ARHUD after-sales calibration method and system based on diagnostic instrument

Abstract

The invention provides an ARHUD after-sale calibration method and system based on a diagnostic apparatus, and the method comprises the steps: accessing the diagnostic apparatus, transmitting a calibration mode starting instruction through the diagnostic apparatus to trigger a head-up display system to enter a calibration mode, and loading a default calibration model and calibration parameters of the system to the head-up display system, and sending a calibration image calling instruction to control the head-up display system to project a calibration image, receiving an adjustment signal of a user to the position of the eye box through a diagnostic instrument, obtaining a distortion adjustment mode selected by the user, collecting the projected calibration image in real time, and dynamically correcting calibration parameters according to the distortion adjustment mode selected by the user. The corrected new calibration parameters are verified, and after verification is passed, a calibration parameter updating instruction is sent through the diagnostic instrument to control the head-up display system to apply the new calibration parameters. According to the invention, the diagnostic instrument of the vehicle machine is combined with an image processing algorithm, so that a convenient and efficient after-sales software distortion correction function without depending on calibration equipment is realized, and the method is suitable for all intelligent network connection vehicles carrying AR HUD (Augmented Reality Head Up Display).

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automotive augmented reality head-up display (ARHUD) technology, in particular to an ARHUD post-sale calibration method based on a diagnostic instrument and an ARHUD post-sale calibration system based on a diagnostic instrument. BACKGROUND

[0002] AR HUD (augmented reality head-up display) is a vehicle-mounted display technology that fuses virtual information with real road scenes, projects key information such as navigation, speed, and driving assistance in front of the driver's line of sight through the windshield, and superimposes it in real time with the actual road environment (such as lanes, vehicles, and pedestrians), reduces the frequency of the driver looking down at the instrument panel, realizes dynamic interaction between man and machine, and improves driving immersion and safety.

[0003] Currently, the ARHUD calibration commonly used in the market relies on factory calibration and special hardware devices (such as calibration versions, gantry frames, cameras, optical benches, etc.), and the default calibration parameters of the ARHUD itself are generally based on the middle eye position. However, in the actual use of the vehicle, the ARHUD may need to match the eye box of drivers of different heights, and the default parameters of the optical machine cannot meet the needs of different customers.

[0004] During the post-sale process, there may also be problems of projection image distortion caused by vehicle collision, windshield replacement, and temperature deformation. The existing post-sale calibration scheme requires the vehicle to be returned to the factory or to use professional calibration equipment for re-calibration, which is time-consuming and costly, and requires a certain level of expertise from the calibration personnel. Other ARHUD software calibration schemes lack a closed-loop correction process linked to the on-board diagnostic system (OBD), and the calibration system is not perfect. SUMMARY

[0005] In view of the above problems, the present application is proposed to provide an ARHUD post-sale calibration method based on a diagnostic instrument and a corresponding ARHUD post-sale calibration system based on a diagnostic instrument, which can overcome the above problems or at least partially solve the above problems.

[0006] The present application discloses an ARHUD post-sale calibration method based on a diagnostic instrument, which comprises:

[0007] Accessing a diagnostic instrument;

[0008] Sending a calibration mode start instruction through the diagnostic instrument to trigger the head-up display system to enter the calibration mode;

[0009] Loading the system default calibration model and calibration parameters into the head-up display system and sending a calibration image calling instruction to control the head-up display system to project the calibration image;

[0010] The diagnostic instrument receives a user adjustment signal for the eyebox position, and obtains a user-selected distortion adjustment mode;

[0011] The projected calibration image is collected in real time, and the calibration parameters are dynamically corrected according to the user-selected distortion adjustment mode;

[0012] The corrected new calibration parameters are verified, and after the new calibration parameters pass the verification, the diagnostic instrument sends a calibration parameter update instruction to control the head-up display system to apply the new calibration parameters.

[0013] Optionally, the projected calibration image is collected in real time, and the calibration parameters are dynamically corrected according to the user-selected distortion adjustment mode, including:

[0014] The projected calibration image is collected in real time, and an image feature recognition algorithm is used to process the calibration image to generate a correction mapping table;

[0015] A linear interpolation algorithm and a matrix algorithm are used to perform coordinate mapping based on the correction mapping table and the user-selected distortion adjustment mode, and the calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters.

[0016] Optionally, the projected calibration image is collected in real time, and an image feature recognition algorithm is used to process the calibration image to generate a correction mapping table, including:

[0017] The FAST feature detection algorithm is used to preliminarily screen the corner points of the calibration image;

[0018] A Gaussian pyramid model is constructed based on the screened corner point positions;

[0019] The corner points are accurately positioned based on the gradient amplitude and direction in the Gaussian pyramid model, and a binary descriptor is generated;

[0020] The binary descriptor is matched and optimized to generate a correction mapping table.

[0021] Optionally, a linear interpolation algorithm and a matrix algorithm are used to perform coordinate mapping based on the correction mapping table and the user-selected distortion adjustment mode, and the calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters, including:

[0022] The coordinate data in the correction mapping table is normalized, and the coordinate position information of adjacent pixels is calculated by a bicubic interpolation algorithm;

[0023] According to the user-selected distortion adjustment mode, a corresponding transformation matrix algorithm is dynamically called to perform coordinate mapping;

[0024] The calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters.

[0025] Optionally, the coordinate mapping is performed according to a corresponding transformation matrix algorithm dynamically invoked according to a user-selected distortion adjustment mode, including:

[0026] When the translation or rotation distortion adjustment is selected, the coordinate mapping is performed based on a rigid body transformation matrix;

[0027] When the trapezoidal distortion adjustment is selected, the coordinate mapping is performed based on a slope matrix in four directions of up, down, left and right of the image;

[0028] When the smile distortion adjustment is selected, the coordinate mapping is performed based on an arc matrix in four directions of up, down, left and right of the image.

[0029] Optionally, the method further includes:

[0030] The calibration image projected in real time is processed by using a multi-thread accelerated image feature recognition algorithm to generate a correction mapping table.

[0031] The application further discloses an ARHUD post-sale calibration system based on a diagnostic instrument, the system comprising:

[0032] A diagnostic instrument access module is configured to access the diagnostic instrument;

[0033] A calibration start module is configured to send a calibration mode start instruction through the diagnostic instrument to trigger the head-up display system to enter a calibration mode;

[0034] An initial calibration information loading module is configured to load a system default calibration model and calibration parameter to the head-up display system, and send a calibration image calling instruction to control the head-up display system to project a calibration image;

[0035] A distortion adjustment mode selection module is configured to receive a user adjustment signal for the eyebox position through the diagnostic instrument, and obtain a user-selected distortion adjustment mode;

[0036] A calibration parameter correction module is configured to collect a projected calibration image in real time and dynamically correct the calibration parameter according to the user-selected distortion adjustment mode;

[0037] A new calibration parameter verification application module is configured to verify the corrected new calibration parameter, and send a calibration parameter update instruction through the diagnostic instrument to control the head-up display system to apply the new calibration parameter after the new calibration parameter is verified.

[0038] Optionally, the calibration parameter correction module comprises:

[0039] A correction mapping table generation sub-module is configured to collect a projected calibration image in real time and process the calibration image by using an image feature recognition algorithm to generate a correction mapping table;

[0040] The coordinate mapping submodule is configured to perform coordinate mapping based on a rectification mapping table and a user-selected distortion adjustment mode by using a linear interpolation algorithm and a matrix algorithm, and update calibration parameters according to a coordinate mapping result to obtain new calibration parameters.

[0041] Optionally, the rectification mapping table generation submodule comprises:

[0042] The corner point preliminary screening unit is configured to perform corner point preliminary screening on the calibration image by using a FAST feature detection algorithm.

[0043] The scale space construction unit is configured to construct a Gaussian pyramid model at the position of the screened corner point.

[0044] The key point accurate positioning and description unit is configured to accurately position the corner point based on a gradient amplitude and a direction in the Gaussian pyramid model, and generate a binary descriptor.

[0045] The optimization processing unit is configured to perform matching and optimization processing on the binary descriptor to generate a rectification mapping table.

[0046] Optionally, the coordinate mapping submodule comprises:

[0047] The adjacent pixel coordinate position calculation unit is configured to normalize coordinate data in the rectification mapping table, and calculate coordinate position information of adjacent pixels by using a bicubic interpolation algorithm.

[0048] The coordinate transformation unit is configured to dynamically call a corresponding transformation matrix algorithm to perform coordinate mapping according to a user-selected distortion adjustment mode.

[0049] The calibration parameter update unit is configured to update calibration parameters according to a coordinate mapping result to obtain new calibration parameters.

[0050] Optionally, the coordinate transformation unit comprises:

[0051] The translation-rotation distortion adjustment unit is configured to perform coordinate mapping based on a rigid body transformation matrix when translation or rotation distortion adjustment is selected.

[0052] The trapezoidal distortion adjustment unit is configured to perform coordinate mapping based on a slope matrix in four directions (up, down, left and right) of the image when trapezoidal distortion adjustment is selected.

[0053] The smile distortion adjustment unit is configured to perform coordinate mapping based on an arc matrix in four directions (up, down, left and right) of the image when smile distortion adjustment is selected.

[0054] Optionally, the system further comprises:

[0055] The multi-thread acceleration module is configured to collect a projected calibration image in real time, and process the calibration image by using a multi-thread accelerated image feature recognition algorithm to generate a rectification mapping table.

[0056] The application comprises the following advantages:

[0057] The ARHUD post-sale calibration method based on a diagnostic instrument of the application accesses the diagnostic instrument, sends a calibration mode starting instruction through the diagnostic instrument to trigger the head-up display system to enter the calibration mode, loads the system default calibration model and calibration parameters to the head-up display system, and sends a calibration image calling instruction to control the head-up display system to project the calibration image, receives the user's adjustment signal for the eyebox position through the diagnostic instrument, and obtains the user's selected distortion adjustment mode, collects the projected calibration image in real time and dynamically corrects the calibration parameters according to the user's selected distortion adjustment mode, verifies the corrected new calibration parameters, and after the new calibration parameters pass the verification, sends a calibration parameter updating instruction through the diagnostic instrument to control the head-up display system to apply the new calibration parameters. The application combines the diagnostic instrument of the vehicle machine with the image processing algorithm, realizes the convenient and efficient post-sale software distortion correction function without relying on calibration equipment, and is suitable for all intelligent networked vehicles equipped with AR HUD. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 is a step flowchart of an ARHUD post-sale calibration method based on a diagnostic instrument provided by an embodiment of the application;

[0059] Figure 2 is a UDS message and response schematic diagram of a diagnostic instrument provided by an embodiment of the application;

[0060] Figure 3 is a flowchart of ARHUD post-sale calibration based on a diagnostic instrument provided by an embodiment of the application;

[0061] Figure 4 is a distortion adjustment mode schematic diagram provided by an embodiment of the application;

[0062] Figure 5 is an image processing flowchart provided by an embodiment of the application. DETAILED DESCRIPTION

[0063] In order to make the above-mentioned purposes, features and advantages of the application more obvious and easy to understand, the application will be further described in detail below with reference to the drawings and specific embodiments.

[0064] Referring to Figure 1 , a step flowchart of an ARHUD post-sale calibration method based on a diagnostic instrument provided by an embodiment of the application is shown, which can specifically include the following steps:

[0065] Step 101, access the diagnostic instrument;

[0066] Step 102, send a calibration mode starting instruction through the diagnostic instrument to trigger the head-up display system to enter the calibration mode;

[0067] Step 103, load the system default calibration model and calibration parameters to the head-up display system, and send a calibration image calling instruction to control the head-up display system to project a calibration image;

[0068] Step 104, receive the user's adjustment signal for the eyebox position through the diagnostic instrument, and obtain the distortion adjustment mode selected by the user;

[0069] Step 105, real-time collection of the projected calibration image and dynamic correction of the calibration parameters according to the distortion adjustment mode selected by the user;

[0070] Step 106, verification of the corrected new calibration parameters, and after the new calibration parameters pass the verification, sending a calibration parameter update instruction through the diagnostic instrument to control the head-up display system to apply the new calibration parameters.

[0071] In an optional embodiment of the present application, real-time collection of the projected calibration image and dynamic correction of the calibration parameters according to the distortion adjustment mode selected by the user include:

[0072] Real-time collection of the projected calibration image and processing of the calibration image using an image feature recognition algorithm to generate a correction mapping table;

[0073] Using a linear interpolation algorithm and a matrix algorithm, coordinate mapping is performed based on the correction mapping table and the distortion adjustment mode selected by the user, and the calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters.

[0074] The present application combines the diagnostic instrument of the vehicle machine with the image processing algorithm to realize the convenient and efficient after-sales software distortion correction function without relying on calibration equipment. Specifically, based on the UDS communication protocol of the diagnostic instrument, by sending a signal instruction, the host computer of the diagnostic instrument and the HUD after-sales calibration software are connected to realize different distortion adjustment functions.

[0075] The diagnostic instrument protocol structure is as follows:

[0076] OBD-II physical layer→CAN bus→ISO-TP transmission layer→UDS service layer

[0077] The UDS message and response are as follows: Figure 2 .

[0078] Referring to Figure 3 , the process of ARHUD after-sales calibration based on the diagnostic instrument is as follows:

[0079] 1. The user enters the calibration mode through the vehicle machine OBD diagnostic instrument;

[0080] 2. Load the default calibration model and parameter information;

[0081] 3. The diagnostic instrument sends a command to the HUD to call up the calibration image display;

[0082] 4. The user adjusts the HUD eyebox position according to the height;

[0083] 5. The user selects different distortion modes through the diagnostic instrument sending a command;

[0084] 6. The image feature recognition algorithm corrects the calibration model in real time, and the multi-thread mode accelerates the calibration model fitting;

[0085] 7. The linear interpolation algorithm and matrix operation are used for coordinate mapping;

[0086] 8. The diagnostic instrument sends a command to update the calibration parameters and apply them, checks the data, and completes the calibration.

[0087] In an optional embodiment of the present application, the projected calibration image is collected in real time, and an image feature recognition algorithm is used to process the calibration image to generate a correction mapping table, comprising:

[0088] The FAST feature detection algorithm is used to preliminarily screen the corner points of the calibration image;

[0089] A Gaussian pyramid model is constructed at the screened corner point positions;

[0090] The corner points are accurately positioned based on the gradient amplitude and direction in the Gaussian pyramid model, and a binary descriptor is generated;

[0091] The binary descriptor is subjected to matching and optimization processing to generate a correction mapping table.

[0092] In the present embodiment, the image feature recognition algorithm used is an improved SIFT-FAST hybrid feature detection method, specifically:

[0093] (1) FAST feature detection (fast preliminary screening)

[0094] The FAST algorithm quickly detects the corner points by comparing the pixel values on the continuous arcs in the neighborhood of the pixel point with the center pixel.

[0095] Adaptive threshold: dynamically adjust the threshold according to the local contrast of the image (use the variance of the surrounding pixels).

[0096] Non-maximum suppression: retain the strongest corner points to avoid dense repetition.

[0097] Formula:

[0098] For the center pixel p and the threshold t, if the continuous N (usually 12) points on the arc satisfy:

[0099] | I(x)-I(p) | > t (bright or dark corner point)

[0100] Where I(x) is the intensity of neighboring pixels, and the adaptive threshold calculation formula is:

[0101] t=α·σ local +β

[0102] Where, σ local α represents the local standard deviation, and β represents the adjustment coefficients.

[0103] (2) Scale-space construction (SIFT improvement)

[0104] Gaussian pyramids are constructed at the corner locations detected by FAST to achieve scale invariance. Specifically, this includes:

[0105] Local scale optimization: Multi-scale space is calculated only around the corners of FAST, reducing computational cost.

[0106] Pyramid downsampling strategy: Use non-linear intervals to improve coverage of key scales.

[0107] formula:

[0108] The Gaussian pyramid layer L(x,y,σ) is obtained by convolving the image I(x,y) with the Gaussian kernel G(x,y,σ).

[0109] L(x,y,σ)=G(x,y,σ)*I(x,y)

[0110] The formula for calculating the scale interval is: σ k =σ0·2 k / s (s is the number of scales per layer, and k is the optimized exponent).

[0111] (3) Precise location and description of key points

[0112] In the scale space near the FAST corner points, gradient information from SIFT is used to accurately locate and generate descriptors. Specifically, this includes:

[0113] Direction assignment: Use the gradient histogram of the FAST corner neighborhood (simplifies calculation).

[0114] Binary descriptors: Improve speed by combining BRIEF or ORB binary descriptor methods.

[0115] formula:

[0116] Gradient magnitude and direction:

[0117]

[0118] Binary descriptors (such as BRIEF):

[0119]

[0120] where p is a key point, and x, y are random point pairs in the neighborhood.

[0121] After feature extraction and initial matching of the collected calibration image by the improved SIFT-FAST hybrid feature detection method, the matching points are screened by the RANSAC algorithm, the inliers are input into the LM optimization algorithm to generate a correction mapping table, and the outliers trigger re-collection of the image.

[0122] In an optional embodiment of the present application, a linear interpolation algorithm and a matrix algorithm are used to perform coordinate mapping based on the correction mapping table and a distortion adjustment mode selected by a user, and the calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters, including:

[0123] The coordinate data in the correction mapping table is normalized, and the coordinate position information of adjacent pixels is calculated by a bicubic interpolation algorithm;

[0124] According to the user-selected distortion adjustment mode, the corresponding transformation matrix algorithm is dynamically called to perform coordinate mapping;

[0125] The calibration parameters are updated according to the coordinate mapping result to obtain new calibration parameters.

[0126] In this embodiment, the coordinate data in the correction mapping table is normalized, and the coordinate position information of adjacent pixels is calculated by a bicubic interpolation algorithm, wherein the coordinate value of a target point (x, y) is obtained by weighted summation of 16 neighborhood pixel coordinates:

[0127]

[0128] In the above formula, (x i ,y j ) is the coordinate of the 16 neighborhood pixels; and R(·) is a cubic weight function.

[0129] After the normalization and interpolation operation of the mapping table, the corresponding transformation matrix is called to perform coordinate transformation based on the user-selected distortion adjustment mode, and the coordinate mapping relationship is updated.

[0130] The coordinate translation and rotation transformation are realized by matrix operation:

[0131]

[0132] where matrix A represents the transformation matrix, represents the original image coordinate data, and represents the target image coordinate data.

[0133] In an optional embodiment of the present application, according to the user-selected distortion adjustment mode, the corresponding transformation matrix algorithm is dynamically called to perform coordinate mapping, including:

[0134] When the translation or rotation distortion adjustment is selected, the coordinate mapping is performed based on the rigid body transformation matrix;

[0135] When the trapezoidal distortion adjustment is selected, the coordinate mapping is performed based on the slope matrix in the four directions of the image;

[0136] When the smile distortion adjustment is selected, the coordinate mapping is performed based on the radian matrix in the four directions of the image.

[0137] In the embodiment, after the coordinate normalization and interpolation operation on the mapping table, the coordinate mapping is performed based on the distortion adjustment mode selected by the user, including the up-down-left-right movement, rotation, trapezoidal, smile arc and other distortion adjustment modes. The translation and rotation transformation refer to the rigid body transformation matrix, the trapezoidal transformation is performed by calculating the slope matrix in the four directions of the image for trapezoidal distortion adjustment, the smile transformation is performed based on the radian matrix in the four directions of the image for distortion adjustment, and finally the coordinate mapping is performed based on the above matrix to update the distortion parameters.

[0138] In an optional embodiment of the present application, the method further comprises:

[0139] The projected calibration image is collected in real time, and the image feature recognition algorithm accelerated by multi-threading is used to process the calibration image to generate a correction mapping table.

[0140] In the embodiment, the calibration model fitting is accelerated by thread division, and the architecture is as follows:

[0141] (1) Main thread: responsible for feature point collection and result verification;

[0142] (2) Worker thread 1: RANSAC algorithm to remove outliers;

[0143] (3) Worker thread 2: LM algorithm to optimize the homography matrix.

[0144] The following is the complete process of the ARHUD post-sales calibration example based on the diagnostic instrument of the present application:

[0145] 1. Initialization phase

[0146] Ignition switch ON;

[0147] The user adjusts the pose state and HUD height gear;

[0148] Send the command to enter the calibration mode through the car diagnostic instrument:

[0149] Start the calibration mode

[0150] 78C 04 31 01 3C 14 AA AA AA

[0151] 79 C 05 71 01 3C 14 XX AA AA

[0152] Load pre-distortion correction model;

[0153] HUD display standard lattice:

[0154] Display calibration picture

[0155] 78 C 05 2F 3C 0C 03 09 AA AA

[0156] 79 C 03 6F 3C 0C AA AA AA AA

[0157] 2, Calibration stage

[0158] User sends calibration adjustment instruction through diagnostic instrument

[0159] Custom UDS service ID:

[0160] "Adjust four-way position and angle"

[0161] 78 C 05 2E 3C 09 XX YY AA AA

[0162] 79 C 03 6E 3C 09 AA AA AA AA

[0163] Image up: YY

[0164] Image down: YY

[0165] Image left: YY

[0166] Image right: YY

[0167] 78 C 05 2E 3C 0A XX YY AA AA

[0168] 79 C 03 6E 3C 0A AA AA AA AA

[0169] Image counterclockwise rotation: YY

[0170] Image clockwise rotation: YY

[0171] "Adjust image distortion"

[0172] 78 C 04 2E 3C 0B XX AA AA AA

[0173] 79 C 03 6E 3C 0B AA AA AA AA

[0174] The diagnostic instrument interface displays 8 distortion pictures, the user selects the corresponding distortion adjustment mode, and automatically sends a 2E command for adjustment. Figure 4 The distortion adjustment mode is shown in the schematic diagram.

[0175] “Post-calibration check”

[0176] 78C 05 2F 3C 0D 03 00 AA AA

[0177] 79C 03 6F 3C 0D AA AA AA AA

[0178] 3. Image processing algorithm real-time correction of distortion correction model:

[0179] A [feature extraction]-->B [initial matching]

[0180] B-->C {RANSAC screening}

[0181] C-->|inliers|D [LM optimization]

[0182] C-->|outliers|E [re-collection]

[0183] D-->F [generate correction mapping table]

[0184] Figure 5 The image processing flowchart is shown.

[0185] 4. Data preprocessing stage

[0186] According to the generated correction mapping table, the coordinate data is normalized, and then the coordinates of adjacent pixels are calculated by the bicubic interpolation algorithm.

[0187] 5. Coordinate mapping

[0188] After image feature recognition and data preprocessing are completed, coordinate mapping is performed based on the user-selected distortion adjustment mode, including translation, rotation, trapezoidal, and smile transformation. Translation and rotation transformation refer to the rigid body transformation matrix, trapezoidal distortion adjustment is performed by calculating the slope matrix in the up, down, left, and right directions of the image, smile transformation is based on the radian matrix in the up, down, left, and right directions of the image for distortion adjustment, and finally coordinate mapping is performed based on the above matrix to update the distortion parameters.

[0189] 6. Distortion parameter writing stage

[0190] After updating the distortion parameters, checksum verification is performed, and the data is written into the EEPROM through UDS.

[0191] Storage format:

[0192] [HEADER][CRC32][Distortion parameters]

[0193] 7、User confirms the calibration effect, and the car machine applies new distortion correction parameters, and exits the calibration mode.

[0194] The present application has the following technical effects:

[0195] 1、User can complete independent calibration (pure software calibration) through the diagnostic instrument of the car machine, without additional calibration equipment;

[0196] 2、Improve data preprocessing efficiency and calibration model fitting accuracy through improved image processing algorithm;

[0197] 3、Multi-threading accelerates model fitting, shortening the HUD after-sales calibration time;

[0198] 4、Cover multiple distortion adjustment modes, support automatic compensation for translation and rotation, and meet the adjustment needs of different users.

[0199] It should be noted that, for the method embodiments, in order to simply describe, they are all described as a series of action combinations, but those skilled in the art should know that the embodiments of the present application are not limited by the action sequence described, because according to the embodiments of the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions involved are not necessarily necessary for the embodiments of the present application.

[0200] The ARHUD after-sales calibration system based on a diagnostic instrument provided in the embodiments of the present application can specifically include the following modules:

[0201] The diagnostic instrument access module is used to access the diagnostic instrument;

[0202] The calibration start module is used to send a calibration mode start instruction through the diagnostic instrument to trigger the head-up display system to enter the calibration mode;

[0203] The initial calibration information loading module is used to load the system default calibration model and calibration parameters to the head-up display system, and send a calibration image calling instruction to control the head-up display system to project the calibration image;

[0204] The distortion adjustment mode selection module is used to receive the user's adjustment signal for the eyebox position through the diagnostic instrument, and obtain the user's selected distortion adjustment mode;

[0205] The calibration parameter correction module is used to collect the projected calibration image in real time and dynamically correct the calibration parameters according to the user's selected distortion adjustment mode;

[0206] The new calibration parameter verification application module is configured to verify the modified new calibration parameter, and send a calibration parameter update instruction to control the head-up display system to apply the new calibration parameter after the new calibration parameter is verified.

[0207] In an optional embodiment of the present application, the calibration parameter modification module comprises:

[0208] The correction mapping table generation submodule is configured to collect the projected calibration image in real time and process the calibration image by using an image feature recognition algorithm to generate a correction mapping table.

[0209] The coordinate mapping submodule is configured to perform coordinate mapping based on the correction mapping table and a distortion adjustment mode selected by the user by using a linear interpolation algorithm and a matrix algorithm, and update the calibration parameter according to the coordinate mapping result to obtain a new calibration parameter.

[0210] In an optional embodiment of the present application, the correction mapping table generation submodule comprises:

[0211] The corner point preliminary screening unit is configured to perform corner point preliminary screening on the calibration image by using a FAST feature detection algorithm.

[0212] The scale space construction unit is configured to construct a Gaussian pyramid model at the position of the screened corner point.

[0213] The key point accurate positioning and description unit is configured to accurately position the corner point based on the gradient amplitude and direction in the Gaussian pyramid model, and generate a binary descriptor.

[0214] The optimization processing unit is configured to perform matching and optimization processing on the binary descriptor to generate the correction mapping table.

[0215] In an optional embodiment of the present application, the coordinate mapping submodule comprises:

[0216] The adjacent pixel coordinate position calculation unit is configured to normalize the coordinate data in the correction mapping table, and calculate the coordinate position information of adjacent pixels by using a bicubic interpolation algorithm.

[0217] The coordinate transformation unit is configured to dynamically call a corresponding transformation matrix algorithm to perform coordinate mapping according to the distortion adjustment mode selected by the user.

[0218] The calibration parameter update unit is configured to update the calibration parameter according to the coordinate mapping result to obtain a new calibration parameter.

[0219] In an optional embodiment of the present application, the coordinate transformation unit comprises:

[0220] The translation and rotation distortion adjustment unit is configured to perform coordinate mapping based on a rigid body transformation matrix when the translation or rotation distortion adjustment is selected.

[0221] a trapezoidal distortion adjustment unit, configured to perform coordinate mapping based on a slope matrix of four directions of up, down, left and right of the image when trapezoidal distortion adjustment is selected;

[0222] a smile distortion adjustment unit, configured to perform coordinate mapping based on a radian matrix of four directions of up, down, left and right of the image when smile distortion adjustment is selected.

[0223] In an optional embodiment of the present application, the system further comprises:

[0224] a multi-thread acceleration module, configured to collect the projected calibration image in real time, and process the calibration image by using a multi-thread accelerated image feature recognition algorithm to generate the correction mapping table.

[0225] For the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts refer to the part of the method embodiment.

[0226] It should be noted that, in this document, the terms such as first and second are merely used to distinguish one entity or operation from another, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, article or device including the element.

[0227] Each embodiment in the specification is described in a relevant manner, and the same and similar parts between each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts refer to the part of the method embodiment.

[0228] The above only describes the preferred embodiments of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A diagnostic instrument based ARHUD post sale calibration method, characterized in that, The method comprises: accessing a diagnostic instrument; sending a calibration mode starting instruction through the diagnostic instrument to trigger the head-up display system to enter a calibration mode; loading system default calibration models and calibration parameters to the head-up display system, and sending a calibration image calling instruction to control the head-up display system to project a calibration image; receiving a user adjustment signal for the eyebox position through the diagnostic instrument, and obtaining a user-selected distortion adjustment mode; real-time collection of the projected calibration image and dynamic correction of the calibration parameters according to the user-selected distortion adjustment mode; verification of the corrected new calibration parameters, and sending a calibration parameter update instruction through the diagnostic instrument to control the head-up display system to apply the new calibration parameters after the new calibration parameters pass the verification.

2. The method of claim 1, wherein, The real-time collection of the projected calibration image and the dynamic correction of the calibration parameters according to the user-selected distortion adjustment mode comprise: real-time collection of the projected calibration image and processing of the calibration image by using an image feature recognition algorithm to generate a correction mapping table; coordinate mapping based on the correction mapping table and the user-selected distortion adjustment mode by using a linear interpolation algorithm and a matrix algorithm, and updating of the calibration parameters according to the coordinate mapping result to obtain new calibration parameters.

3. The method of claim 2, wherein, The real-time collection of the projected calibration image and the processing of the calibration image by using an image feature recognition algorithm to generate a correction mapping table comprise: preliminary screening of corner points of the calibration image by using a FAST feature detection algorithm; construction of a Gaussian pyramid model at the positions of the screened corner points; precise positioning of the corner points based on gradient amplitudes and directions in the Gaussian pyramid model, and generation of binary descriptors; matching and optimization processing of the binary descriptors to generate the correction mapping table.

4. The method of claim 2, wherein, The coordinate mapping based on the correction mapping table and the user-selected distortion adjustment mode by using a linear interpolation algorithm and a matrix algorithm, and the updating of the calibration parameters according to the coordinate mapping result to obtain new calibration parameters comprise: normalization processing of coordinate data in the correction mapping table, and calculation of coordinate position information of adjacent pixels by using a bicubic interpolation algorithm; dynamic calling of a corresponding transformation matrix algorithm according to the user-selected distortion adjustment mode to perform coordinate mapping; updating of the calibration parameters according to the coordinate mapping result to obtain new calibration parameters.

5. The method of claim 4, wherein, The dynamic calling of the corresponding transformation matrix algorithm according to the user-selected distortion adjustment mode to perform coordinate mapping comprises: when translation or rotation distortion adjustment is selected, coordinate mapping is performed based on a rigid body transformation matrix; when trapezoidal distortion adjustment is selected, coordinate mapping is performed based on slope matrices of four directions of the image; when smile distortion adjustment is selected, coordinate mapping is performed based on radian matrices of four directions of the image.

6. The method of claim 1, wherein, The method further comprises: real-time collection of the projected calibration image, and processing of the calibration image by using a multi-thread accelerated image feature recognition algorithm to generate a correction mapping table.

7. A diagnostic instrument based ARHUD post sale calibration system, characterized in that, The system comprises: a diagnostic instrument access module for accessing a diagnostic instrument; a calibration starting module for sending a calibration mode starting instruction through the diagnostic instrument to trigger the head-up display system to enter a calibration mode; an initial calibration information loading module for loading system default calibration models and calibration parameters to the head-up display system, and sending a calibration image calling instruction to control the head-up display system to project a calibration image; The distortion adjustment mode selection module is configured to receive a user adjustment signal for the eyebox position via the diagnostic instrument and obtain a user-selected distortion adjustment mode. The calibration parameter correction module is configured to collect a projected calibration image in real time and dynamically correct the calibration parameter according to the user-selected distortion adjustment mode. The new calibration parameter verification application module is configured to verify the corrected new calibration parameter, and send a calibration parameter update instruction to control the head-up display system to apply the new calibration parameter after the new calibration parameter passes the verification.

8. The system of claim 7, wherein, The calibration parameter correction module includes: The correction mapping table generation submodule is configured to collect a projected calibration image in real time and process the calibration image using an image feature recognition algorithm to generate a correction mapping table. The coordinate mapping submodule is configured to perform coordinate mapping based on the correction mapping table and the user-selected distortion adjustment mode using a linear interpolation algorithm and a matrix algorithm, and update the calibration parameter according to the coordinate mapping result to obtain a new calibration parameter.

9. The system of claim 8, wherein, The correction mapping table generation submodule includes: The corner point preliminary screening unit is configured to perform corner point preliminary screening on the calibration image using a FAST feature detection algorithm. The scale space construction unit is configured to construct a Gaussian pyramid model at the screened corner point position. The key point precise positioning and description unit is configured to precisely position the corner point based on a gradient amplitude and direction in the Gaussian pyramid model, and generate a binary descriptor. The optimization processing unit is configured to perform matching and optimization processing on the binary descriptor to generate a correction mapping table.

10. The system of claim 8, wherein, The coordinate mapping submodule includes: The adjacent pixel coordinate position calculation unit is configured to normalize the coordinate data in the correction mapping table, and calculate coordinate position information of adjacent pixels using a bicubic interpolation algorithm. The coordinate transformation unit is configured to dynamically call a corresponding transformation matrix algorithm to perform coordinate mapping according to the user-selected distortion adjustment mode. The calibration parameter update unit is configured to update the calibration parameter according to the coordinate mapping result to obtain a new calibration parameter.

Images