A method and device for detecting projection effect of vehicle-mounted HUD system

Abstract

The invention discloses a method and device for detecting the projection effect of a vehicle-mounted HUD system. The method comprises the following steps: S1, placing a calibration plate in front of the vehicle, calibrating a human eye simulation device, and establishing a conversion relationship between the calibration plate coordinate system and the vehicle coordinate system; S2, adjusting the position of the human eye simulation device, measuring the visible area and the best visible area of ​​the HUD; the maximum inscribed rectangular area of ​​the best visible area is recorded as the EyeBox area of ​​the HUD system to be tested; S3, collecting the calibration plate image and the HUD projection virtual image image obtained by the human eye simulation device at each viewpoint in the EyeBox area, and calculating the virtual image imaging distance (VID), contrast, distortion, brightness uniformity, brightness adjustment range, ghosting, and field of view FOV.

Description

Technical Field

[0001] The present invention relates to the technical field of HUD detection, and in particular to a method and device for detecting projection effects of a vehicle-mounted HUD system. Background Art

[0002] The in-vehicle HUD projects driving assistance information onto the windshield and then reflects it into the driver's forward field of view, which can not only expand the driver's environmental perception information, but also prevent the driver from looking down at the instruments too much, thereby effectively improving driving safety.

[0003] In order to ensure that all functional indicators of the vehicle HUD can meet the design requirements and use requirements before it is put into use on the market, the vehicle HUD system needs to be tested before leaving the factory, and important parameters such as the distance, distortion, and EyeBox area size of the HUD projected virtual image need to be tested to determine whether it is qualified.

[0004] The vehicle-mounted HUD system consists of several major parts, including the car windshield, HUD optical machine, control module, etc. If there are processing or design defects in any part, it may affect the user's safety and experience. For this reason, it is necessary to conduct a comprehensive test on the projection effect of the vehicle-mounted HUD system to ensure that all functional indicators of the vehicle-mounted HUD can meet the design requirements and usage requirements before it is put on the market. At present, domestic and foreign researchers have conducted a lot of research on the imaging quality of HUD virtual images. For example, the application with application publication number CN108132156A discloses a full-function detection method for automotive HUD, which provides a device and method for detecting parameters such as the position, viewing angle, and screen backlight brightness of the HUD projection virtual image. However, there is still a lack of detection of important parameters such as virtual image size, distortion, and ghosting. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiency that the HUD detection method in the prior art cannot comprehensively and accurately detect the HUD projection effect, and to provide a method and device for detecting the projection effect of a vehicle-mounted HUD system.

[0006] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0007] A method for detecting projection effect of a vehicle-mounted HUD system comprises the following steps:

[0008] S1, place a calibration plate in front of the vehicle, calibrate the human eye simulation device, and establish the conversion relationship between the calibration plate coordinate system and the vehicle coordinate system;

[0009] S2, adjusting the position of the human eye simulation device, measuring the visible area and the best visible area of ​​the HUD; the visible area is a set of positions where at least one camera in the human eye simulation device can completely capture the virtual image of the HUD projection; the best visible area is a set of positions where all cameras of the human eye simulation device can completely capture the virtual image of the HUD projection; the maximum inscribed rectangular area of ​​the best visible area is recorded as the EyeBox area of ​​the HUD system to be tested;

[0010] S3, using a human eye simulation device to capture the calibration plate image and the HUD projected virtual image at each viewpoint in the EyeBox area, and calculate the virtual image imaging distance;

[0011] According to the preset ROI area, obtain the ROI area of ​​the HUD projected virtual image captured by the human eye simulation device at each viewpoint in the EyeBox area, and calculate the contrast;

[0012] When the human eye simulation device is at each viewpoint in the EyeBox area, the average value of all corner point distortions of the HUD projected virtual image collected is recorded as the HUD system distortion of each viewpoint in the EyeBox area; wherein the corner point distortion is calculated as follows: a projection plane is set according to the virtual image imaging distance, and the corner points in the projected virtual image are mapped on the projection plane and recorded as projection points; the distance between the projection point and the theoretical position point of the corner point on the projection plane is recorded as the corner point distortion;

[0013] Adjust the brightness value output by the HUD system, and collect HUD projected virtual images of the human eye simulation device at the same position in the EyeBox area when the HUD system outputs different brightness values; calculate the brightness uniformity of the HUD projected virtual image at different brightness values ​​of the HUD system; calculate the average brightness of the HUD projected virtual image when the brightness value output by the HUD system is the minimum and the brightness value is the maximum, and obtain the brightness adjustment range of the HUD projected virtual image; determine whether there is ghosting at each point of the HUD projected virtual image at different brightnesses according to the adaptive threshold, and if there is ghosting, cluster it according to the change of pixel grayscale value to obtain the range of the ghosting of each point on the image;

[0014] Adjust the output brightness value of the HUD system to the maximum value, collect the HUD projected virtual image and calibration plate image of the human eye simulation device at the center of the EyeBox area, and calculate the field of view FOV based on the four boundary corner points of the maximum inscribed rectangle of the HUD projected virtual image.

[0015] Preferably, the implementation method of step S2 is as follows: a binocular camera is used as a human eye simulation device, a collaborative robot is used to adjust the position of the binocular camera, and the image projected by the HUD system is set to a checkerboard image.

[0016] Preferably, the step S3 calculates the virtual image imaging distance by the following steps:

[0017] Calibrate the external parameters of the human eye simulation device according to the calibration plate images captured by the human eye simulation device with known parameters at multiple viewpoints in the EyeBox area;

[0018] The corner point extraction algorithm is used to obtain the coordinates of the corner points in the HUD projected virtual image image taken by each camera of the human eye simulation device, and the corresponding corner points in the HUD projected virtual image image taken by different cameras of the human eye simulation device at the same viewpoint are matched; and the spatial three-dimensional coordinates of each corner point are calculated;

[0019] The average value of the depth information of each corner point of the HUD projected virtual image of each viewpoint in the EyeBox area is calculated and recorded as the virtual image imaging distance.

[0020] Preferably, the step S3 calculates the contrast by the following steps:

[0021] Use the human eye simulation device to obtain the HUD projected virtual image in the EyeBox area, select multiple local areas of the HUD projected virtual image according to the preset ROI area, and calculate the local area contrast in each local area using the following formula:

[0022]

[0023] in: Represents the grayscale difference between adjacent pixels; The grayscale difference between adjacent pixels is The pixel distribution probability when ;

[0024] The average value of the contrast of multiple local areas is the contrast of the viewpoint.

[0025] Preferably, the step S3 calculates the distortion by the following steps:

[0026] The HUD projected virtual image collected by the human eye simulation device is preprocessed, and then the projection points of each corner point of the HUD projected virtual image image mapped on the projection plane are obtained respectively; the projection plane is a plane perpendicular to the Y axis and has the same depth as the HUD virtual image imaging distance; all projection points form a mapping area, and the maximum inscribed rectangular area of ​​the mapping area is obtained. The maximum inscribed rectangular area is sampled and quantized according to the resolution of the HUD input image, and the coordinates of the projection point are calculated; the distortion is calculated according to the coordinates of the input image and the coordinates of the projection point; the projection point P'(x', y') of the input image after projection mapping, the distortion λ at point P is expressed as:

[0027]

[0028] Where (x 0 ,y 0 ) are the coordinates of the center point of the image.

[0029] Then the HUD distortion observed at the viewpoint of the camera image is expressed as:

[0030]

[0031] Among them, n is the number of feature points, which is used to calculate the average distortion in the EyeBox area.

[0032] Preferably, the step S3 calculates and detects the brightness uniformity of the HUD projected virtual image through the following steps:

[0033] The brightness value output by the HUD system is adjusted from the minimum to the maximum, and the HUD projection virtual image is captured at the same position of the EyeBox area by a binocular camera when the HUD system outputs different brightness values;

[0034] The brightness uniformity of the projected virtual image is calculated when the HUD system outputs different brightness values. The brightness uniformity at a certain brightness value is calculated by sequentially counting the number of HUD projected virtual images captured in multiple custom areas (e.g. Figure 3 The grayscale average value of the ROI area in the image is calculated by comparing the grayscale average values ​​of multiple custom areas to obtain the maximum grayscale average value I max and the minimum grayscale average value I min , the ratio of the minimum grayscale average to the maximum grayscale average I min / I max It is the numerical value of brightness uniformity.

[0035] Preferably, the step S3 calculates and detects the brightness range of the HUD projected virtual image by the following steps:

[0036] Calculate the average grayscale value of all custom areas when the HUD system outputs the minimum brightness value, and record it as the minimum brightness And the average of the grayscale averages of all custom areas when the brightness value is maximum is recorded as the maximum brightness β; the HUD brightness adjustment range [α, β] is obtained.

[0037] Preferably, the step S3 detects ghosting by the following steps:

[0038] Analyze the change of gray value of each point in the HUD projected virtual image when the HUD system outputs different brightness values, obtain the adaptive threshold according to the change of image gray value, and process the gray image according to the adaptive threshold to determine whether there is ghosting at each point. If there is ghosting, cluster it according to the change of pixel gray value, obtain the range of ghosting of each point on the image, and calculate the ghosting size by the following formula:

[0039] G = arctan(h / f)

[0040] Where h is the diameter of the ghost range fitting circle, and f is the focal length of the camera.

[0041] Preferably, the step S3 calculates and measures the field of view angle FOV by the following steps:

[0042] Adjust the HUD system output brightness value to the maximum, and capture the HUD projection image and calibration plate image at the center of the EyeBox area using a human eye simulation device;

[0043] The human eye simulation device is calibrated using the calibration plate image to obtain the external parameters of the human eye simulation device; the HUD projection virtual image images obtained by each camera of the human eye simulation device are preprocessed respectively to obtain the four boundary corner points of the maximum inscribed rectangle of the HUD projection virtual image, and the four corner points are mapped to the projection plane, where the projection plane is a plane with the same depth as the HUD virtual image imaging distance, and the spatial three-dimensional coordinates of the boundary corner points are obtained, and the horizontal field of view angle, vertical field of view angle and diagonal field of view angle are calculated according to the spatial three-dimensional coordinates of the viewpoint at the center position of the EyeBox area and the boundary corner points.

[0044] A projection effect detection device for a vehicle-mounted HUD system comprises a calibration plate, a human eye simulation device, a position adjustment device and a controller; the output end of the human eye simulation device is connected to the input end of the controller, and the output end of the controller is connected to the input end of the position adjustment device; the controller is used to control the position adjustment device to adjust the position of the human eye simulation device according to the above method, and to collect images taken by the human eye simulation device to detect the projection effect.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention provides a method and device for detecting the projection effect of a vehicle-mounted HUD system, which can detect various indicators of the projection effect of the vehicle-mounted HUD system with high efficiency and high accuracy through comprehensive detection and analysis of the visible area, virtual image imaging distance, contrast, distortion, brightness uniformity, brightness adjustment range, ghosting and field of view FOV of the HUD system. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of a method for detecting projection effect of a vehicle-mounted HUD system according to exemplary embodiment 1 of the present invention;

[0047] Figure 2 is a schematic diagram of a checkerboard image according to exemplary embodiment 1 of the present invention;

[0048] Figure 3 Schematic diagram of the ROI region of exemplary embodiment 1 of the present invention;

[0049] Figure 4It is a schematic projection plane diagram of exemplary embodiment 1 of the present invention;

[0050] Figure 5 Schematic diagram of the field of view FOV of exemplary embodiment 1 of the present invention. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below in conjunction with test examples and specific implementation methods. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0052] Example 1

[0053] like Figure 1 As shown, this embodiment provides a method for detecting projection effect of a vehicle-mounted HUD system, comprising the following steps:

[0054] S1, place a calibration plate in front of the vehicle, calibrate the human eye simulation device, and establish the conversion relationship between the calibration plate coordinate system and the vehicle coordinate system;

[0055] S2, adjusting the position of the human eye simulation device, measuring the visible area and the best visible area of ​​the HUD; the visible area is a set of positions where at least one camera in the human eye simulation device can completely capture the virtual image of the HUD projection; the best visible area is a set of positions where all cameras of the human eye simulation device can completely capture the virtual image of the HUD projection; the maximum inscribed rectangular area of ​​the best visible area is recorded as the EyeBox area of ​​the HUD system to be tested;

[0056] S3, using a human eye simulation device to capture the calibration plate image and the HUD projected virtual image at each viewpoint in the EyeBox area, and calculate the virtual image imaging distance;

[0057] According to the preset ROI area, obtain the ROI area of ​​the HUD projected virtual image captured by the human eye simulation device at each viewpoint in the EyeBox area, and calculate the contrast;

[0058] Use a human eye simulation device to collect HUD projected virtual image images at each viewpoint in the EyeBox area, calculate the average value of corner distortion of all corner points in the virtual image, and record it as the HUD projected image distortion of each viewpoint in the EyeBox area; the calculation method of corner distortion is as follows: set a projection plane according to the virtual image imaging distance, map the corner points in the projected virtual image onto the projection plane, and record them as projection points; the distance between the projection point and the theoretical position point of the corner point on the projection plane is recorded as corner distortion;

[0059] Adjust the brightness value output by the HUD system, and collect HUD projected virtual images of the human eye simulation device at the same position in the EyeBox area when the HUD system outputs different brightness values; calculate the brightness uniformity of the HUD projected virtual image at different brightness values ​​of the HUD system; calculate the average brightness of the HUD projected virtual image when the brightness value output by the HUD system is the minimum and the brightness value is the maximum, and obtain the brightness adjustment range of the HUD projected virtual image; determine whether there is ghosting at each point of the HUD projected virtual image at different brightnesses according to the adaptive threshold, and if there is ghosting, cluster it according to the change of pixel grayscale value to obtain the range of the ghosting of each point on the image;

[0060] Adjust the output brightness value of the HUD system to the maximum value, collect the HUD projected virtual image and calibration plate image of the human eye simulation device at the center of the EyeBox area, and calculate the field of view FOV based on the four boundary corner points of the largest inscribed rectangle in the area where the corner points of the HUD projected virtual image are located.

[0061] According to the four boundary corner points of the largest inscribed rectangle in the area where each corner point of the HUD projection virtual image is located, the spatial position of the rectangle in the projection plane can be calculated using the camera parameters, which can represent the size of the virtual image.

[0062] Through comprehensive testing and analysis of the HUD system's visible area, virtual image distance, contrast, distortion, brightness uniformity, brightness adjustment range, ghosting, virtual image size, and field of view FOV, we can efficiently and accurately test various indicators of the vehicle-mounted HUD system's projection effect, which is convenient for the optimal design of the vehicle-mounted HUD system.

[0063] Specifically, the implementation method of step S2 is as follows: a binocular camera is used as a human eye simulation device, a mechanical arm or a collaborative robot is used to adjust the position of the binocular camera, and the HUD projection image is set to a checkerboard image (such as Figure 2As shown, the pattern rules of the chessboard image are convenient for calculation). Extend the robotic arm or collaborative robot into the vehicle where the HUD product to be tested is installed, and move the robotic arm or collaborative robot from left to right, from top to bottom, and from front to back in sequence with a set step length. After each movement, use the binocular camera to shoot the calibration plate in front of the vehicle, and then use the binocular camera to shoot the HUD projection virtual image, that is, the virtual image of the chessboard image projected by the HUD, to obtain the HUD projection virtual image image. If the left and right cameras of the binocular camera can completely shoot the rectangular frame in the chessboard image, the viewpoint at the location of the optical center of the binocular camera is in the best visible area of ​​the HUD product, and the above image is stored. If any of the left and right cameras of the binocular camera can completely shoot the rectangular frame in the chessboard image, the viewpoint at the location of the optical center of the binocular camera is in the visible area of ​​the HUD product, and the above image is stored. The largest inscribed rectangular area of ​​the best visible area is the EyeBox area of ​​the HUD product to be tested. The EyeBox area of ​​the vehicle-mounted HUD represents the eye activity area where the driver can fully view the HUD projection image.

[0064] Specifically, step S3 calculates the virtual image imaging distance through the following steps:

[0065] Calibrate the external parameters of the human eye simulation device according to the calibration plate images captured by the human eye simulation device with known parameters at multiple viewpoints in the EyeBox area;

[0066] The corner point extraction algorithm is used to obtain the coordinates of the corner points in the HUD projected virtual image image taken by each camera of the human eye simulation device, and the corresponding corner points in the HUD projected virtual image image taken by different cameras of the human eye simulation device at the same viewpoint are matched; and the spatial three-dimensional coordinates of each corner point are calculated;

[0067] Calculate the average value of the depth information of each corner point of the HUD projected virtual image at each viewpoint in the EyeBox area, and record it as the virtual image imaging distance (VID). The depth information is the value of the corner point in the Y-axis direction, and the coordinate system takes the front of the vehicle as the positive direction of the Y-axis.

[0068] Exemplarily, the calibration plate image collected by the binocular camera with known camera parameters (intrinsic parameters and distortion coefficients) is used to calculate the calibration camera external parameters using the Zhang Zhengyou calibration method. The external parameters are used to represent the relative position relationship between the camera coordinate system and the world coordinate system (vehicle coordinate system). The coordinates of the checkerboard corner points in the HUD projection virtual image image respectively taken by the left and right cameras of the binocular camera are obtained by the corner point extraction algorithm, and the corresponding corner points on the views of the left and right cameras are matched by binocular matching. According to the matching corner point pairs obtained above, the spatial three-dimensional coordinates of each corner point are calculated using the binocular parallax principle. According to the above steps, the depth information of each corner point of the HUD projection virtual image image at each viewpoint in the EyeBox area can be obtained, and the average value can be calculated as the HUD projection virtual image imaging distance. In addition, the spatial three-dimensional coordinates of the corner points obtained at each viewpoint are surface fitted to obtain the HUD projection virtual image imaging surface at each viewpoint, and the HUD projection virtual image imaging distance can be intuitively viewed.

[0069] Exemplarily, step S3 calculates the contrast by the following steps:

[0070] Use a binocular camera to obtain the HUD projected virtual image in the EyeBox area, select multiple local areas of the HUD projected virtual image according to the preset ROI area, and calculate the local area contrast in each local area using the following formula:

[0071]

[0072] in: Represents the grayscale difference between adjacent pixels; The grayscale difference between adjacent pixels is The pixel distribution probability when ;

[0073] The average value of the contrast of multiple local areas is the contrast of the viewpoint. Specifically, the preset ROI area can be Figure 3 The area shown in the dotted box.

[0074] Exemplarily, step S3 calculates the distortion by the following steps:

[0075] Use a binocular camera to capture the HUD projection virtual image (the chessboard virtual image of the HUD projection) at each viewpoint in the EyeBox area, recorded as the HUD projection virtual image image, and perform the following steps:

[0076] The HUD projected virtual image captured by the binocular camera is subjected to preprocessing operations such as cropping and binarization, and then the projection points of each corner point of the HUD projected virtual image of the left and right views are mapped to the projection plane respectively; the projection plane is a plane perpendicular to the Y axis and has the same depth as the HUD virtual image imaging distance; all projection points form a mapping area, and the maximum inscribed rectangular area of ​​the mapping area is obtained. The maximum inscribed rectangular area is sampled and quantized according to the resolution of the HUD input image, and the coordinates of the projection point are calculated; the distortion is calculated according to the coordinates of the input image and the coordinates of the projection point; the point P (x, y) of the input image, the projection point P' (x', y') after projection mapping, the distortion λ at point P is expressed as:

[0077]

[0078] Where (x 0 ,y 0 ) are the coordinates of the center point of the image.

[0079] Then the HUD distortion observed at the viewpoint of the camera image is expressed as:

[0080]

[0081] Where n is the number of feature points, which can be used to calculate the average distortion of the HUD projected virtual image at each viewpoint in the EyeBox area.

[0082] like Figure 4 As shown in the figure, there is a point P(x, y) in the HUD input image. Point P is projected by the HUD and mapped to the projection plane to obtain point P'; the maximum inscribed rectangle R of the area formed by the combination of all the points in the original input image after mapping and projection is obtained; the maximum inscribed rectangle R is sampled and quantized according to the resolution of the original input image, and the coordinates of point P' (x', y') that are consistent with the coordinates of point P are calculated. Then the distortion of the point can be calculated in the same coordinate system, and the distortion of the image can be calculated by statistical methods.

[0083] Exemplarily, step S3 calculates and detects the brightness adjustment range and brightness uniformity of the HUD projected virtual image through the following steps:

[0084] The brightness value output by the HUD system is adjusted from the minimum to the maximum, and the HUD projection virtual image is captured at the same position of the EyeBox area by the binocular camera when the HUD system outputs different brightness values. Since the HUD projection virtual image captured at the center position has the best effect, this application captures the HUD projection virtual image at the center position for the analysis of the brightness adjustment range and brightness uniformity.

[0085] The brightness uniformity of the projected virtual image is calculated when the HUD system outputs different brightness values. The brightness uniformity at a certain brightness value is calculated by sequentially counting the number of HUD projected virtual images captured in multiple custom areas (e.g. Figure 3 The grayscale average value of the ROI area in the image is calculated by comparing the grayscale average values ​​of multiple custom areas to obtain the maximum grayscale average value I max and the minimum grayscale average value I min , the ratio of the minimum grayscale average to the maximum grayscale average I min / I max The brightness uniformity value is the value of brightness uniformity. The brightness uniformity of the HUD system under each brightness setting is obtained in turn. The larger the value of brightness uniformity, the more uniform the brightness of the HUD output image.

[0086] The analysis process of the brightness adjustment range is as follows: Calculate the average grayscale average value of all custom areas when the HUD system outputs the minimum brightness value, and record it as the minimum brightness The mean of the grayscale averages of all custom areas when the brightness value is maximum is recorded as the maximum brightness β, and the HUD brightness adjustment range [α, β] is obtained.

[0087] Exemplarily, step S3 detects ghosting by the following steps:

[0088] The brightness value output by the HUD system is adjusted from the minimum to the maximum, and the HUD projection virtual image is captured at the same position of the EyeBox area by a binocular camera when the HUD system outputs different brightness values. The gray value change of each point in each HUD projection virtual image is analyzed, and an adaptive threshold is obtained according to the gray value change of the image. The gray image is processed according to the adaptive threshold to determine whether there is ghosting at each point. If there is ghosting, clustering is performed according to the change of pixel gray value, and the range of the ghost of each point on the image is obtained. The ghost size is calculated by the following formula:

[0089] G = arctan(h / f)

[0090] Where h is the diameter of the ghost range fitting circle, and f is the focal length of the camera.

[0091] Exemplarily, step S3 calculates the measured field of view angle FOV by the following steps:

[0092] First, adjust the HUD system output brightness value to the maximum, and use the human eye simulation device to capture the HUD projection image and the calibration plate image at the center of the EyeBox area;

[0093] Then, the human eye simulation device is calibrated using the calibration plate image to obtain the external parameters of the human eye simulation device; the HUD projection virtual image images obtained by each camera of the human eye simulation device are preprocessed respectively to obtain the four boundary corner points of the maximum inscribed rectangle of the HUD projection virtual image, and the four corner points are mapped to the projection plane (a plane with the same depth as the HUD virtual image imaging distance) to obtain the spatial three-dimensional coordinates of the boundary corner points, and the horizontal field of view angle, vertical field of view angle and diagonal field of view angle are calculated according to the spatial three-dimensional coordinates of the viewpoint at the center position of the EyeBox area and the boundary corner points.

[0094] FOV (Field of View) refers to the angle between the edge of the part observable by the human eye and the line connecting the center of the human pupil in the virtual image formed by the HUD projection, including the horizontal field of view angle, the vertical field of view angle, and the diagonal field of view angle. The larger the field of view angle, the stronger the sense of immersion brought by the HUD device. This embodiment captures the image at the center of the EyeBox area to clearly capture the HUD projected virtual image, and then pre-processes the HUD projected virtual image through image processing operations such as grayscale to obtain the boundary corner points and calculate the field of view angle. The calculation process of the field of view angle is as follows:

[0095] Assuming the above, the four boundary corner points of the HUD projected virtual image observed at the viewpoint E are arranged in a clockwise direction with the point in the upper left corner as the starting point, and are recorded as points A (Xa, Y0, Za), B (Xb, Y0, Zb), C (Xc, Y0, Zc), and D (Xd, Y0, Zd); since the four corner points are all mapped to the projection plane, that is, their Y-axis coordinate values ​​are the same, which are all the HUD virtual image imaging distance value Y0; the position of the binocular camera of the human eye simulation device is E (Xe, Ye, Ze);

[0096] Then the horizontal field of view FOVw can be expressed as:

[0097] FOVw=arccos([(Xa-Xe)(Xb-Xe)+(Ya-Ye)(Yb-Ye)] / |EA||EB|)

[0098] Where EA=(Xa-Xe, Ya-Ye), EB=(Xb-Xe, Yb-Ye);

[0099] The vertical field of view FOVh can be expressed as:

[0100] FOVh=arccos([(Xa-Xe)(Xd-Xe)+(Ya-Ye)(Yd-Ye)] / |EA||ED|)

[0101] Where ED = (Xd-Xe, Yd-Ye);

[0102] The diagonal field of view FOVv can be expressed as:

[0103] FOVv=arccos([(Xa-Xe)(Xc-Xe)+(Ya-Ye)(Yc-Ye)] / |EA||EC|)

[0104] Where EC = (Xc-Xe, Yc-Ye).

[0105] Through comprehensive testing and analysis of the HUD system's visible area, virtual image distance, contrast, distortion, brightness uniformity, brightness adjustment range, ghosting, and field of view FOV, we can efficiently and accurately test various indicators of the vehicle-mounted HUD system's projection effect.

[0106] Example 2

[0107] The present embodiment provides a projection effect detection device for a vehicle-mounted HUD system, including a calibration plate, a human eye simulation device, a position adjustment device and a controller; the output end of the human eye simulation device is connected to the input end of the controller, and the output end of the controller is connected to the input end of the position adjustment device; the controller is used to control the position adjustment device to adjust the position of the human eye simulation device according to the method described in Example 1, and to collect images taken by the human eye simulation device to detect the projection effect.

[0108] Specifically, the camera used in the human eye simulation device of this embodiment uses a high-resolution fixed-focus binocular camera to better simulate the visual effect of the human eye, and the distance between the left and right cameras of the binocular camera is 61mm. The mechanical arm used in the position adjustment device fixes the camera on the mechanical arm through a camera bracket, which is convenient for simulating the situation of human eyes watching HUD projection at different positions; in order to adjust the position more accurately, this embodiment uses a six-axis high-precision collaborative robot.

[0109] According to the vehicle model used by the HUD system to be tested, the vehicle-mounted HUD system projection effect detection device is installed, and the position and posture of the mechanical arm are initialized. At the same time, a calibration plate is placed in front of the vehicle equipped with the HUD system to be tested, and a conversion relationship between the calibration plate coordinate system and the vehicle coordinate system is established. The controller controls the position adjustment device to adjust the position of the human eye simulation device according to the method described in Example 1, and collects images taken by the human eye simulation device to detect the projection effect.

[0110] The above is only a detailed description of the specific implementation of the present invention, rather than a limitation of the present invention. Various substitutions, modifications and improvements made by those skilled in the relevant art without departing from the principle and scope of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting projection effect of a vehicle-mounted HUD system, characterized in that: The following steps are involved: S1, place a calibration plate in front of the vehicle, calibrate the human eye simulation device, and establish the conversion relationship between the calibration plate coordinate system and the vehicle coordinate system; S2, adjusting the position of the human eye simulation device, measuring the visible area and the best visible area of ​​the HUD; the visible area is a set of positions where at least one camera in the human eye simulation device can completely capture the virtual image of the HUD projection; the best visible area is a set of positions where all cameras of the human eye simulation device can completely capture the virtual image of the HUD projection; the maximum inscribed rectangular area of ​​the best visible area is recorded as the EyeBox area of ​​the HUD system to be tested; S3, using a human eye simulation device to capture the calibration plate image and the HUD projected virtual image at each viewpoint in the EyeBox area, and calculate the virtual image imaging distance; According to the preset ROI area, obtain the ROI area of ​​the HUD projected virtual image captured by the human eye simulation device at each viewpoint in the EyeBox area, and calculate the contrast; When the human eye simulation device is at each viewpoint in the EyeBox area, the average value of all corner point distortions of the HUD projected virtual image collected is recorded as the HUD system distortion of each viewpoint in the EyeBox area; wherein the corner point distortion is calculated as follows: a projection plane is set according to the virtual image imaging distance, and the corner points in the projected virtual image are mapped on the projection plane and recorded as projection points; the distance between the projection point and the theoretical position point of the corner point on the projection plane is recorded as the corner point distortion; Adjust the brightness value output by the HUD system, and collect HUD projected virtual images of the human eye simulation device at the same position in the EyeBox area when the HUD system outputs different brightness values; calculate the brightness uniformity of the HUD projected virtual image at different brightness values ​​of the HUD system; calculate the average brightness of the HUD projected virtual image when the brightness value output by the HUD system is the minimum and the brightness value is the maximum, and obtain the brightness adjustment range of the HUD projected virtual image; determine whether there is ghosting at each point of the HUD projected virtual image at different brightnesses according to the adaptive threshold, and if there is ghosting, cluster it according to the change of pixel grayscale value to obtain the range of the ghosting of each point on the image; Adjust the output brightness value of the HUD system to the maximum value, collect the HUD projected virtual image and calibration plate image of the human eye simulation device at the center of the EyeBox area, and calculate the field of view angle FOV according to the four boundary corner points of the maximum inscribed rectangle of the HUD projected virtual image; The four boundary corner points of the HUD projected virtual image observed at viewpoint E are arranged in a clockwise direction starting from the point in the upper left corner, and are recorded as points A (Xa, Ya, Za), B (Xb, Yb, Zb), C (Xc, Yc, Zc), and D (Xd, Yd, Zd); Since the four corner points are all mapped to the projection plane, their Y-axis coordinate values ​​are the same, which are all the HUD virtual image imaging distance value Y0; the position of the binocular camera of the human eye simulation device is E (Xe, Ye, Ze); Then the horizontal field of view FOVw can be expressed as: FOVw=arccos([(Xa-Xe)(Xb-Xe)+(Ya-Ye)(Yb-Ye)] / |EA||EB|); Where EA=(Xa-Xe, Ya-Ye), EB=(Xb-Xe, Yb-Ye); The vertical field of view FOVh can be expressed as: FOVh=arccos([(Xa-Xe)(Xd-Xe)+(Ya-Ye)(Yd-Ye)] / |EA||ED|) Where ED = (Xd-Xe, Yd-Ye); The diagonal field of view FOVv can be expressed as: FOVv=arccos([(Xa-Xe)(Xc-Xe)+(Ya-Ye)(Yc-Ye)] / |EA||EC|) Where EC = (Xc-Xe, Yc-Ye); According to the four boundary corner points of the largest inscribed rectangle in the area where each corner point of the HUD projected virtual image is located, the spatial position of the rectangle in the projection plane can be calculated using the camera parameters, which can represent the size of the virtual image.

2. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The implementation method of step S2 is as follows: a binocular camera is used as a human eye simulation device, a collaborative robot is used to adjust the position of the binocular camera, and the image projected by the HUD system is set to a checkerboard image.

3. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates the virtual image imaging distance by the following steps: Calibrate the external parameters of the human eye simulation device according to the calibration plate images captured by the human eye simulation device with known parameters at multiple viewpoints in the EyeBox area; The corner point extraction algorithm is used to obtain the coordinates of the corner points in the HUD projected virtual image image taken by each camera of the human eye simulation device, and the corresponding corner points in the HUD projected virtual image image taken by different cameras of the human eye simulation device at the same viewpoint are matched; and the spatial three-dimensional coordinates of each corner point are calculated; The average value of the depth information of each corner point of the HUD projected virtual image of each viewpoint in the EyeBox area is calculated and recorded as the virtual image imaging distance.

4. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates the contrast by the following steps: Use the human eye simulation device to obtain the HUD projected virtual image in the EyeBox area, select multiple local areas of the HUD projected virtual image according to the preset ROI area, and calculate the local area contrast in each local area using the following formula: in: Represents the grayscale difference between adjacent pixels; The grayscale difference between adjacent pixels is The pixel distribution probability when ; The average value of the contrast of multiple local areas is the contrast of the viewpoint.

5. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates the distortion by the following steps: The HUD projected virtual image collected by the human eye simulation device is preprocessed, and then the projection points of each corner point of the HUD projected virtual image image mapped on the projection plane are obtained respectively; the projection plane is a plane perpendicular to the Y axis and has the same depth as the HUD virtual image imaging distance; all projection points form a mapping area, and the maximum inscribed rectangular area of ​​the mapping area is obtained. The maximum inscribed rectangular area is sampled and quantized according to the resolution of the HUD input image, and the coordinates of the projection point are calculated; the distortion is calculated according to the coordinates of the input image and the coordinates of the projection point; the projection point P'(x', y') of the input image after projection mapping, the distortion λ at point P is expressed as: Where (x0, y0) is the coordinate of the center point of the image; Then the HUD distortion observed at the viewpoint of the camera image is expressed as: Among them, n is the number of feature points, which is used to calculate the average distortion in the EyeBox area.

6. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates and detects the brightness uniformity of the HUD projected virtual image through the following steps: The brightness value output by the HUD system is adjusted from the minimum to the maximum, and the HUD projection virtual image is captured at the same position of the EyeBox area by a binocular camera when the HUD system outputs different brightness values; The brightness uniformity of the projected virtual image when the HUD system outputs different brightness values ​​is calculated respectively; the brightness uniformity at a certain brightness value is calculated by sequentially counting the grayscale average values ​​of the captured HUD projected virtual image in multiple custom areas, comparing the grayscale average values ​​of multiple custom areas, and obtaining the maximum grayscale average value I max and the minimum grayscale average value I min , the ratio of the minimum grayscale average to the maximum grayscale average I min / I max It is the numerical value of brightness uniformity.

7. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates and detects the brightness range of the HUD projected virtual image through the following steps: Calculate the average of the grayscale averages of all custom areas when the HUD system outputs the minimum brightness value, recorded as the minimum brightness α; and the average of the grayscale averages of all custom areas when the brightness value is the maximum, recorded as the maximum brightness β; Get the HUD brightness adjustment range α, β.

8. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 detects ghosting by the following steps: Analyze the gray value changes of each point in the HUD projected virtual image when the HUD system outputs different brightness values, obtain an adaptive threshold according to the image gray change, and process the gray image according to the adaptive threshold to determine whether there is a ghost at each point. If there is a ghost, cluster it according to the change of pixel gray value, obtain the range of the ghost of each point on the image, and calculate the ghost size K by the following formula: K = arctanh / f Where h is the diameter of the ghost range fitting circle, and f is the focal length of the camera.

9. The method for detecting projection effect of a vehicle-mounted HUD system according to claim 1, characterized in that: The step S3 calculates and measures the field of view FOV by the following steps: Adjust the HUD system output brightness value to the maximum, and capture the HUD projection image and calibration plate image at the center of the EyeBox area using a human eye simulation device; The human eye simulation device is calibrated using the calibration plate image to obtain the external parameters of the human eye simulation device; the HUD projection virtual image images obtained by each camera of the human eye simulation device are preprocessed respectively to obtain the four boundary corner points of the maximum inscribed rectangle of the HUD projection virtual image, and the four corner points are mapped to the projection plane, where the projection plane is a plane with the same depth as the HUD virtual image imaging distance, and the spatial three-dimensional coordinates of the boundary corner points are obtained, and the horizontal field of view angle, vertical field of view angle and diagonal field of view angle are calculated according to the spatial three-dimensional coordinates of the viewpoint at the center position of the EyeBox area and the boundary corner points.

10. A vehicle-mounted HUD system projection effect detection device, characterized in that: It includes a calibration board, a human eye simulation device, a position adjustment device and a controller; the output end of the human eye simulation device is connected to the input end of the controller, and the output end of the controller is connected to the input end of the position adjustment device; The controller is used to control the position adjustment device to adjust the position of the human eye simulation device according to the method described in any one of claims 1 to 9, and to collect images taken by the human eye simulation device to detect the projection effect.

Images