1 =50° / s²); (3) Predict the angle at t=0.2s: θ 0.2 =13.25+35×0.1+0.5×50×(0.1)²=13.25+3.5+0.25=17°; (4) Calculate the angular velocity at t=0.2s: v 0.2 =35 + 50 × 0.1 = 40° / s; (5) Predict the angle at t=0.3s: θ 0.3 =17+40×0.1+0.5×50×(0.1)²=17+4+0.25=21.25°.

[0049] The final result showed that the wiper will gradually swing from 10° to 21.25° within 0.3 seconds. The error between this result and the actual wiper movement trajectory was tested to be ≤3°, which translates to a displacement error of ≤15mm on the windshield, meeting the system accuracy requirements.

[0050] Furthermore, after determining the predicted position angle, the occlusion area is determined based on the predicted position angle. This is a spatial mapping process, which is the transformation from the wiper position angle to the HUD pixel coordinates. The core of spatial mapping is to establish a one-to-one correspondence between the mechanical swing angle of the wiper and the pixel coordinate system of the HUD display area. The transformed HUD coordinate range is the predicted occlusion range of the future wiper. The specific real-time conversion calculation steps are as follows: Step 1, Angle-Physical Position Conversion Based on the wiper mounting wheelbase L, the predicted wiper position angle θ t+n Convert to the physical projection coordinates of the windshield wipers on the windshield (X) phy ,Y phy The calculation formula is X. phy =L×cosθ t+n Y phy =L×sinθ t+n , where (X) phy ,Y phy The origin is the projection point of the wiper motor shaft on the windshield.

[0051] Step 2, Physical Location-HUD Projection Conversion Using pre-calibrated HUD field of view parameters, the physical coordinates (X, Y, F) on the windshield are displayed. phy ,Y phy This is converted into physical coordinates of the HUD virtual display plane. For example, a 10° HUD field of view corresponds to a 20cm horizontal area on the windshield; the physical position is proportionally converted into a relative position (X) within the HUD plane. HUD ,Y HUD ).

[0052] Step 3, HUD plane physical coordinates to pixel coordinates conversion Based on the HUD resolution (1920×720), the physical coordinates of the HUD plane are converted to pixel coordinates. Assuming a horizontal physical width of 15cm corresponds to 1920 pixels, the pixel conversion factor kx = 1920 / 15 = 128 pixels / cm. Similarly, the vertical factor ky is derived, resulting in the final pixel coordinates (X...). pix ,Y pix ) = (X HUD ×kx,Y HUD ×ky).

[0053] It should be noted that the feature points (such as the center point / top left corner point) of the occluded area corresponding to the wiper swing angle are pixel coordinates (X). pix ,Y pix Only by combining the physical dimensions and motion trajectory of the wiper with the pre-calibrated mapping relationship (calibration library) can the complete occlusion rectangle area be derived.

[0054] Here, the single pixel coordinates serve as the reference point, and in the spatial mapping, the pixel coordinates (X) are calculated using the wiper angle. pix ,Y pix (e.g., (260, 150)) is a feature reference point of the obstructed area (e.g., the top left corner vertex, center point, or geometric center of the wiper obstruction area). The function of this reference point is to serve as the positioning anchor point of the obstructed area, determining the approximate position of the area in the HUD coordinate system; then, combined with the physical properties of the wiper (length, width) and the pre-calibrated size ratio, it is expanded into a complete rectangular area.

[0055] The complete occlusion rectangle was derived using the following method: During the initial calibration phase, when recording the angle-area correspondence, not only were the coordinates of the reference point corresponding to each wiper angle recorded, but also the width and height pixel values ​​of the wiper obstructing the HUD at that angle were recorded simultaneously (e.g., when the wiper angle is 10°, the obstructed area is 100px wide and 200px high). These dimensional data were calculated based on the actual physical dimensions of the wiper (e.g., 60cm long and 10cm wide) and the HUD projection ratio.

[0056] During real-time calculation, the coordinates of the reference point are first calculated using the angle (e.g., (260, 150)). Then, the width (100px) and height (200px) corresponding to that angle are retrieved from the calibration library to finally determine the complete rectangular area: the coordinates of the top left corner = (reference point x, reference point y), and the coordinates of the bottom right corner = (reference point x + width, reference point y + height). For example, when the width of the occluded area is 100px and the height is 200px, the occluded rectangular area is (260, 150) - (360, 350). The determination of the rectangular area depends on the dimensional data from the calibration stage and the physical properties of the wiper; it is a combination of the reference point and fixed / dynamic dimensions.

[0057] The derivation process from the reference point to the rectangular region is essentially the engineering implementation of pre-stored mapping relationships in the calibration phase and rule-based expansion in the real-time phase. It relies entirely on the physical properties of the wiper, the HUD projection characteristics, and the pre-calibration data to ensure the accuracy and real-time performance of the derivation results. The following is a detailed derivation process broken down into stages, including mathematical calculations, calibration logic, and engineering details.

[0058] The core of the calibration phase is data storage. Through multi-point calibration, a mapping table is established between the wiper angle, the reference point coordinates, and the occlusion size (pixels). The specific operations are as follows: For each of the multiple key angles (e.g., 0°~90°) within the full swing range of the windshield wipers (e.g., 0°~90°), record two key pieces of information: the coordinates of the reference point (P). x ,P y ) and occlusion size (W _px H _px Reference point coordinates (P) x ,P y The top left corner is the reference point for the wiper occlusion area; the occlusion size (W) is the reference point for the wiper occlusion area. _px H _px The value represents the pixel width and pixel height of the HUD area obstructed by the wiper at that angle, calculated from the physical size of the wiper.

[0059] To obtain the physical dimensions of the windshield wiper, the actual physical parameters of the wiper are fixed, including wiper length, wiping width, and wiping height. For example: wiper length L = 60cm, wiping width W... _phy =10cm, wiper height H _phy =20cm; where, W _phy H _phy It's not the full size of the wiper blade, but the physical size of the overlapping portion between the wiper blade and the HUD projection area.

[0060] Calculate the pixel conversion factor and, considering the HUD projection characteristics (e.g., a horizontal physical width of 15cm corresponds to 1920 pixels), obtain the following: Horizontal conversion factor kx = HUD horizontal pixel count / HUD horizontal physical width; Vertical conversion factor ky = HUD vertical pixel count / HUD vertical physical width. For example, if a HUD horizontal physical width of 15cm corresponds to 1920 pixels, kx = 1920px / 15cm = 128px / cm; if a HUD horizontal physical width of 6cm corresponds to 720 pixels, ky = 720px / 6cm = 120px / cm.

[0061] Physical dimensions are converted to pixel dimensions. The pixel size obscured by the windshield wiper is calculated as: physical size × conversion factor. W _px =W _phy ×kx;H _px =H _phy ×ky.

[0062] To correct projection deviations, since the windshield is curved, the wiper blades' projections will be slightly distorted at different angles. After correction using a "perspective transformation matrix," the corrected dimensions (W) at each calibration angle are finally determined. _cal H _cal ), ensuring that the mapping error is ≤2px.

[0063] like Figure 4 The table shown is an example of a calibration mapping table for wiper angle, reference point coordinates, and occlusion size (pixels).

[0064] During real-time calculations, the wiper position angle at the next preset time is calculated, and combined with the mapping table of wiper swing angle—reference point coordinates—coverage size (pixels), the occlusion rectangular area can be derived. Taking a calculated wiper swing angle θ = 21.25° as an example, the derivation of the occlusion rectangular area is explained below: Step 1: Query the calibration mapping table (interpolate and supplement non-calibrated angles).

[0065] Since there is no direct data for 21.25° in the calibration mapping table, the reference point and dimensions corresponding to this angle need to be calculated through linear interpolation, as follows: Consult the calibration mapping table to find the two adjacent calibration angles of 21.25°: 20° (Px1 =250,P y1 =150; W1=100px, H1=200px) and 30° (P x2 =300,P y2 =140; W2=102px, H2=200px).

[0066] Calculate the interpolation weights: α = (21.25° - 20°) / (30° - 20°) = 0.125.

[0067] Interpolation calculation of the reference point and dimensions for 21.25°: P x =P x1 +α×(P x2 -P x1 =250 + 0.125 × (300 - 250) = 256.25px (rounded to 256px); P y =P y1 +α×(P y2 -P y1 =150+0.125×(140-150)=148.75px (rounded to 149px). W _px =W1+α×(W2-W1)=100+0.125×(102-100)=100.25px (rounded to 100px). H _px =H1+α×(H2-H1)=200+0.125×(200-200)=200px; The final result is the reference point (256, 149) and the occlusion size (100px, 200px) corresponding to 21.25°.

[0068] Step 2: Expand the rectangle into four vertices according to the reference point rule.

[0069] Since the reference point is the top-left vertex, the coordinates of the four vertices of the rectangle can be directly derived from the reference point and the dimensions, as follows: Top left vertex (X1, Y1) = reference point (P) x ,P y = (256, 149); Top right vertex (X2, Y2) = (P x +W _px ,P y ) = (256 + 100, 149) = (356, 149); The lower left vertex (X3, Y3) = (P x ,P y+H _px ) = (256, 149 + 200) = (256, 349); The bottom right vertex (X4, Y4) = (P x +W _px ,P y +H _px ) = (256 + 100, 149 + 200) = (356, 349); Step 3: Edge cropping (to avoid exceeding the HUD display area) The display area of ​​a HUD is fixed, such as 1920×720 pixels, with coordinates ranging from 0 to 1919 and 0 to 719. The derived rectangle needs to undergo boundary checks and cropping to prevent it from exceeding the display area and causing rendering errors. Specifically: If X1 < 0, cut to X1 = 0; If X2 > 1919, then cut to X2 = 1919; If Y1 < 0, cut to Y1 = 0; If Y2 > 719, then cut to Y2 = 719.

[0070] In this example, (256,149)-(356,349) is completely within the HUD range and does not require clipping. The final output occlusion rectangle area is (256,149)-(356,349).

[0071] It should be noted that the occlusion duration is the time from when the wiper edge enters the occlusion area to when it leaves the occlusion area. Knowing the wiper's angular velocity and acceleration, the occlusion duration can be calculated. After determining the occlusion duration, it is checked whether it exceeds a first time threshold. If it exceeds the first time threshold, the focal point area of ​​the view in the HUD display area needs to be determined, and the overlap rate between the occlusion area and the focal point area needs to be calculated. In this embodiment, the first time threshold is set to 0.2s.

[0072] Specifically, determining the focal point area of ​​the gaze within the HUD display area includes: determining the focal coordinates of the gaze within the HUD display area, and expanding the area around these focal coordinates to obtain the focal point area. In this embodiment, motion tracking technology can be used to capture the driver's gaze focus in real time, perform gaze focus mapping, convert the driver's eye movements into precise coordinates within the HUD display area, and expand this into a dynamic area, which is the focal point area. This provides a basis for calculating the overlap rate between the occluded area and the focal point.

[0073] When the driver uses it for the first time, they establish an eye movement-display coordinate mapping relationship by focusing on a 9-point calibration pattern on the HUD (pausing at each point for 2 seconds). In subsequent actual use, the pupil center (x) is detected by an infrared camera. _eye,y _eye The HUD display area's gaze focus coordinates (x, y) are calculated using a pre-trained neural network model. _hud ,y _hud The calculation formula is as follows: (x _hud ,y _hud )=(w x ×(x _eye -x0) / w eye_x +x0,w y ×(y _eye -y0) / w eye_y +y0); (x _hud ,y _hud ) represents the target output parameter, which is the coordinate of the gaze focus in the pixel coordinate system of the HUD display area. The origin of the coordinate system (0,0) is the upper left corner of the HUD display area. (x) _eye ,y _eye The input parameter is the pixel coordinates of the pupil center detected by the infrared camera in the camera image coordinate system, with the origin of the coordinate system being the upper left corner of the camera's captured image. w x w y The pixel width and height of the HUD display area are 1920 pixels and 720 pixels, respectively, in this embodiment. w eye_x w eye_y The pixel width and height of the image captured by the infrared camera are 640 pixels and 480 pixels, respectively, in this embodiment, corresponding to the camera module. (x0, y0) are calibration parameters, which are the origin coordinates of the HUD display area in the driver's field of vision, determined by 9-point calibration. For example, after calibration, x0 = 50 pixels and y0 = 30 pixels, representing the reference position of the upper left corner of the HUD in the driver's field of vision.

[0074] The core of this formula is to achieve the scaling and datum calibration of two coordinate systems, which consists of two steps: Relative offset calculation, (x _eye -x0) / w eye_x and (y) _eye -y0) / w eye_y Calculate the lateral and longitudinal relative offset ratios of the pupil center relative to the calibration origin to eliminate the influence of the deviation between the camera and the HUD installation position. HUD coordinate scaling, via w x and w yThe offset ratio is scaled to convert the relative position in the camera coordinate system into the actual pixel coordinates of the HUD display area. Finally, the calibration origin (x0, y0) is superimposed to ensure the accuracy of the coordinate mapping.

[0075] For example, w is known x =1920, w y =720, w eye_x =640, w eye_y =480, x0=50, y0=30, the infrared camera detected x _eye =320, y _eye =240, then x _hud =1920×(320-50) / 640+50=1920×0.421875+50=810, y _hud =720×(240-30) / 480+30=720×0.4375+30=345, that is, the coordinates of the focus of the line of sight on the HUD are (810,345).

[0076] In this embodiment, considering the stringent requirements for computing power and real-time performance in automotive scenarios, a lightweight MobileNetV3 neural network model with 3 fully connected layers is adopted for calculating the gaze focus coordinates. The specific design is as follows: The neural network model's structure includes a feature extraction layer, a regression layer, and a post-processing layer. The feature extraction layer employs the MobileNetV3-Small architecture, replacing traditional convolutions with depthwise separable convolutions, reducing computational cost by 70% while retaining the ability to extract eye features. The input is a 640×480 pixel eye image captured by an infrared camera, and the output is a 128-dimensional feature vector. The regression layer consists of three fully connected layers, successively compressing the feature vector dimension from 128 to 64 and then to 32, ultimately outputting a 2-dimensional vector, representing the initial coordinates of the gaze focus in the infrared camera image coordinate system. The post-processing layer adds a BatchNorm layer and a Dropout layer (dropout rate 0.2) to prevent overfitting and improve prediction stability under complex lighting conditions.

[0077] The training process of the neural network model includes dataset construction, training parameters, and model deployment. Dataset construction involves collecting eye data from 200 drivers of different genders and ages, covering eight driving conditions including normal vision, strabismus, low light, and strong light. 1000 frames are collected for each condition, containing eye images and corresponding HUD focus coordinate labels (obtained through 9-point calibration). Training parameters utilize the Adam optimizer with an initial learning rate of 0.001, decreasing by 10% every 10 epochs. The loss function is mean squared error (MSE), aiming to minimize the deviation between predicted coordinates and true labels. Model deployment involves quantizing the model to INT8 precision using TensorRT and deploying it to an NXP.MX8 QuadMax processor. The single-frame processing time is ≤10ms, meeting real-time requirements.

[0078] Compared to traditional geometric localization algorithms, this model can automatically learn multi-dimensional features such as pupil shape, iris texture, and eyelid occlusion. It is more adaptable to scenarios such as drivers wearing ordinary glasses or squinting slightly, and its coordinate prediction error is ≤3px, which is better than the 5px error of OpenCV geometric algorithms.

[0079] Furthermore, although the gaze focus coordinates are a single pixel, the driver's visual attention is not limited to that point, but rather to a certain area centered on that point. The gaze focus area is determined by employing a focus area expansion method. Specifically, based on human visual characteristics, the gaze focus coordinates (x, y, z) within the HUD display area are... _hud ,y _hud The focal area is expanded to a 30×30 pixel square (this size has been ergonomically tested and conforms to the normal visual focusing range of a driver). The coordinate range of the expanded area is (x... _hud -15,y _hud -15) to (x _hud +15,y _hud +15), if it exceeds the HUD display boundary, it will be cropped.

[0080] Furthermore, calculating the regional overlap rate between the occluded region and the focal point region includes: determining the number of overlapping pixels in the overlapping portion between the occluded region and the focal point region; and calculating the regional overlap rate between the occluded region and the focal point region based on the number of overlapping pixels and the number of pixels in the focal point region. In this embodiment, the coverage degree of the occluder on the key visual region is directly quantified by statistically analyzing the ratio of the number of overlapping pixels to the total number of pixels in the focal point region. Compared to traditional coarse estimations based on bounding boxes or region centers, this method can more accurately reflect the actual intensity of visual interference.

[0081] Furthermore, if the duration of occlusion exceeds a first time threshold and the regional overlap rate is greater than a preset regional overlap rate threshold, before the wiper enters the occluded area, image compensation is performed on the HUD display information in the overlapping area. This includes: acquiring the occluded external real-world image information and preprocessing and enhancing the external real-world image information to obtain an information mask; and compensating and reconstructing the HUD display information in the focal area using the information mask. In this embodiment, the preset regional overlap rate threshold is set to 30%. In other possible embodiments, the regional overlap rate threshold can be set to other values.

[0082] This scenario typically occurs in moderate rain, with the wipers oscillating at a moderate speed (0.8-1.5 m / s), and the obstructed area precisely covering the core HUD information the driver is focused on (such as navigation guidance and lane departure warning). For example, when the driver is looking at the left-turn navigation arrow on the HUD, the wipers will be oscillating over this area, resulting in an overlap of up to 45%.

[0083] In this embodiment, the HUD display information in the overlapping area is reconstructed and compensated using the ADAS camera image, solving the problem of the driver not being able to see the real-world road conditions when the wipers are sweeping across the outer surface of the windshield. The ADAS camera image is not the virtual information such as navigation and vehicle speed displayed by the HUD itself, but refers to the real-time road video stream collected by one or more cameras in front of the vehicle for use by advanced driver assistance systems. ADAS cameras are usually the main camera module deployed on the inside of the vehicle's windshield. When the wipers obscure the core information area of ​​the HUD that the driver is focusing on, the system needs to do more than just adjust the HUD display content; it needs to intelligently reconstruct the portion of the real-world scene obscured by the wipers on the actual windshield and compensate it into the HUD display.

[0084] Specifically, when compensating for ADAS camera images, firstly, Zhang Zhengyou's calibration method is used to correct ADAS image distortion. A lightweight U-Net model is then used to perform semantic segmentation on the real road image captured by the ADAS camera, accurately extracting key road condition information that is obscured (such as lane lines and traffic signs). Simultaneously, redundant background elements such as the sky and vegetation are removed, generating an information mask that only contains the outlines of key information or transparent channels. For example, only white lane lines and left-turn arrows are visible, while the rest are transparent. Then, the extracted information mask is blended and rendered with the native HUD virtual interface, accurately compensating for overlapping areas. Through adaptive brightness adjustment and edge feathering, seamless integration of the compensated image with the original display is ensured. Finally, the image is output to the HUD with an ultra-low latency of ≤50ms, creating a clear and uninterrupted digital view for the driver even in obscured conditions.

[0085] Additionally, the compensation priority for critical information in the HUD display, such as lane lines, vehicles ahead, and traffic signs, is set to the highest level, taking precedence over non-critical information like the normal navigation background. When the information mask is integrated with the native HUD virtual interface for rendering, this critical information is preserved and cannot be covered by the information mask, while non-critical information can be covered by the information mask.

[0086] It's important to note that during the fusion rendering process, the ADAS image isn't simply pasted onto the physical location on the windshield obscured by the wipers like a sticker. Instead, compensation information, as part of the HUD virtual image, is projected onto the overlapping area. That is, within the driver's focus area, the portion of the real road condition obscured by the wipers is reconstructed using the ADAS camera image and overlaid on the corresponding position on the HUD virtual interface. This allows the driver to see both the HUD navigation and the complete real road conditions.

[0087] Furthermore, if the duration of occlusion exceeds a first time threshold and the area overlap rate is less than or equal to a preset area overlap rate threshold, before the wiper enters the occluded area, the HUD display information in the overlapping area is compensated, including: extracting core abstract information from the HUD display information in the overlapping area, performing lightweight processing on the core abstract information, and compensating and reconstructing the HUD display information in the overlapping area based on the processed core abstract information.

[0088] This scenario typically occurs in light rain, with slow wiper speeds (<0.8m / s), obstructed areas at the edge of the HUD (such as the fuel consumption display area), and the driver's focus on the vehicle speed display area in the center of the HUD. For example, the overlap between the obstructed area and the focus area is only 15%.

[0089] In this scenario, the windshield wipers obstruct the HUD display in the overlapping area, making it blurry and difficult for the driver to see clearly. Lightweight compensation is applied to the HUD display in the overlapping area. The goal of this compensation is not to restore the true field of vision, but rather to maintain the visibility of the HUD user interface. The compensation data originates from the HUD's own virtual UI data, aiming to simplify complex UI icons into a low-interference outline. Specifically, information is filtered, retaining only the core abstract information from the HUD display within the obstructed area and removing complex image content. For example, when the navigation area is obstructed, only the white outline of the navigation arrow is retained, removing the arrow fill color and background texture; when the speed-related area is obstructed, only the numerical outline is retained, simplifying the display style. Given the predicted obstruction area (obstruction mask) from the wipers, compensation content (core abstract information) is generated directly based on the outline of the mask. After adjusting the brightness according to ambient light, no complex color adjustments are needed; only the brightness of the compensation content is kept consistent with the original information to ensure it does not interfere with the driver's focus on the focal area.

[0090] It should be noted that when performing lightweight compensation on the HUD display information in the overlapping area, the geometric boundary of the occlusion mask is directly reused without additional calculation of the shape of the compensation area, and the core abstract information after compensation is integrated into the overlapping area.

[0091] Furthermore, if the duration of the obstruction is less than a first time threshold, before the wiper enters the obstructed area, the HUD display information obscured by the wiper is offset. This includes: determining the time when the wiper reaches the obstructed area, and, based on the line of sight focus, offsetting the HUD display information obscured by the wiper towards the line of sight focus by a preset distance within a preset offset time. The preset offset time and preset distance are set according to actual conditions. This scenario applies to heavy rain, where the wipers swing at high speed (>1.5m / s). The duration of each instance of obscuring a specific area of ​​the HUD is extremely short (e.g., 0.1s). If compensation is only made after the obstruction occurs, it can easily cause information flickering. For example, if the wipers swing across the HUD navigation area at high speed, the obstruction lasts for 0.15s.

[0092] In this scenario, continuous visibility of information is ensured by predicting the wiper trajectory in real time and dynamically adjusting the information position. Specifically, based on a second-order kinematic model, the precise time T for the wiper to reach the obstructed area is predicted, and information offset is initiated 50ms in advance. Using the coordinates of the viewpoint as a reference, the obstructed HUD information is offset 50-80 pixels towards the viewpoint. The offset distance is dynamically adjusted according to the wiper obstruction range to ensure that the offset information is not on the wiper's trajectory. The information offset process uses linear movement with a movement time of 20ms to avoid visual shock caused by instantaneous information movement. When the wiper leaves the obstructed area, the offset information is reset to its original position in the same linear manner, with the reset time synchronized with the wiper's departure time to ensure the continuity of information display.

[0093] Additionally, when a sensor or algorithm malfunction is detected, a redundancy mechanism is triggered and logged. This redundancy mechanism includes resetting redundant sensors / watchdog timers, downgrading the display of basic information, and issuing alarms. Specifically, when the main camera malfunctions, the redundant sensor automatically switches to the in-vehicle monitoring camera (640×480 resolution) for auxiliary compensation. The watchdog timer (e.g., MAX16997) checks the main control chip status every 200ms; if no heartbeat signal is received within the timeout period, a forced reset is initiated and the system switches to basic mode (displaying only vehicle speed and RPM). Specifically, when eye-tracking fails, the HUD display position is automatically adjusted based on the wiper position and vehicle speed (e.g., moving the vehicle speed indicator to the top edge of the windshield). When the ADAS camera image becomes blurry (e.g., during heavy rain), image compensation stops, and the driver is alerted via a buzzer and a red warning box.

[0094] In this embodiment, by predicting the wiper obstruction area and duration, and proactively compensating or shifting the displayed information before obstruction occurs, a smooth transition is achieved, ensuring that critical driving data, such as navigation or alerts in the focus area, remains visible. Combining real-time prediction, this approach is more proactive than passive adjustment. For prolonged obstruction, the overlap rate is analyzed based on the focus area to accurately compensate the most critical area; for short-term obstruction, an offset strategy is employed, adapting to different scenarios to avoid over-processing and ensuring that resources, such as processing power, are only used where necessary. Information offset is used for brief obstructions to smoothly move information, avoiding the frustration of information flickering or disappearing, and maintaining a natural and smooth interface. Focus area compensation prioritizes areas of user interest when the overlap rate is high, improving comfort. This method, through threshold judgment and dynamic compensation / offset, achieves safe, efficient, and user-friendly information management, improving driving safety, optimizing resource utilization, and enhancing the user experience.

[0095] Example 2 This embodiment proposes a HUD display adjustment system, including: an information acquisition module 100, a HUD module 200, and a processing module 300. The modules interact with each other through a controller local area network (CAN bus), and the communication protocol follows the ISO11898 standard with a baud rate set to 500kbps.

[0096] The information acquisition module 100 includes a Hall sensor, a forward-facing camera, an infrared camera, an ADAS camera, and an environmental perception sensor.

[0097] A Hall effect sensor, used to collect wiper angle data, is mounted on the wiper motor shaft.

[0098] A forward-facing camera is used to capture images of the windshield.

[0099] An infrared camera, used to capture the driver's eye movements and collect images of the driver's eyes, is installed on the top of the dashboard; an ADAS camera, used to collect real-time images of the external environment in front of the vehicle.

[0100] Environmental perception sensors include a multispectral ambient light sensor (covering 380-1100nm) that detects ambient light intensity (0.1-100,000 lux) and color temperature (2700-6500K), with a response time ≤50ms. A rain sensor (installed inside the windshield) detects rainfall levels (0-5) through capacitance changes. A vehicle speed sensor (from the vehicle's ECU) acquires real-time vehicle speed (accuracy ±1km / h).

[0101] The HUD module 200 is used for projecting and displaying information, including a light source system, a mirror assembly, and an electronic control unit. It employs waveguide technology (10°×6° field of view), supports a resolution of 1920×720, and has a dynamic brightness range of 10-1000 nits.

[0102] The processing module 300 is used to predict the area obstructed by the wiper in the HUD display area at the next preset time and determine the duration of obstruction. If the obstruction duration exceeds a first time threshold, it determines the focal point area of ​​the line of sight in the HUD display area and calculates the overlap rate between the obstructed area and the focal point area. If the overlap rate is greater than a preset overlap rate threshold, it performs image compensation on the HUD display information in the overlapping area before the wiper enters the obstructed area; otherwise, it performs minor compensation on the HUD display information in the overlapping area before the wiper enters the obstructed area. If the obstruction duration is less than the first time threshold, it shifts the HUD display information obstructed by the wiper before the wiper enters the obstructed area. The processing module can use an NXPi.MX8QuadMax processor (supporting multi-sensor parallel processing) and run a real-time operating system based on QNX.

[0103] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of this application. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of this application. All such changes and modifications are intended to be included within the scope of this application as claimed in the appended claims.

[0104] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0105] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed.

[0106] The various component embodiments of this application can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some modules according to the embodiments of this application. This application can also be implemented as an apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such an implementation of this application can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0107] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0108] Although the description of this application has been made in conjunction with the specific embodiments described above, it will be apparent to those skilled in the art that many substitutions, modifications, and variations can be made based on the foregoing. Therefore, all such substitutions, modifications, and variations are included within the spirit and scope of the appended claims.

Claims

1. A method for adjusting a HUD display, characterized in that, The method includes: Predict the area of ​​the wiper that will be obstructed by the windshield wiper in the HUD display area at the next preset time, and determine the duration of the obstruction; If the duration of the occlusion exceeds the first time threshold, the focal point area of ​​the view in the HUD display area is determined, and the area overlap rate between the occlusion area and the focal point area is calculated. If the overlap rate of the regions is greater than a preset overlap rate threshold, the HUD display information in the overlapping regions is compensated; otherwise, the HUD display information in the overlapping regions is lightly compensated. If the duration of the obstruction is less than a first time threshold, the HUD display information that will be obscured by the wiper will be shifted.

2. The HUD display adjustment method according to claim 1, characterized in that, The predicted area where the wiper will block the light in the HUD display area at the next preset time includes: Calculate the predicted position angle of the wiper at the next preset time, and determine the occlusion area based on the predicted position angle.

3. The HUD display adjustment method according to claim 2, characterized in that, The calculation of the predicted position angle of the wiper at the next preset time includes: Acquire windshield wiper angle data, which includes: the current position angle, angular velocity, and angular acceleration of the windshield wiper. The angle data is calibrated and fused to obtain calibrated angle data; Based on the calibrated angle data, the predicted position angle of the wiper at the next preset time is calculated.

4. The HUD display adjustment method according to claim 3, characterized in that, The step of calibrating and fusing the angle data to obtain calibrated angle data includes: Dynamically generate wiper shielding masks; The visual angle of the windshield wiper is determined by the masking film. Based on the visual angle, the angle data is calibrated and fused to obtain calibrated angle data.

5. The HUD display adjustment method according to claim 1, characterized in that, The determination of the focal point area in the HUD display area includes: Within the HUD display area, determine the focal coordinates of the line of sight, and expand the area around the focal coordinates to obtain the focal area of ​​the line of sight.

6. The HUD display adjustment method according to claim 5, characterized in that, The calculation of the overlap rate between the occluded area and the focal point area includes: Determine the number of overlapping pixels in the overlapping portion between the occluded area and the focal point area, and calculate the regional overlap rate between the occluded area and the focal point area based on the number of overlapping pixels and the number of pixels in the focal point area.

7. The HUD display adjustment method according to claim 1, characterized in that, The process of compensating for the HUD display information in the overlapping area includes: Obtain the obscured external real-world scene information, and preprocess and enhance the external real-world scene information to obtain an information mask; Based on the information mask, the HUD display information in the overlapping area is compensated and reconstructed.

8. The HUD display adjustment method according to claim 1, characterized in that, The light compensation for the HUD display information in the overlapping area includes: Extract the core abstract information from the HUD display information of the overlapping area, perform lightweight processing on the core abstract information, and compensate and reconstruct the HUD display information of the overlapping area based on the processed core abstract information.

9. The HUD display adjustment method according to claim 1, characterized in that, The offset will affect the HUD display information that is obscured by the windshield wipers, including: Determine the time when the wiper reaches the blocked area. Based on the focal point of the line of sight, within a preset offset time, shift the HUD display information blocked by the wiper towards the focal point of the line of sight by a preset distance.

10. A HUD display adjustment system, characterized in that, include: The information acquisition module is used to collect wiper angle data; HUD module, used to project and display information; The processing module is used to predict the area of ​​the wiper obstruction in the HUD display area at the next preset time and determine the duration of obstruction. If the duration of the obstruction exceeds the first time threshold, the focal point area of ​​the view in the HUD display area is determined, and the overlap rate between the obstructed area and the focal point area is calculated; if the overlap rate is greater than a preset overlap rate threshold, the HUD display information in the overlapping area is compensated; otherwise, the HUD display information in the overlapping area is lightly compensated; if the duration of the obstruction is less than the first time threshold, the HUD display information obstructed by the wiper is shifted.

Images