HUD image display calibration method and device and storage medium

Abstract

The invention discloses an HUD image display calibration method and device and a storage medium, and relates to the technical field of HUD image calibration, and the method comprises the following steps: obtaining the eye movement ranges of different users in different use regions when driving a vehicle according to the use regions of the vehicle, and recording the eye movement ranges as initial eye movement range data; performing eye movement range preprocessing on the initial eye movement range data of the different use areas to obtain reference eye movement range data of the different use areas; performing eye box range acquisition processing to obtain eye box range position data of different use areas; hUD image display is calibrated in the vehicle manufacturing process based on the corrected eye box data of the different use areas; the method is used for solving the problem that when a vehicle-mounted HUD is calibrated in the vehicle manufacturing process through an existing HUD image calibration technology, a calibration eye box cannot be adjusted according to the eye movement range of a user in a vehicle using area, and therefore the calibrated HUD can adapt to users in different areas.

Description

Technical Field

[0001] This invention relates to the field of HUD image calibration technology, specifically to a HUD image display calibration method, device, and storage medium. Background Technology

[0002] HUD image calibration technology refers to a technical system that uses hardware adjustments, software algorithms, or multi-sensor fusion to address deviations in images projected onto a head-up display due to device characteristics, environmental interference, or changes in the user's viewing angle during the display process. Ultimately, it ensures that the image is clear, accurate, and precisely matches the real scene in the user's field of vision.

[0003] Existing HUD image calibration technologies, when calibrating in-vehicle HUDs during vehicle manufacturing, often require using industrial cameras to simulate the driver's field of vision. The cameras are placed within a designated eyebox area for calibration. However, the eyebox area is typically based on industry-standard ergonomic data, an abstract generalization of the average user. But height data for users of different genders and regions often vary, with some regions exhibiting differences exceeding the coverage of the average model. If calibration is performed using a uniform eyebox, many users in certain regions may find their eyes outside the standard eyebox area, resulting in unclear image visibility and requiring manual adjustment, thus impacting user experience. User experience is affected; if the eye box range is set too large to cover all possible situations, it will inevitably lead to a cumbersome calibration process and excessively high calibration costs. For example, patent application CN112164377A discloses a self-adaptive method for HUD image correction. The eye box used in the HUD image correction process is based on industry-standard ergonomic data and is not adjusted according to regional differences. The corrected image may not be suitable for users in all regions. Therefore, existing HUD image calibration technologies cannot adjust the calibration eye box according to the eye activity range of users in different regions when calibrating in-vehicle HUDs during vehicle manufacturing, thus failing to adapt the calibrated HUD to users in different regions. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in the prior art. It obtains the eye movement range of different users driving a vehicle in different usage areas, recording this as initial eye movement range data. The initial eye movement range data for different usage areas is preprocessed to obtain reference eye movement range data for each area. Based on the reference eye movement range data, eye box range acquisition processing is performed to obtain eye box range position data for different usage areas. Based on the corrected eye box data for different usage areas, the HUD image display is calibrated during vehicle manufacturing. This addresses the problem that existing HUD image calibration technologies, when calibrating in-vehicle HUDs during vehicle manufacturing, cannot adjust the calibration eye box according to the eye movement range of users in different usage areas, thus failing to adapt the calibrated HUD to users in different regions.

[0005] To achieve the above objectives, in a first aspect, this application provides a HUD image display calibration method, comprising the following steps: Based on the vehicle's usage area, obtain the eye movement range of different users driving the vehicle in different usage areas, and record it as the initial eye movement range data; Initial eye movement range data for different usage areas were preprocessed to obtain reference eye movement range data for different usage areas. Based on reference eye movement range data, the eye box range acquisition process is performed to obtain eye box range location data for different usage areas; Based on the correction eyebox data for different usage areas, the HUD image display is calibrated during the vehicle manufacturing process.

[0006] Furthermore, based on the vehicle's usage area, the eye movement range of different users driving the vehicle in different usage areas is obtained and recorded as initial eye movement range data, including the following sub-steps: Let any type of vehicle be designated as the first vehicle, and obtain all the usage areas of the first vehicle, which are designated as the total usage area; divide the total usage area into multiple usage areas according to geographical and administrative regions, and designate any one of the usage areas as the first usage area; A spatial rectangular coordinate system is established within the first vehicle, denoted as the first spatial coordinate system. For all first vehicles in the first usage area, the position of the driver's eyes in the first spatial coordinate system is acquired during normal driving and at the first time interval. The number of acquisitions is recorded, and the data is classified and merged according to the corresponding driver, denoted as the initial eye movement range data. The initial eye movement range data of all usage areas is collected repeatedly, where the first time interval is t1.

[0007] Furthermore, the initial eye movement range data for different usage areas are preprocessed to obtain reference eye movement range data for different usage areas, including the following sub-steps: For the initial eye movement range data of the first usage area, the number of collections for all drivers are arranged in descending order and recorded as the first number sequence; the last k1% of the first number sequence is recorded as the second number sequence, where k1% is the set percentage; Divide the second number sequence into k2 groups in equal order, and denot them as number groups 1-k2 in sequence. Calculate the average number of each group, and denot it as CP1-CPk2 in sequence, where k2 is the set number. Calculate the difference between the average values ​​of any two adjacent frequencies in CP1-CPk2, denoted as the average difference RCI, where RCI = CPi - CPi-1; and obtain the maximum average difference RCn, and denote the frequency array n corresponding to the average frequency CPn as the first array, and obtain the minimum value of the first array, denoted as the frequency threshold CS0; Drivers with fewer than CS0 data collections are categorized as "too few drivers," while those with CS0 or more data collections are categorized as "reference drivers." Data from "too few drivers" are removed from the initial eye movement range data, resulting in the baseline eye movement range data for the first usage area.

[0008] Furthermore, the initial eye movement range data for different usage areas are preprocessed to obtain reference eye movement range data for different usage areas, which includes the following sub-steps: For the basic eye-tracking range data of the first usage area, any reference driver in the basic eye-tracking range data is designated as the first driver; for the position of the first driver's eye in the first spatial coordinate system for each acquisition, the midpoint of the line connecting the left and right eye coordinates is obtained and designated as the eye point, and its coordinates are obtained. The average values ​​of the x-coordinate, y-coordinate, and vertical coordinate of all eye points of the first driver are obtained respectively and denoted as PX, PY, and PZ in order. The point with coordinates (PX, PY, PZ) is designated as the mean point. The distance from each eye point to the mean point is calculated, and the average value LP and standard deviation LB are obtained. For each eye point, if the distance to the mean point is within [LP-k3*LB, LP+k3*LB], it is considered a normal eye point; otherwise, it is marked as an abnormal eye point, where k3 is the scaling factor.

[0009] Furthermore, the initial eye movement range data for different usage areas are preprocessed to obtain reference eye movement range data for different usage areas, which includes the following sub-steps: The minimum distance from the eye point to the mean point is denoted as the expansion distance; and any normal eye point is denoted as the first eye point. For the first eye point, calculate the average distance from the nearest normal eye point (k4) to the first eye point, and record it as the nearest average distance of the first eye point; repeat this process to obtain the nearest average distance of all normal eye points; and arrange them in ascending order, and record them as the nearest average sequence, where k4 is the set number; The normal eye points corresponding to the nearest average distance of the top k5% of the nearest average sequence are recorded as candidate eye points, and any candidate eye point is recorded as the second eye point, where k5% is a set percentage. For the second eye point, take the second eye point as the center and the corresponding nearest average distance as the initial radius. Each time, increase the radius by the expansion distance and count the number of normal eye points contained in the spherical region corresponding to each radius. Calculate the eye point density based on the corresponding spherical volume. Repeat the expansion to obtain the maximum eye point density and the corresponding radius, which are recorded as the first density and the first density radius of the second eye point, respectively. Repeatedly obtain the first number density and first density radius of all candidate eye points, and record the candidate eye point with the largest first number density as the eye movement reference point, the corresponding first density radius as the eye movement reference radius, and the corresponding spherical region as the eye movement reference region. Repeatedly acquire the eye movement reference points and eye movement reference radii of all reference drivers in the basic eye movement range data. After completion, obtain the reference eye movement range data of the first use area. Repeatedly acquire the reference eye movement range data of all use areas.

[0010] Furthermore, based on the reference eye-tracking range data, the eye box range acquisition process is performed to obtain the eye box range location data for different usage areas, including the following sub-steps: For the reference eye movement range data of the first use area, the vertical coordinates of the eye movement reference points of all reference drivers are obtained and arranged in ascending order, which is recorded as the eye point height sequence. The eye point height sequence is divided into two subsequences using a density clustering algorithm, which are recorded in ascending order as the equivalent female sequence and the equivalent male sequence, respectively. The reference drivers corresponding to the vertical coordinates in the equivalent female sequence and the equivalent male sequence are respectively labeled as the equivalent female driver and the equivalent male driver. The reference eye movement range data are respectively equivalent to the female eye movement data and the equivalent male eye movement data according to the corresponding reference driver. The equivalent female eye movement data and the equivalent male eye movement data are recorded as the first eye movement range data.

[0011] Furthermore, the process of obtaining eye box range location data for different usage areas based on reference eye movement range data includes the following sub-steps: Obtain the number of times each reference driver is sampled in the first eye movement range data, sum them up to get the total number of samples, and calculate the proportion of the number of samples for each reference driver to the total number of samples, which is recorded as the sampling weight of the corresponding reference driver. All data collection weights are arranged in descending order, and the reference drivers corresponding to the top k6% of data collection weights are recorded as high-frequency drivers, where k6% is the set percentage; Obtain the minimum bounding cuboid of all high-frequency driver eye-tracking reference points, denoted as the high-frequency eye-tracking range. Select k4 points uniformly within the high-frequency eye-tracking range, denoted as candidate representative points, where k7 is the set number. Any candidate representative point is designated as the first representative point. If the first representative point is located within the eye-tracking reference area of ​​a high-frequency driver, the corresponding high-frequency driver's collection weight is accumulated. This accumulation is repeated to obtain the total accumulated weight of the first representative point, which is denoted as the total consensus score. Repeatedly obtain the total consensus of all candidate representative points, and record the candidate representative point with the highest total consensus as the general representative point; obtain the smallest circumscribed sphere with the general representative point as the center and the sum of the collection weights of the reference drivers contained therein is greater than k8%, and record it as the general representative sphere; obtain the smallest circumscribed cuboid of the general representative sphere and record it as the eye box reference range; after completion, obtain the eye box range position information, where k8% is the set percentage; Repeatedly acquire the eyebox range position information of the equivalent female eye movement data and the equivalent male eye movement data of the first use area to obtain the eyebox range position data of the first use area, and repeat acquire the eyebox range position data of all use areas.

[0012] Furthermore, based on the correction eyebox data for different usage areas, the calibration of the HUD image display during vehicle manufacturing includes the following sub-steps: For the location data of the eye box range in the first use area, based on the eye box reference ranges corresponding to the equivalent female driver and the equivalent male driver, the union of the two eye box reference ranges is obtained and recorded as the calibration eye box range of the first use area. The calibration eye box ranges of all use areas are obtained repeatedly. After the vehicle is fully assembled, the corresponding calibration eye box range is obtained according to the vehicle's usage area. The detection camera is placed within the calibration eye box range to capture the reflected image of the HUD projected onto the glass in real time. The display position, display size, display angle, distortion, ghosting, and field of view of the HUD image are then calibrated.

[0013] Secondly, this application provides an electronic device including a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method described above are performed.

[0014] Thirdly, this application provides a storage medium on which a computer program is stored, which, when executed by a processor, performs the steps of the method described above.

[0015] The beneficial effects of this invention are as follows: This invention obtains the eye movement range of different users driving a vehicle in different usage areas, based on the vehicle's usage area, and records it as initial eye movement range data; it performs eye movement range preprocessing on the initial eye movement range data for different usage areas to obtain reference eye movement range data for different usage areas; it performs eye box range acquisition processing based on the reference eye movement range data to obtain eye box range position data for different usage areas; and it calibrates the HUD image display during the vehicle manufacturing process based on the corrected eye box data for different usage areas. When calibrating the in-vehicle HUD during vehicle manufacturing, the calibration eye box can be adjusted according to the eye movement range of users in the vehicle's usage area, thereby adapting the calibrated HUD to users in different regions. This invention calculates averages by grouping data and excludes drivers with insufficient data by abrupt changes in the average, retaining only representative drivers to ensure the reliability of subsequent analysis and avoid bias from manually setting absolute thresholds. It automatically identifies eye-tracking reference points and their radii by utilizing the local density and spherical expansion around each point, generating an eye-tracking reference region. Its advantage lies in considering both the optimal concentration of local point density and ensuring the final region's inclusiveness of most normal eye points, thus improving the accuracy and robustness of subsequent eye-tracking device positioning. Furthermore, it sets collection weights based on the proportion of collection times and selects eye-tracking reference points based on overall consensus. This ensures that the actual user's gaze characteristics are reflected to the greatest extent, giving higher weight to users who drive frequently, making the resulting eye-tracking reference range closer to actual usage scenarios. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the steps of the method of the present invention; Figure 2 This is a flowchart of the eye-tracking reference region acquisition process of the present invention; Figure 3 This is a flowchart illustrating the process of obtaining the reference range for the eye box according to the present invention. Figure 4 This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0017] 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 embodiments of the present invention, and not all embodiments. 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.

[0018] Example 1, please refer to Figure 1 As shown, this application provides a HUD image display calibration method, including the following steps: Step S1: Based on the vehicle's usage area, obtain the eye movement range of different users driving the vehicle in different usage areas, and record it as the initial eye movement range data; Step S1 includes the following sub-steps: Step S101: Designate any type of vehicle as the first vehicle, obtain all the usage areas of the first vehicle, and designate them as the overall usage range; divide the overall usage range into multiple usage areas according to geographical and administrative regions, and designate any one of the usage areas as the first usage area; the area division can capture the differences in body shape and driving posture of users in different regions, making subsequent calibration more targeted and avoiding "one-size-fits-all" approaches that may cause inaccurate image viewing or user discomfort when users in certain regions use the vehicle. Step S102: Establish a spatial rectangular coordinate system within the first vehicle, denoted as the first spatial coordinate system; for all first vehicles in the first usage area, acquire the position of the driver's eyes in the first spatial coordinate system during normal driving and at the first time interval, record the number of acquisitions, and classify and merge them according to the corresponding driver, denoted as the initial eye movement range data; because there may be multiple people driving the same vehicle, all acquired data should be classified according to the driver, which can be done by using facial recognition or iris recognition to assign a unique number to each driver, thus classifying the data. Step S103: Repeatedly collect initial eye movement range data for all areas of use, wherein the first time interval is t1; in this embodiment, the first time interval is 0.5 seconds; In the specific implementation process, height is the core indicator of human body proportions, and there are significant differences in the average height and height distribution of people in different regions; while the core function of the eye box is to ensure that the user's eyes are still within a clear field of view when moving, and its parameters need to match the eye position distribution of the target population; and the height differences in different regions will indirectly change the range and position of the eye box by affecting the spatial position of the eyes, thereby affecting the direction and magnitude of HUD image correction.

[0019] Step S2 involves preprocessing the initial eye movement range data for different usage areas to obtain reference eye movement range data for each usage area. Step S2 includes the following sub-steps: Step S201: For the initial eye-tracking range data of the first usage area, arrange the number of collections for all drivers in descending order and record it as the first number sequence; record the last k1% of the first number sequence as the second number sequence, where k1% is a set percentage; in this embodiment, k1%=40%, and k1% can be set according to the actual application scenario; because the users with more collections in the previous stage have enough data and do not need to be filtered, they can be removed first for the sake of calculation simplicity; Step S202: Divide the second number sequence into k2 groups in equal order, and denot them as number groups 1-k2 in sequence. Calculate the average number of occurrences for each group, denoted as CP1-CPk2 in sequence, where k2 is the set number. In this embodiment, k2=20, and k2 can be set according to the actual application scenario. Step S203: Calculate the difference between the average values ​​of any two adjacent frequency values ​​in CP1-CPk2, denoted as the average difference RCI, where RCI = CPi - CPi-1; and obtain the maximum average difference RCn, and denote the frequency array n corresponding to the average frequency value CPn as the first array, obtain the minimum value of the first array, and denote it as the frequency threshold CS0; through group aggregation, the drivers are grouped according to the amount of data collected, and the most significant dividing point in the distribution of data collection is automatically located using the maximum difference principle, and the frequency threshold is dynamically determined.

[0020] Step S204: Drivers with fewer than CS0 sampling times are recorded as drivers with too few sampling times, and drivers with more than or equal to CS0 sampling times are recorded as reference drivers; remove the data of drivers with too few sampling times from the initial eye movement range data, and obtain the basic eye movement range data of the first use area; remove a few samples with too few sampling times to avoid them from interfering too much with subsequent statistics, while retaining sufficient data that best represents the driving population in this area. Step S205: For the basic eye movement range data of the first usage area, any reference driver in the basic eye movement range data is denoted as the first driver; for the position of the first driver's eyes in the first spatial coordinate system each time, the midpoint of the line connecting the left eye coordinate and the right eye coordinate is obtained and denoted as the eye point, and the coordinates are obtained. The positions of the two eyes are combined into a single midpoint for convenient subsequent calculation.

[0021] Step S206: Obtain the average abscissa, average ordinate, and average ordinate of all eye points of the first driver, and denote them as PX, PY, and PZ respectively in order; denote the point with coordinates (PX, PY, PZ) as the mean point, calculate the distance from each eye point to the mean point, and obtain the average value LP and standard deviation LB. Step S207: For each eye point, if the distance to the mean point is within [LP-k3*LB, LP+k3*LB], it is considered a normal eye point; otherwise, it is marked as an abnormal eye point, where k3 is a scaling factor. Occasional abnormal points or measurement noise are removed to ensure that subsequent data is based on normal eye point data. In this embodiment, k3=3.

[0022] For step S208, please refer to... Figure 2 As shown, the minimum distance from the eye point to the mean point is denoted as the expansion distance; and any normal eye point is denoted as the first eye point; the expansion distance is used as the step size for the subsequent spherical region radius growth. Step S209: For the first eye point, calculate the average distance from the nearest normal eye point (k4) to the first eye point, and record it as the nearest average distance of the first eye point; repeatedly obtain the nearest average distance of all normal eye points; and arrange them in ascending order, and record them as the nearest average sequence, where k4 is the set number; in this embodiment, k4=5, k4 can be set according to the actual application scenario, but it should not be too large in order to balance the amount of calculation; the smaller the nearest average distance, the denser the area around the first eye point; Step S210: Record the normal eye points corresponding to the nearest average distance of the top k5% of the nearest average sequence as candidate eye points, and record any candidate eye point as the second eye point, where k5% is a set percentage; in this embodiment, k5%=30%, and the ratio can be flexibly adjusted.

[0023] Step S211: For the second eye point, with the second eye point as the center and the corresponding nearest average distance as the initial radius, the radius is increased by the expansion distance each time, and the number of normal eye points contained in the spherical region corresponding to each radius is counted. Based on the corresponding spherical volume, the eye point number density is calculated. The expansion is repeated to obtain the maximum eye point number density and the corresponding radius, which are recorded as the first number density and the first density radius of the second eye point, respectively. Eye point number density = number of normal eye points contained / corresponding spherical volume. The core position and its coverage area that best represent the driver's eye position distribution are automatically identified to ensure the accuracy of the eye movement reference area.

[0024] Step S212: Repeatedly obtain the first number density and first density radius of all candidate eye points, and record the candidate eye point with the largest first number density as the eye movement reference point, record the corresponding first density radius as the eye movement reference radius, and record the corresponding spherical region as the eye movement reference region; if there are multiple largest first number densities, select the candidate eye point with the largest first density radius and record it as the eye movement reference point. Step S213: Repeatedly acquire the eye movement reference points and eye movement reference radii of all reference drivers in the basic eye movement range data. After completion, the reference eye movement range data of the first use area is obtained. Repeatedly acquire the reference eye movement range data of all use areas. In the specific implementation process, because the number of collections by most users will form a relatively concentrated normal range, while users with too few collections will form a clear gap from this range, the average value is calculated by simple grouping to find the jump point of the gap as a threshold, avoiding the deviation of manually setting absolute thresholds. By utilizing the local density and spherical expansion around each point, eye-tracking reference points and their radii are identified, and an eye-tracking reference region is generated. This takes into account the optimal concentration of local point density and ensures that the final region is inclusive of most normal eye points, avoiding the subjectivity of setting the radius based on experience, and making the eye-tracking reference region more closely match the density characteristics of the data itself. This improves the accuracy and robustness of subsequent eye box positioning.

[0025] Step S3 involves obtaining eye box range data based on reference eye movement range data to acquire eye box range location data for different usage areas. Step S3 includes the following sub-steps: Step S301: For the reference eye movement range data of the first usage area, obtain the vertical coordinates of all reference drivers' eye movement reference points and arrange them in ascending order, denoted as the eye point height sequence. Use a density clustering algorithm to divide the eye point height sequence into two subsequences, denoted in ascending order as the equivalent female sequence and the equivalent male sequence, respectively. The vertical coordinates of the eye movement reference points are the spatial position parameters of the driver's eyes in the vertical direction, which are directly related to height. At the group level, since there is a significant difference in average height between men and women, men usually have a higher average value. Their eye point height distribution will show a bimodal trend, that is, forming high-density areas around two different center values. The male group is concentrated in the higher interval, and the female group is concentrated in the lower interval.

[0026] Step S302: The reference drivers corresponding to the vertical coordinates in the equivalent female sequence and the equivalent male sequence are marked as the corresponding equivalent female driver and equivalent male driver, respectively. The reference eye movement range data is then converted into equivalent female eye movement data and equivalent male eye movement data according to the corresponding reference driver. The equivalent female eye movement data and equivalent male eye movement data are recorded as the first eye movement range data. Using the physiological difference of gender, the overall data is automatically stratified to provide a basis for subsequent fitting of the eye box range of different gender groups, ensuring that subsequent calibration is equally accurate for drivers of different genders. For step S303, please refer to [link / reference]. Figure 3As shown, the number of times each reference driver is sampled in the first eye movement range data is obtained and summed to obtain the total number of samplings. The proportion of the sampling times of each reference driver to the total number of samplings is calculated and recorded as the sampling weight of the corresponding reference driver. Drivers with more sampling times, that is, users who drive vehicles more frequently, have higher weights, which can make subsequent positioning more consistent with actual usage scenarios. Step S304: Arrange all the collection weights in descending order, and record the reference drivers corresponding to the top k6% of the collection weights as high-frequency drivers, where k6% is a set percentage; in this embodiment, k6% = 60%, which can be adjusted flexibly. Step S305: Obtain the minimum circumscribed cuboid of all high-frequency driver eye-tracking reference points, denoted as the high-frequency eye-tracking range. Select k7 points uniformly within the high-frequency eye-tracking range, denoted as candidate representative points, where k7 is the set number. In this embodiment, k7=2000, which can be set according to the actual application scenario. The larger k7 is, the more accurate the general representative point, but the greater the computational load will be.

[0027] Step S306: Record any candidate representative point as the first representative point. If the first representative point is located in the eye-tracking reference area of ​​a high-frequency driver, then accumulate the corresponding high-frequency driver's collection weight. Repeat the accumulation to obtain the total accumulated weight of the first representative point, which is recorded as the total consensus degree. The total consensus degree indicates how many users' eye-tracking reference areas cover a candidate representative point. The higher the value, the more the point can represent the dense area of ​​most people. Step S307: Repeatedly obtain the total consensus of all candidate representative points, and record the candidate representative point with the highest total consensus as the general representative point; obtain the smallest circumscribed sphere with the general representative point as the center, whose sum of the collection weights of the reference drivers is greater than k8%, and record it as the general representative sphere; obtain the smallest circumscribed cuboid of the general representative sphere, and record it as the eye box reference range; after completion, the eye box range position information is obtained, where k8% is the set percentage; in this embodiment, k8%=90%, that is, the collection weight of the vast majority of drivers, which can be adjusted appropriately; Step S308: Repeatedly acquire the eye box range position information of the equivalent female eye movement data and the equivalent male eye movement data of the first use area to obtain the eye box range position data of the first use area, and repeat to acquire the eye box range position data of all use areas; In practice, the "equivalent female sequence" and "equivalent male sequence" are not strictly classified according to gender labels, but are functionally equivalent groups formed according to the actual distribution characteristics of eye point height. For example, the vertical coordinates of the eye movement reference points of some tall women may fall into the equivalent male sequence, and the vertical coordinates of the eye movement reference points of some short men may fall into the equivalent female sequence.

[0028] Step S4 involves calibrating the HUD image display during vehicle manufacturing based on the corrected eyebox data for different usage areas. Step S4 includes the following sub-steps: Step S401: For the eye box range position data of the first use area, according to the eye box reference range corresponding to the equivalent female driver and the equivalent male driver, obtain the union of the two eye box reference ranges, and record it as the calibration eye box range of the first use area. Repeat the acquisition of the calibration eye box range of all use areas. The calibration eye box range covers all typical eye ranges of male and female drivers in the current use area, ensuring that no matter the driver's height or sitting posture, their eyes can fall within the calibration area. Step S402: After the vehicle is fully assembled, obtain the corresponding calibration eye box range according to the vehicle's usage area, place the detection camera within the calibration eye box range, and capture the reflected image of the HUD projected onto the glass in real time. Then, calibrate the display position, display size, display angle, distortion, ghosting, and field of view of the HUD image. In the actual implementation process, the calibration eye box range parameters of the region where the vehicle will be sold can be automatically retrieved from the database or assembly instructions for calibration; ensuring that the HUD image presents the best display effect when users in each region start using the vehicle.

[0029] Example 2, please refer to Figure 4 As shown, Figure 4 A schematic diagram of an electronic device is provided, which may include a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other via the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, steps such as those in a HUD image display calibration method are performed to achieve the following functions: Based on the vehicle's usage area, the eye movement range of different users driving the vehicle in different usage areas is obtained and recorded as initial eye movement range data; eye movement range preprocessing is performed on the initial eye movement range data for different usage areas to obtain reference eye movement range data for different usage areas; eye box range acquisition processing is performed based on the reference eye movement range data to obtain eye box range position data for different usage areas; and the HUD image display is calibrated during vehicle manufacturing based on the corrected eye box data for different usage areas.

[0030] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and used as independent products, or stored in a computer-readable storage medium when in use. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0031] Example 3: This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it performs the steps of a HUD image display calibration method to achieve the following functions: Based on the vehicle's usage area, it acquires the eye movement range of different users driving the vehicle in different usage areas, recording it as initial eye movement range data; it performs eye movement range preprocessing on the initial eye movement range data for different usage areas to obtain reference eye movement range data for different usage areas; it performs eye box range acquisition processing based on the reference eye movement range data to obtain eye box range position data for different usage areas; and it calibrates the HUD image display during vehicle manufacturing based on the corrected eye box data for different usage areas.

[0032] Based on the above description of the embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the above technical solutions, in essence or in terms of their contribution to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or certain parts of the embodiments.

[0033] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces. The indirect coupling or communication connection between systems, modules, and units may be electrical, mechanical, or other forms.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for calibrating HUD image display, characterized in that, Includes the following steps: Based on the vehicle's usage area, obtain the eye movement range of different users driving the vehicle in different usage areas, and record it as the initial eye movement range data; Initial eye movement range data for different usage areas were preprocessed to obtain reference eye movement range data for different usage areas. Based on reference eye movement range data, the eye box range acquisition process is performed to obtain eye box range location data for different usage areas; Based on the correction eyebox data for different usage areas, the HUD image display is calibrated during the vehicle manufacturing process.

2. The HUD image display calibration method according to claim 1, characterized in that, Based on the vehicle's usage area, the eye movement range of different users driving the vehicle in different usage areas is obtained and recorded as initial eye movement range data, including the following sub-steps: Let any type of vehicle be designated as the first vehicle, and obtain all the usage areas of the first vehicle, which are designated as the total usage area; divide the total usage area into multiple usage areas according to geographical and administrative regions, and designate any one of the usage areas as the first usage area; Establish a spatial rectangular coordinate system within the first vehicle, denoted as the first spatial coordinate system; For all first vehicles in the first usage area, when the first vehicle is driving normally, the position of the driver's eyes in the first spatial coordinate system is acquired at the first time interval. The number of acquisitions is recorded and the data is classified and merged according to the corresponding driver, and recorded as the initial eye movement range data. The initial eye movement range data of all usage areas is collected repeatedly, where the first time interval is t1.

3. The HUD image display calibration method according to claim 2, characterized in that, The initial eye movement range data for different usage areas are preprocessed to obtain reference eye movement range data for different usage areas, including the following sub-steps: For the initial eye movement range data of the first usage area, the number of collections for all drivers are arranged in descending order and recorded as the first number sequence; the last k1% of the first number sequence is recorded as the second number sequence, where k1% is the set percentage; Divide the second number sequence into k2 equal groups, and denot them sequentially as number groups 1-k2. Calculate the average number of each group, and denot it sequentially as CP1-CP2. k2 Where k2 is the number set; Calculate CP1-CP k2 The difference between the average values ​​of any two adjacent frequencies is denoted as the average difference RC. i , among which, RC i =CP i -CP i-1 And obtain the maximum average difference RC. n And the corresponding average number of times CP n The corresponding subarray n is denoted as the first array. The minimum value of the first array is denoted as the threshold CS0. Drivers with fewer than CS0 data collections are categorized as "too few drivers," while those with CS0 or more data collections are categorized as "reference drivers." Data from "too few drivers" are removed from the initial eye movement range data, resulting in the baseline eye movement range data for the first usage area.

4. The HUD image display calibration method according to claim 3, characterized in that, The process of preprocessing the initial eye movement range data for different usage areas to obtain reference eye movement range data for different usage areas also includes the following sub-steps: For the basic eye-tracking range data of the first usage area, any reference driver in the basic eye-tracking range data is designated as the first driver; for the position of the first driver's eye in the first spatial coordinate system for each acquisition, the midpoint of the line connecting the left and right eye coordinates is obtained and designated as the eye point, and its coordinates are obtained. The average values ​​of the x-coordinate, y-coordinate, and vertical coordinate of all eye points of the first driver are obtained respectively and denoted as PX, PY, and PZ in order. The point with coordinates (PX, PY, PZ) is designated as the mean point. The distance from each eye point to the mean point is calculated, and the average value LP and standard deviation LB are obtained. For each eye point, if the distance to the mean point is within [LP-k3*LB, LP+k3*LB], it is considered a normal eye point; otherwise, it is marked as an abnormal eye point, where k3 is the scaling factor.

5. The HUD image display calibration method according to claim 4, characterized in that, The process of preprocessing the initial eye movement range data for different usage areas to obtain reference eye movement range data for different usage areas also includes the following sub-steps: The minimum distance from the eye point to the mean point is denoted as the expansion distance; and any normal eye point is denoted as the first eye point. For the first eye point, calculate the average distance from the nearest k4 normal eye point to the first eye point, and record it as the nearest average distance of the first eye point; repeat this process to obtain the nearest average distance of all normal eye points. Arrange them in ascending order, and denote them as the most recent average sequence, where k4 is the number of sets; The normal eye points corresponding to the nearest average distance of the top k5% of the nearest average sequence are recorded as candidate eye points, and any candidate eye point is recorded as the second eye point, where k5% is a set percentage. For the second eye point, take the second eye point as the center and the corresponding nearest average distance as the initial radius. Each time, increase the radius by the expansion distance and count the number of normal eye points contained in the spherical region corresponding to each radius. Calculate the eye point density based on the corresponding spherical volume. Repeat the expansion to obtain the maximum eye point density and the corresponding radius, which are recorded as the first density and the first density radius of the second eye point, respectively. Repeatedly obtain the first number density and first density radius of all candidate eye points, and record the candidate eye point with the largest first number density as the eye movement reference point, the corresponding first density radius as the eye movement reference radius, and the corresponding spherical region as the eye movement reference region. Repeatedly acquire the eye movement reference points and eye movement reference radii of all reference drivers in the basic eye movement range data. After completion, obtain the reference eye movement range data of the first use area. Repeatedly acquire the reference eye movement range data of all use areas.

6. The HUD image display calibration method according to claim 5, characterized in that, The process of obtaining eye box range based on reference eye movement range data to obtain eye box range location data for different usage areas includes the following sub-steps: For the reference eye movement range data of the first use area, the vertical coordinates of the eye movement reference points of all reference drivers are obtained and arranged in ascending order, which is recorded as the eye point height sequence. The eye point height sequence is divided into two subsequences using a density clustering algorithm, which are recorded in ascending order as the equivalent female sequence and the equivalent male sequence, respectively. The reference drivers corresponding to the vertical coordinates in the equivalent female sequence and the equivalent male sequence are respectively labeled as the equivalent female driver and the equivalent male driver. The reference eye movement range data are respectively equivalent to the female eye movement data and the equivalent male eye movement data according to the corresponding reference driver. The equivalent female eye movement data and the equivalent male eye movement data are recorded as the first eye movement range data.

7. The HUD image display calibration method according to claim 6, characterized in that, The process of obtaining eye box range location data for different usage areas based on reference eye movement range data also includes the following sub-steps: Obtain the number of times each reference driver is sampled in the first eye movement range data, sum them up to get the total number of samples, and calculate the proportion of the number of samples for each reference driver to the total number of samples, which is recorded as the sampling weight of the corresponding reference driver. All data collection weights are arranged in descending order, and the reference drivers corresponding to the top k6% of data collection weights are recorded as high-frequency drivers, where k6% is the set percentage; Obtain the minimum bounding cuboid of all high-frequency driver eye-tracking reference points, denoted as the high-frequency eye-tracking range. Select k7 points uniformly within the high-frequency eye-tracking range, denoted as candidate representative points, where k7 is the set number. Any candidate representative point is designated as the first representative point. If the first representative point is located within the eye-tracking reference area of ​​a high-frequency driver, the corresponding high-frequency driver's collection weight is accumulated. This accumulation is repeated to obtain the total accumulated weight of the first representative point, which is denoted as the total consensus score. Repeatedly obtain the total consensus of all candidate representative points, and record the candidate representative point with the highest total consensus as the general representative point; obtain the smallest circumscribed sphere with the general representative point as the center and the sum of the collection weights of the reference drivers contained therein is greater than k8%, and record it as the general representative sphere; obtain the smallest circumscribed cuboid of the general representative sphere and record it as the eye box reference range; after completion, obtain the eye box range position information, where k8% is the set percentage; Repeatedly acquire the eyebox range position information of the equivalent female eye movement data and the equivalent male eye movement data of the first use area to obtain the eyebox range position data of the first use area, and repeat acquire the eyebox range position data of all use areas.

8. The HUD image display calibration method according to claim 7, characterized in that, The calibration of HUD image display during vehicle manufacturing, based on corrected eyebox data for different usage areas, includes the following sub-steps: For the location data of the eye box range in the first use area, based on the eye box reference ranges corresponding to the equivalent female driver and the equivalent male driver, the union of the two eye box reference ranges is obtained and recorded as the calibration eye box range of the first use area. The calibration eye box ranges of all use areas are obtained repeatedly. After the vehicle is fully assembled, the corresponding calibration eye box range is obtained according to the vehicle's usage area. The detection camera is placed within the calibration eye box range to capture the reflected image of the HUD projected onto the glass in real time. The display position, display size, display angle, distortion, ghosting, and field of view of the HUD image are then calibrated.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing computer-readable instructions that, when executed by the processor, perform the steps of the method as described in any one of claims 1-8.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it performs the steps of the method as described in any one of claims 1-8.

Images