VR / AR glasses optical adjustment and detection method

Abstract

The invention discloses an optical adjustment and detection method for VR / AR glasses. The method comprises the following steps: 1) adjustment; (11) focusing; 12) adjusting the posture; 2) detecting; 21) performing distortion detection; 22) carrying out MTF detection; 23) detecting the uniformity of dirt and brightness; and 24) artifact detection. And 3) recording and judging a result. According to the method, the visual guidance technology and the optical performance real-time feedback are combined, the camera is used for capturing the image for analysis, the position and the inclination angle of the VR / AR lens group can be accurately measured, and accurate positioning of the optical element in the assembling process is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of optical alignment, and in particular to an optical alignment and detection method for VR / AR glasses. Background Art

[0002] With the progress of optical design and optical processing technologies, advanced optical systems have made remarkable developments in multiple fields. Whether it is a telescope in the aerospace field or short-focus projection, virtual reality (VR), and augmented reality (AR) in daily life, they all show extensive applications and developments. These advanced optical systems have put forward higher installation accuracy requirements for the position and attitude of optical devices, making precision alignment a key link in the production of optical systems. Currently, the evaluation of alignment performance mainly relies on human eye vision, lacking a unified evaluation standard, and relevant research is still relatively scarce.

[0003] Most current optical system alignment means focus on evaluating image quality through an interferometer, and the alignment process is guided by wavefront aberration theory. Although wavefront aberration can obtain the optical performance of an optical system, the interference method is limited by the field of view of the interferometer. In addition, the price of the interferometer is also relatively expensive. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, it is desirable to provide an optical alignment and detection method for VR / AR glasses, which combines vision guidance technology with real-time feedback of optical performance, uses a camera to capture and analyze images, can accurately measure the position and tilt angle of the VR / AR lens group, and ensure the accurate positioning of optical elements during the assembly process.

[0005] An optical alignment and detection method for VR / AR glasses provided by the present invention includes the following steps:

[0006] 1) Alignment;

[0007] 11) Focusing: A black-and-white stripe pattern is displayed on the screen, the camera obtains the screen image I through the VR / AR lens group, analyzes the MTF of the screen image I, moves the VR / AR lens group in the direction close to or away from the camera, and obtains the position of the VR / AR lens group corresponding to the maximum MTF to complete focusing;

[0008] 12) Pose adjustment: An alignment pattern is displayed on the screen, the camera obtains the screen image II through the VR / AR lens group, analyzes the screen image II to obtain the translation amount and rotation angle of the VR / AR lens group, and adjusts the pose of the VR / AR lens group according to the translation amount and rotation angle of the VR / AR lens group;

[0009] 13) Glue fixing: Glue is applied between the VR / AR lens groups to fix them;

[0010] 2) Detection;

[0011] 21) Distortion detection: The camera captures the distortion detection pattern displayed on the screen through the VR / AR lens group, calculates the deviation between the ideal and actual imaging positions, and evaluates the distortion;

[0012] 22) MTF detection: The camera captures the MTF stripe pattern displayed on the screen through the VR / AR lens group, selects the detection area, calculates the stripe contrast, and obtains the MTF values of different fields of view;

[0013] 23) Dirt and brightness uniformity detection: The camera captures the white field pattern displayed on the screen through the VR / AR lens group, performs frequency domain processing and gray scale analysis to detect dirt and evaluate the brightness uniformity;

[0014] 24) Artifact detection: The camera captures the artifact detection pattern displayed on the screen through the VR / AR lens group, divides the ROI, enhances the gray scale and performs threshold segmentation, and calculates the relative proportion of artifacts.

[0015] 3) Result recording and determination;

[0016] Record the detection results of various performance indicators, compare with the design requirements, and determine whether the product is qualified.

[0017] Furthermore, in the step 11), the MTF of the screen image a is analyzed by the stripe contrast method, including the following steps:

[0018] 111) Perform mean filtering along the stripe direction to make the gray scale uniform within the black and white areas of the stripe, and eliminate the errors of the camera;

[0019] 112) According to the interval of the stripe, at the center position of the stripe, find the maximum value I of the gray scale within each interval maxi and the minimum value I mini , and calculate the MTF within each interval i as:

[0020]

[0021] where i = 1, 2, ···, n, and n is the total number of intervals;

[0022] 113) Average the MTF values of each interval to obtain the final MTF value;

[0023]

[0024] Furthermore, in the step 12), the alignment pattern includes 1 group of MTF detection stripes at the center and 8 groups of MTF detection stripes evenly distributed on the outer periphery of the central MTF detection stripes; the analysis of the screen image II includes the following steps:

[0025] 121) Threshold segmentation is performed on each MTF detection stripe in the screen image II through the positioning circles;

[0026] 122) Each closed area enclosed by the positioning circles is filled to obtain nine circles;

[0027] 123) The nine circles are distinguished by connected component judgment, the circles are respectively fitted, the center of the fitted circle is obtained, and the position of the center is marked;

[0028] 124) The center circle of the pattern is located by the center coordinates, a horizontal auxiliary line and a vertical auxiliary line are established at the position of the center circle of the pattern, other circles are located according to the center coordinates of other circles, and a pattern coordinate system is established from the center coordinates of other original circles in the four directions of up, down, left, and right;

[0029] 125) Obtain the coordinates (Δ ximage , Δ yimage ) of the center of the pattern coordinate system in the camera coordinate system, and calculate the translation amount of the VR / AR lens group;

[0030]

[0031] where, x c is the translation amount of the VR / AR lens group in the x-axis direction; y c is the translation amount of the VR / AR lens group in the y-axis direction;

[0032] P x and P y are conversion coefficients;

[0033] 126) Calculate the rotation angle of the VR / AR lens group according to the MTF of the screen image II in the x and y directions;

[0034]

[0035] where, R x is the rotation angle value of the VR / AR lens group around the x-axis; R y is the rotation angle value of the VR / AR lens group around the y-axis;

[0036] α and β are adjustment coefficients;

[0037] MTFL is the MTF value at the left end of the x-axis in the screen image II; MTFR is the MTF value at the right end of the x-axis in the screen image II;

[0038] MTFU is the MTF value at the top end of the y-axis in the screen image II; MTFD is the MTF value at the bottom end of the y-axis in the screen image II.

[0039] Further, in step 21), the captured pattern is subjected to threshold segmentation, and the opening operation is performed to remove possible noise, and all the dots are extracted; circle fitting is performed on each dot to obtain the center of the fitted circle, the centers are sorted to obtain the center position of the central circle; from the adjacent points, the center position of the ideal image is calculated; the deviation between the ideal center position and the imaging center position is the current target distortion; the deviation of the outermost periphery of the camera field of view is used for distortion evaluation; at the same time, after calibrating the forward center, the deviation between the center of the camera coordinate system and the center position of the central circle is the final deviation of the VR / AR lens group.

[0040] Further, in step 22), a detection area is generated from the captured MTF fringe pattern within a set range, the MTF fringe pattern is subjected to threshold segmentation, fully dilated, the fringe patterns are connected to form a rectangle, the opening operation is performed on each area to remove noise, the MTF fringe area is obtained by screening according to the area, and the center of the circumscribed rectangle is used as the center of the fringe area, and the contrast of the fringe is read according to the offset of the MTF fringe; the calculated MTF value is compared with the nominal value of the camera lens at the current frequency, and the lens measurement error is compensated for the MTF detection result.

[0041] Further, in step 23), the captured white-field pattern is subjected to frequency-domain filtering, inverse Fourier transform, and then a threshold is set, and the corresponding defect marks are obtained by screening according to the threshold; the average gray levels of the middle area and several peripheral areas are extracted from the pattern under the white field, and the gray-level distribution state of the entire area is recorded; the brightness of the peripheral and middle areas is compared to obtain the brightness uniformity.

[0042] Further, in step 24), an ROI is divided in the artifact detection pattern, the ROI is subjected to gray-level enhancement to increase the gray-level difference, segmented by a threshold, and appropriately eroded to reduce the interference of the artifact periphery, and the ratio of the maximum gray-level data in the artifact of each area to the gray-level mean of the original image is used to obtain the relative ratio of the artifact.

[0043] Further, an auxiliary system is used to complete the assembly, adjustment and detection of the VR / AR lens group; the auxiliary system includes a six-degree-of-freedom platform, a manual linear displacement stage and a screen; the six-degree-of-freedom platform adsorbs and fixes the VR / AR lens group through a negative-pressure air valve and drives it to move along the x-axis, y-axis or z-axis direction, and rotate around the x-axis, y-axis or z-axis;

[0044] The manual linear displacement stage is arranged below the VR / AR lens group, and a camera is fixedly installed on the manual linear displacement stage and driven to move along the x-axis, y-axis or z-axis direction;

[0045] The screen is arranged above the VR / AR lens group and is used to display the patterns for assembly, adjustment or detection.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] The method of the present invention combines vision guidance technology with real-time feedback of optical performance. It uses a camera to capture and analyze images, and guides the alignment according to performance indicators such as MTF, breaking through the limitation of traditional alignment methods that rely on mechanical positioning. It can accurately measure the position and tilt angle of the VR / AR lens group, ensuring the accurate positioning of optical elements during the assembly process. The active alignment method improves the accuracy and stability of the optical system, ensures the accurate transmission of the optical path, and thus achieves higher-quality imaging results.

[0048] The method of the present invention constructs an intelligent system covering multi-dimensional performance detection such as distortion, MTF, dirt, and artifacts. By designing special patterns and algorithms, it realizes comprehensive, rapid, and accurate detection of the alignment results.

[0049] It should be understood that the content described in the Summary of the Invention section is not intended to limit the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Other features, objects, and advantages of the present invention will become more apparent by reading the detailed description of the non-limiting embodiments with reference to the following drawings:

[0051] Figure 1 is a schematic structural diagram of the auxiliary system;

[0052] Figure 2 is a schematic diagram for analyzing MTF by the contrast method;

[0053] Figure 3 In FIG., FIG. a is the alignment pattern, and FIG. b is a partially enlarged schematic diagram of the alignment pattern;

[0054] Figure 4 is a schematic diagram of the processing flow of the alignment pattern. Among them, FIG. a is threshold segmentation, FIG. b is filling the area, and FIG. c is marking the center of the circle;

[0055] Figure 5 is a schematic diagram of the alignment analysis method;

[0056] Figure 6 is a schematic diagram of the distortion detection pattern;

[0057] Figure 7 is a schematic diagram of the processing flow of the distortion detection pattern; among them, FIG. a is the captured image, FIG. b is extracting the dot, and FIG. c is obtaining the result;

[0058] Figure 8 In FIG., FIG. a is the MTF fringe pattern, and FIG. b is a partially enlarged schematic diagram of the MTF fringe pattern;

[0059] Figure 9 Figure 1 is a schematic diagram of the MTF detection process; Figure a shows the captured image, Figure b shows the selected ROI, Figure c shows the threshold segmentation, and Figure d shows the obtained MTF area.

[0060] Figure 10 is a schematic diagram of a white field pattern;

[0061] Figure 11 This is a flow chart for pattern processing for dirt and brightness uniformity detection. Figure a is a white field defect image, figure b is a magnified image of the defect, and figure c is the defect detection result.

[0062] Figure 12 This is a schematic diagram of brightness balance detection;

[0063] Figure 13 is a schematic diagram of an artifact detection pattern;

[0064] Figure 14 Schematic diagram of artifacts;

[0065] Figure 15 Schematic diagram of the artifact analysis process; Figure a shows the ROI selection, and Figure b shows the artifact area screening.

[0066] Numbers in the figure: 1. Six-degree-of-freedom platform; 2. Manual linear translation stage; 3. VR / AR lens group; 4. Camera. DETAILED DESCRIPTION

[0067] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the invention are shown in the accompanying drawings.

[0068] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0069] Please refer to Figures 1 to 15 The embodiment of the present invention provides a method for optical assembly and detection of VR / AR glasses, including the following steps:

[0070] 1) Installation and adjustment;

[0071] 11) Focusing: The screen displays a black and white stripe pattern. Camera 4 captures screen image I through VR / AR lens assembly 3. The MTF of screen image I is analyzed. VR / AR lens assembly 3 is moved toward or away from camera 4 to determine the position of VR / AR lens assembly 3 corresponding to the maximum MTF, completing focusing.

[0072] 12) Pose adjustment: The alignment pattern is displayed on the screen. The camera 4 obtains the screen image II through the VR / AR lens group 3, analyzes the screen image II to obtain the translation amount and rotation angle of the VR / AR lens group 3, and adjusts the pose of the VR / AR lens group 3 according to the translation amount and rotation angle of the VR / AR lens group 3;

[0073] 13) Glue fixation: Glue is applied between the VR / AR lens groups 3 to fix them;

[0074] 2) Detection;

[0075] 21) Distortion detection: The camera 4 takes pictures of the distortion detection pattern displayed on the screen through the VR / AR lens group 3, calculates the deviation between the ideal and actual imaging positions, and evaluates the distortion;

[0076] 22) MTF detection: The camera 4 takes pictures of the MTF fringe pattern displayed on the screen through the VR / AR lens group 3, selects the detection area, calculates the fringe contrast, and obtains the MTF values of different fields of view;

[0077] 23) Dirt and brightness uniformity detection: The camera 4 takes pictures of the white field pattern displayed on the screen through the VR / AR lens group 3, performs frequency domain processing and gray scale analysis to detect dirt and evaluate the brightness uniformity;

[0078] 24) Artifact detection: The camera 4 takes pictures of the artifact detection pattern displayed on the screen through the VR / AR lens group 3, divides the ROI, enhances the gray scale and performs threshold segmentation, and calculates the relative proportion of artifacts.

[0079] 3) Result recording and determination;

[0080] Record the detection results of various performance indicators, compare them with the design requirements, and determine whether the product is qualified.

[0081] In this embodiment, the VR / AR lens group 3 includes a left lens group and a right lens group, and the alignment of each of them is adjusted by the camera 4 respectively. Through visual guidance and image analysis, high-precision alignment between the screen and the VR / AR lens group 3 is achieved. After the alignment is completed, the left lens group and the right lens group are fixed by applying glue. The centering deviation is controlled within 1 pixel, and the field angle deviation is less than 0.01°, significantly improving the optical performance.

[0082] Moreover, comprehensive and accurate detection of various performance indicators of the optical system is carried out to ensure the consistency of product quality and reduce the defective rate. This application provides a unified evaluation standard for the alignment of the optical system, provides favorable technical support for related research, and has extremely high promotion value.

[0083] In a preferred embodiment, as Figure 2 shown, in step 11), the MTF of the screen image a is analyzed by the fringe contrast method, including the following steps:

[0084] 111) Perform mean filtering along the stripe direction to make the gray levels uniform within the black and white areas of the stripes and eliminate the errors of camera 4;

[0085] 112) According to the interval of the stripes, at the center position of the stripes, find the maximum value I of the gray level within each interval maxi and the minimum value I mini , and calculate the MTF within each interval i as:

[0086]

[0087] where i = 1, 2, ···, n, and n is the total number of intervals;

[0088] 113) Average the MTF values of each interval to obtain the final MTF value;

[0089]

[0090] In this embodiment, the primary task of alignment is still focusing. The parsing process of MTF is as Figure 2 shown. In the pattern design, black and white stripes are used to calculate the MTF of a specific frequency. The gray vertical stripe is the window for mean filtering, and the horizontal stripes with alternating dark and light gray are the range for solving the contrast. The sharpness is evaluated through the MTF value of the central field of view. Control the lifting of the VR / AR lens group 3 and find the position of the VR / AR lens group 3 corresponding to the maximum MTF to complete autofocus.

[0091] In a preferred embodiment, as shown in Figure 3 , 4 and 5, in step 12), the alignment pattern includes 1 group of MTF detection stripes at the center and 8 groups of MTF detection stripes evenly distributed on the periphery of the central MTF detection stripes; the parsing of the screen image II includes the following steps:

[0092] 121) Perform threshold segmentation on each MTF detection stripe in the screen image II through the positioning circles respectively;

[0093] 122) Fill each closed area surrounded by the positioning circles to obtain 9 circles;

[0094] 123) Distinguish the 9 circles by connected component judgment, fit the circles respectively, obtain the centers of the fitted circles, and mark the positions of the centers;

[0095] 124) Locate the center circle of the pattern from the center coordinates of the circles, establish a horizontal auxiliary line and a vertical auxiliary line at the position of the center circle of the pattern, locate the other circles according to the center coordinates of the other circles, and establish a pattern coordinate system from the center coordinates of the other circles in the four directions of up, down, left, and right;

[0096] 125) Obtain the coordinates (Δ ximage , Δ yimage ) of the center of the pattern coordinate system in the camera coordinate system, and calculate the translation amount of the VR / AR lens group;

[0097]

[0098] Among them, x c is the translation amount of the VR / AR lens group in the x-axis direction; y c is the translation amount of the VR / AR lens group in the y-axis direction;

[0099] P x and P y are conversion coefficients;

[0100] 126) Calculate the rotation angle of the VR / AR lens group according to the MTF of the screen image II in the x and y directions;

[0101]

[0102] Among them, R x is the rotation angle value of the VR / AR lens group around the x-axis; R y is the rotation angle value of the VR / AR lens group around the y-axis;

[0103] α and β are adjustment coefficients;

[0104] MTFL is the MTF value at the left end of the x-axis in the screen image II; MTFR is the MTF value at the right end of the x-axis in the screen image II;

[0105] MTFU is the MTF value at the top end of the y-axis in the screen image II; MTFD is the MTF value at the bottom end of the y-axis in the screen image II.

[0106] In this embodiment, the acquisition of the position and attitude adjustment amount of the VR / AR lens group 3 is achieved by analyzing the positioning circle of the projection pattern, and the image processing flow is as Figure 4 shown. Threshold segmentation can obtain each element (as shown in Figure 4 figure a), fill the closed area to obtain 9 circles (as shown in Figure 4 figure b), distinguish the 9 circles by connected domain judgment, respectively fit the circles, obtain the center of the fitted circle, and mark the center position (as shown in Figure 4 figure c). The alignment analysis is performed by the center of the positioning circle of the captured pattern as shown in 5, and the translation amount and rotation angle of the VR / AR lens group are calculated.

[0107] In a preferred embodiment, as Figure 6 and Figure 7As shown in the figure, in step 21), the captured pattern is subjected to threshold segmentation, and the possible noise is removed by opening operation, and all the circular dots are extracted; circle fitting is performed on each circular dot to obtain the center of the fitted circle, the centers are sorted to obtain the center position of the central circle; the center position of the ideal imaging is calculated from the adjacent points; the deviation between the ideal center position and the imaging center position is the current target distortion; the deviation at the outermost periphery of the camera's field of view is used for distortion evaluation; at the same time, after calibrating the forward center, the deviation between the center of the camera coordinate system and the center position of the central circle is the final deviation of the VR / AR lens group.

[0108] In this embodiment, the process of capturing and processing the distortion is as Figure 7 shown. By performing threshold segmentation on the captured pattern (such as Figure 7 the figure a shown), the possible noise is removed by opening operation, and all the circular dots are extracted (such as Figure 7 the figure b shown). Circle fitting is performed on each circle to obtain the center of the fitted circle (such as Figure 7 the figure c shown). The deviation between the camera center and the central circle is the final deviation of the VR / AR lens group installation.

[0109] In a preferred embodiment, as Figure 8 and Figure 9 shown, in step 22), a detection area is generated in the captured MTF fringe pattern within a set range, the MTF fringe pattern is subjected to threshold segmentation, fully dilated, the fringe patterns are connected to form a rectangle, opening operation is performed on each area to remove noise, the MTF fringe area is obtained by screening according to the area, and the center of the circumscribed rectangle is used as the center of the fringe area, and the contrast of the fringe is read according to the offset of the MTF fringe; the calculated MTF value is compared with the nominal value of the camera lens at the current frequency, and the lens measurement error is compensated for the MTF detection result.

[0110] In this embodiment, the MTF fringe pattern is captured (such as Figure 8 shown), and the MTF of different fields of view is detected. The detection process is as Figure 9 shown. A detection area is generated within a set range (such as Figure 9 the circular ring framed area shown in the figure b), to prevent detecting areas outside the lens. Subsequently, the image is subjected to threshold segmentation, fully dilated, the fringe patterns are connected to form a rectangle (such as the figure c shown in Figure 9), the MTF fringe area is obtained by screening according to the area (such as Figure 9 the figure d shown), the ratio with the nominal value is calculated, and the lens measurement error is compensated for the MTF detection result.

[0111] In a preferred embodiment, as Figure 10 、 11As shown in FIGS. 11 and 12, in step 23), the captured white-field pattern is subjected to frequency-domain filtering and inverse Fourier transform, and then a threshold is set, and corresponding defect marks are obtained by screening according to the threshold; the average gray levels of regions including the middle region and several peripheral regions are extracted from the pattern under the white field, and the gray-level distribution state of the entire region is recorded; the brightness of the peripheral and middle regions is compared to obtain the brightness uniformity.

[0112] In this embodiment, the white-field pattern mainly detects defects in the picture. There are some ripple textures in the white field. In this case, the image needs to be Fourier-transformed. Frequency-domain processing has very good advantages in defect detection with ripples and vignetting. The image processing process is as Figure 11 shown, the captured image is as Figure 11 shown in FIG. a in Figure 11 ; the specific morphology of the defect is as Figure 11 shown in FIG. b in Figure 12 . To prevent over-saturated exposure, a relatively dark shooting method is adopted, and the defect marks are as shown in FIG. c in

[0113] Figure 13 . The average gray levels of the 5 regions as 14 shown in FIG. 15 are extracted from the pattern under the white field, so as to obtain the brightness uniformity.

[0114] In this embodiment, the artifact is caused by the refraction of light in the lens due to the phenomenon of double refraction of light. The artifact has a great interference on the imaging of the VR system and will have a great impact on the image quality. The artifact is as Figure 14 shown, and the artifact usually images at the defocus position, and its shape depends on the displayed pattern.

[0115] The artifact and the background base color have similar gray levels. To extract the artifact, the ROI needs to be divided first to prevent the pattern itself from being included in the calculation. Then, the gray level of the ROI region is enhanced, and the gray-level difference is enlarged by using the power method, and is segmented by a threshold, and appropriate erosion is performed to reduce the interference of the periphery of the artifact. The ratio of the maximum gray-level data in the artifact of each region to the gray-level mean value of the original image is used to obtain the relative ratio of the artifact. The analysis process is as Figure 15 shown.

[0116] In a preferred embodiment, as Figure 1As shown in the figure, an auxiliary system is adopted to complete the assembly, adjustment and detection of the VR / AR lens group 3; the auxiliary system includes a six-degree-of-freedom platform 1, a manual linear displacement stage 2 and a screen; the six-degree-of-freedom platform 1 adsorbs and fixes the VR / AR lens group 3 through a negative pressure air valve, and drives it to move along the x-axis, y-axis or z-axis, and rotate around the x-axis, y-axis or z-axis;

[0117] The manual linear displacement stage 2 is arranged below the VR / AR lens group 3, and a camera 4 is fixedly installed on the manual linear displacement stage 2, and drives it to move along the x-axis, y-axis or z-axis;

[0118] The screen is arranged above the VR / AR lens group 3 and is used to display the patterns for assembly, adjustment or detection.

[0119] In this embodiment, for the three-direction translation and three-direction rotation degrees of freedom of the VR / AR lens group 3, a precise motion mechanism is designed to achieve the coordinated adjustment of each degree of freedom, and solve the problem of high-precision matching between the VR / AR lens group 3 and the screen under complex optical paths. The auxiliary system has sufficient degrees of freedom and flexibility of motion, can adapt to the assembly, adjustment and detection requirements of VR / AR glasses of different models and specifications, and has good versatility and expandability.

[0120] During the assembly and adjustment process, after obtaining the adjustment amounts such as the translation amount and rotation angle of the VR / AR lens group 3, it is fed back to the auxiliary system for adjustment; the automated assembly and adjustment process and real-time performance feedback greatly shorten the assembly and adjustment time, improve production efficiency, and meet the requirements of large-scale production.

[0121] In the description of this specification, terms such as "connection", "installation", "fixation" should all be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected, or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific situations.

[0122] In the description of this specification, the description of terms such as "one embodiment", "some embodiments" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0123] The above is only the preferred embodiment of this application and is not used to limit this application. For those skilled in the art, this application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included within the protection scope of this application.

Claims

1. An optical alignment and detection method for VR / AR glasses, characterized in that, It includes the following steps: 1) Assembly and alignment; 11) Focusing: The screen displays a black and white stripe pattern. The camera obtains the screen image I through the VR / AR lens group, analyzes the MTF of the screen image I, moves the VR / AR lens group in the direction of approaching or moving away from the camera, and obtains the position of the VR / AR lens group corresponding to the maximum MTF to complete the focusing; 12) Pose adjustment: The screen displays an assembly and alignment pattern. The camera obtains the screen image II through the VR / AR lens group, analyzes the screen image II to obtain the translation amount and rotation angle of the VR / AR lens group, and adjusts the pose of the VR / AR lens group according to the translation amount and rotation angle of the VR / AR lens group; 13) Glue dot fixing: Glue is applied between the VR / AR lens groups to fix them; 2) Detection; 21) Distortion detection: The camera takes a picture of the distortion detection pattern displayed on the screen through the VR / AR lens group, calculates the deviation between the ideal and actual imaging positions, and evaluates the distortion; 22) MTF detection: The camera takes a picture of the MTF stripe pattern displayed on the screen through the VR / AR lens group, selects the detection area, calculates the stripe contrast, and obtains the MTF values of different fields of view; 23) Dirt and brightness uniformity detection: The camera takes a picture of the white field pattern displayed on the screen through the VR / AR lens group, performs frequency domain processing and gray scale analysis to detect dirt and evaluate the brightness uniformity; 24) Artifact detection: The camera takes a picture of the artifact detection pattern displayed on the screen through the VR / AR lens group, divides the ROI, performs gray scale enhancement and threshold segmentation, and calculates the relative proportion of artifacts. 3) Result recording and determination; Record the detection results of various performance indicators, compare them with the design requirements, and determine whether the product is qualified.

2. The VR / AR glasses optical alignment and detection method according to claim 1, characterized in that In the step 11), the stripe contrast method is used to analyze the MTF of the screen image a, including the following steps: 111) Perform mean filtering along the stripe direction to make the gray scale uniform in the black and white areas of the stripe and eliminate the errors of the camera; 112) According to the interval of the stripes, at the center position of the stripes, find the maximum value I of the gray level within each interval maxi and the minimum value I mini , and calculate the MTF within each interval i as follows: where i = 1, 2, ···, n, and n is the total number of intervals; 113) Average the MTF values of each interval to obtain the final MTF value; 3. The VR / AR glasses optical alignment and detection method according to claim 1, wherein, In the step 12), the assembly and alignment pattern includes 1 group of MTF detection stripes located in the center and 8 groups of MTF detection stripes evenly distributed on the outer periphery of the central MTF detection stripes; the analysis of the screen image II includes the following steps: 121) Perform threshold segmentation on each MTF detection stripe in the screen image II through the positioning circle respectively; 122) Fill each closed area surrounded by each positioning circle to obtain 9 circles; 123) Distinguish the 9 circles by connected domain judgment, fit the circles respectively, obtain the center of the fitted circle, and mark the center position; 124) Locate the center circle of the pattern from the center coordinates of the circle, establish a horizontal auxiliary line and a vertical auxiliary line at the position of the center circle of the pattern, locate other circles according to the center coordinates of other circles, and establish a pattern coordinate system from the center coordinates of other circles in the four directions of up, down, left, and right; 125) Obtain the coordinates (Δx image , Δy image ) of the center of the pattern coordinate system in the camera coordinate system, and calculate the translation amount of the VR / AR lens group; where x c is the translation amount of the VR / AR lens group in the x-axis direction; y c is the translation amount of the VR / AR lens group in the y-axis direction; Px and Py are conversion coefficients; 126) Calculate the rotation angle of the VR / AR lens group according to the MTF of the screen image II in the x and y directions; Among them, R x is the rotation angle value of the VR / AR lens group around the x-axis; R y is the rotation angle value of the VR / AR lens group around the y-axis; α and β are adjustment coefficients; MTF L is the MTF value at the left end of the x-axis in the screen image II; MTF R is the MTF value at the right end of the x-axis in the screen image II; MTF U is the MTF value at the top of the y-axis in the screen image II; MTF D is the MTF value at the bottom of the y-axis in the screen image II.

4. The VR / AR glasses optical alignment and detection method according to claim 1, wherein In step 21), threshold segmentation is performed on the captured pattern, and possible noise points are removed through opening operation to extract all circular dots. Circle fitting is performed on each circular dot to obtain the center of the fitted circle, and the centers are sorted to obtain the center position of the central circle. The center position of the ideal imaging is calculated from adjacent points. The deviation between the ideal center position and the imaging center position is the current distortion of the target. The deviation at the outermost periphery of the camera field of view is used for distortion evaluation. At the same time, after calibrating the forward center, the deviation between the center of the camera coordinate system and the center position of the central circle is the final deviation of the VR / AR lens group.

5. The VR / AR glasses optical alignment and detection method according to claim 1, characterized in that, In step 22), a detection area is generated in the captured MTF fringe pattern within a set range. Threshold segmentation is performed on the MTF fringe pattern, and it is fully dilated to connect the fringe patterns to form a rectangle. Opening operation is performed on each area to remove noise. The MTF fringe area is obtained by screening according to the area, and the center of the circumscribed rectangle is used as the center of the fringe area. The contrast of the fringe is read according to the offset of the MTF fringe. The calculated MTF value is compared with the nominal value of the camera lens at the current frequency to compensate for the lens measurement error in the MTF detection result.

6. The VR / AR glasses optical alignment and detection method according to claim 1, wherein In step 23), frequency domain filtering is performed on the captured white field pattern, followed by inverse Fourier transform. Then, a threshold is set, and corresponding defect marks are obtained by screening according to the threshold. The average gray levels of the middle area and several peripheral areas are extracted from the pattern under the white field, and the gray level distribution state of the entire area is recorded. The ratio of the brightness of the peripheral and middle areas is obtained to get the brightness uniformity.

7. The VR / AR glasses optical alignment and detection method according to claim 1, wherein In step 24), an ROI is divided in the artifact detection pattern. Gray level enhancement is performed on the ROI to increase the gray level difference, and it is segmented by a threshold and appropriately eroded to reduce the interference from the periphery of the artifact. The ratio of the maximum gray level data in the artifact of each area to the gray level mean of the original image is obtained to get the relative ratio of the artifact.

8. The VR / AR glasses optical alignment and detection method according to claim 1, wherein The assembly, adjustment, and detection of the VR / AR lens group are completed using an auxiliary system. The auxiliary system includes a six-degree-of-freedom platform, a manual linear displacement stage, and a screen. The six-degree-of-freedom platform adsorbs and fixes the VR / AR lens group through a negative pressure air valve and drives it to move in the x-axis, y-axis, or z-axis direction, and rotate around the x-axis, y-axis, or z-axis. The manual linear displacement stage is arranged below the VR / AR lens group, and a camera is fixedly installed on the manual linear displacement stage and driven to move in the x-axis, y-axis, or z-axis direction. The screen is arranged above the VR / AR lens group and is used to display the patterns for assembly, adjustment, or detection.

Images