An imaging colorimeter-based OLED panel optical detection method and device

Abstract

The present application relates to the technical field of display panel optical detection, solves the problems of alignment deviation and lens distortion when using an imaging colorimeter in the prior art, and the difficulty in high-precision coordinate positioning of the pixels of OLED panels with barrel-shaped, pillow-shaped distortion, irregular cutting and the like, and provides an OLED panel optical detection method and device based on an imaging colorimeter, which establishes a one-to-one mapping relationship between a pixel digital map and a physical pixel, and uses a grid positioning mechanism of bidirectional auxiliary coordinates to effectively overcome the problems of image geometric deformation and pixel misalignment caused by the tilting, rotation or lens distortion of the imaging colorimeter. Compared with the traditional global rigid matching algorithm, the present application can independently detect and collect data for each pixel digital map, significantly improving the spatial correspondence accuracy of the light color extraction. The finally generated light color synthesis map is strictly the same size as the panel physical resolution, solving the technical problems of light color data misalignment and residual image after repair.

Description

Technical Field

[0001] This invention relates to the field of optical inspection technology for display panels, and in particular to an optical inspection method and device for OLED panels based on an imaging colorimeter. Background Technology

[0002] Organic light-emitting diode (OLED) display panels are widely used in smartphones, televisions, automotive displays, and other fields due to their advantages such as self-illumination, high contrast, and flexible display capabilities. In the OLED panel production process, precise testing of the panel's brightness and color is a crucial step in ensuring the quality of the finished product.

[0003] Imaging colorimeter-based detection technology has become the mainstream solution in the industry. Equipped with a scientific-grade CMOS sensor and an XYZ filter conforming to the CIE standard observer function, the imaging colorimeter can acquire high-precision brightness and color distribution information across the entire field of view in a single measurement, combining the advantages of fast measurement speed and high color accuracy.

[0004] However, in actual industrial testing, due to factors such as the magnification of the imaging system, the relative positional deviation between the panel and the imaging colorimeter, and lens distortion, the correspondence between the physical pixel luminescent area and the sensor imaging pixel is not a simple integer multiple mapping, which leads to spatial misalignment when synthesizing luminance and chromaticity data, making it difficult for current OLED panel testing applications based on imaging colorimeters.

[0005] To address the aforementioned issues, US Patent 10971044B2 proposes a method for improving equivalent resolution using fractional pixel and subpixel displacement sampling. However, the core technology of this patent focuses on improving resolution through hardware translation and multi-frame image interpolation reconstruction, and does not systematically solve the following problem: how to perform high-precision coordinate positioning of pixels in OLED panels with barrel, pincushion, or irregular shapes in a single-frame image with alignment deviations and lens distortion.

[0006] Therefore, there is an urgent need for an optical detection solution that can achieve high-precision pixel-level luminance and color extraction for irregularly shaped OLED panels. Summary of the Invention

[0007] In view of this, embodiments of the present invention provide an optical inspection method and device for OLED panels based on an imaging colorimeter, in order to solve the problems of alignment deviation and lens distortion in existing technologies that use imaging colorimeters, and the difficulty in performing high-precision coordinate positioning of pixels in OLED panels with barrel, pincushion, or irregular shapes.

[0008] In a first aspect, embodiments of the present invention provide an optical detection method for an OLED panel based on an imaging colorimeter, used to detect the luminance and color gamut of an OLED panel under test, wherein the OLED panel under test comprises a plurality of physical pixels; the method includes: The system controls the display of a calibration image on the OLED panel under test, and controls the movement of the imaging colorimeter and / or the OLED panel under test through an adjustment mechanism, so that the position parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements. The test image is displayed on the OLED panel under test, and the display area is imaged using the imaging colorimeter to obtain a test digital image. The test digital image includes multiple pixel digital images, each physical pixel corresponds to one pixel digital image, and one pixel digital image includes multiple imaging pixels. A pixel in the detected digital image is determined as a reference point and its coordinates are set. All pixel images are located along the row direction of the reference point, with the reference point as the origin, to determine auxiliary row coordinate axes. All pixel images are also located along the column direction of the reference point, with the reference point as the origin, to determine auxiliary column coordinate axes. The remaining pixel images are then located according to the auxiliary row and column coordinate axes. For each pixel digital image, a corresponding digital detector is set. The average brightness and average chromaticity values ​​obtained by each digital detector are used as the brightness and chromaticity measurement values ​​of the corresponding pixel digital image. A brightness and chromaticity composite image with the same size as the resolution of the OLED panel under test is generated and saved as a brightness and chromaticity data file.

[0009] By establishing a one-to-one mapping between pixel digital images and physical pixels, and utilizing a gridded positioning mechanism with bidirectional row and column auxiliary coordinates, this method effectively overcomes image geometric distortion and pixel misalignment problems caused by imaging colorimeter tilt, rotation, or lens distortion. Compared to traditional global rigid matching algorithms, this scheme can perform independent digital detection and data acquisition for each pixel digital image, significantly improving the spatial correspondence accuracy of luminance and chrominance extraction. The final generated luminance and chrominance composite image is strictly the same size as the panel's physical resolution, and the data file can be directly used for subsequent Demura compensation without conversion errors, thus solving the technical problems of luminance and chrominance data misalignment and afterimages caused by misalignment.

[0010] As an optional implementation, the OLED panel optical inspection method based on an imaging colorimeter further includes: controlling the imaging colorimeter to perform at least three non-overlapping translations relative to the OLED panel under test, with each translation distance being a multiple of the size of the imaging pixel, to achieve at least one of the following operating modes: First mode: Increase the resolution of the imaging colorimeter to improve the spatial resolution of the detected digital image; Second mode: Each physical pixel of the OLED panel under test is measured multiple times and sampled at least once by an imaging pixel to obtain a true color detection digital image; the imaging pixel includes at least three types of pixels: red, green and blue.

[0011] This embodiment utilizes a pixel interleaving algorithm to effectively increase the spatial sampling frequency of the imaging colorimeter by employing non-overlapping imaging pixel translation, enabling the imaging colorimeter to detect even minute pixel-level brightness unevenness. Furthermore, by ensuring that each physical pixel is independently measured by imaging pixels with R / G / B three-color responses, colorimetric measurement artifacts introduced by spatial deviations of the Bayer array or filters are eliminated, achieving true-color, high-fidelity luminance and chromaticity data acquisition.

[0012] As an optional implementation, the step of controlling the movement of the imaging colorimeter and / or the OLED panel under test through an adjustment mechanism to ensure that the position parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements specifically includes: Identify at least three white circular spots from the calibration image and obtain the calibration coordinates of each white circular spot. Calculate the center offset, horizontal tilt angle, vertical tilt angle, and rotation angle of the center of the calibration image relative to the field of view center of the imaging colorimeter based on the calibration coordinates. Calculate the spatial modulation based on the horizontal and vertical line pairs of the calibration image. Using the center offset, the imaging horizontal tilt angle, the vertical tilt angle, the rotation angle, and the spatial modulation degree as feedback parameters, the adjustment mechanism is driven to move the imaging colorimeter and / or the OLED panel under test to perform one or more combinations of X-axis translation, Y-axis translation, Z-axis translation, rotation around the X-axis, rotation around the Y-axis, and rotation around the Z-axis. Until the center offset is less than 0.2 mm, the deviations of the imaging horizontal tilt angle, the vertical tilt angle, and the rotation angle are all less than 0.2°, and the spatial modulation is between 0.8 and 0.99.

[0013] This embodiment addresses the shortcomings of existing technologies that employ manual or semi-automatic focusing, such as poor accuracy and low consistency, by providing a six-degree-of-freedom automated calibration strategy based on image feedback. By calculating the spatial modulation, it not only ensures the perpendicularity of the imaging colorimeter's optical axis but also confirms the optimal focusing plane. By controlling the center offset to within 0.2mm and the deviations in the imaging horizontal tilt angle, vertical tilt angle, and rotation angle to within 0.2°, it eliminates systematic errors caused by image blur, field-of-view eccentricity, and perspective distortion at the hardware level, providing stable and high-quality raw image input for subsequent pixel-level high-precision positioning.

[0014] As an optional implementation, the step of locating all the pixel digital images along the row direction of the reference point with the reference point as the origin to determine the auxiliary row coordinate axis includes: Within a preset first search window, identify the pixel digital image corresponding to the reference point and perform a first-step search along the row direction of the physical pixels; use the first search window to identify adjacent pixel digital images along the row direction and record the centroid coordinates of the adjacent pixel digital images; using the adjacent pixel digital images as new starting points, repeat the first-step search until all identifiable pixel digital images in the row direction are located; arrange the centroid coordinates of all located pixel digital images in sequence to form the auxiliary row coordinate axis; The step of locating all the pixel digital images along the column direction of the reference point with the reference point as the origin, and determining the auxiliary column coordinate axis, includes: Within a preset second search window, the pixel digital image corresponding to the reference point is identified, and a second step search is performed along the column direction of the physical pixel; the second search window is used to identify the pixel digital images adjacent along the column direction, and the centroid coordinates of the adjacent pixel digital images are recorded; the adjacent pixel digital images are used as new starting points, and the second step search is repeated until all identifiable pixel digital images in the column direction are located; the centroid coordinates of all located pixel digital images are arranged in sequence to form the auxiliary column coordinate axis.

[0015] As an optional implementation, the process of locating all the pixel digital images along the row direction of the reference point with the reference point as the origin to determine the auxiliary row coordinate axis, and locating all the pixel digital images along the column direction of the reference point with the reference point as the origin to determine the auxiliary column coordinate axis, includes: If a non-lit pixel digital image is encountered, the interpolated coordinates of the non-lit pixel digital image are obtained by interpolating the coordinates of the pixel digital images corresponding to two adjacent lit physical pixels.

[0016] In the actual production of OLED panels, there may be situations where pixels are completely damaged or a large number of pixels do not emit light when detecting low grayscale / black scenes. Traditional centroid localization algorithms fail when dealing with non-illuminated areas of the OLED panel. However, this embodiment fills in the coordinates of non-illuminated physical pixels by interpolating the coordinates of two adjacent lit physical pixels, ensuring the integrity of the physical pixel array mapping grid and improving the robustness of generating complete resolution luminance and chrominance data files.

[0017] As an optional implementation, the step of locating the coordinates of the remaining pixel digital image based on the auxiliary row coordinate axis and the auxiliary column coordinate axis includes: There is an auxiliary column coordinate axis value corresponding to the row direction of the remaining pixel digital image, and there is an auxiliary row coordinate axis value corresponding to the column direction of the remaining pixel digital image. The coordinates of the remaining pixel digital image are calculated by a vector operation algorithm based on the auxiliary column coordinate axis value and the auxiliary row coordinate axis value.

[0018] When inspecting large-size OLED panels, the barrel or pincushion distortion commonly found in imaging colorimeter lenses can lead to severe cumulative errors in traditional methods relying on fixed-step search, resulting in mispositioning of pixels in edge regions. This embodiment utilizes auxiliary row and column coordinates established through prior scanning and employs a vector algorithm for two-dimensional surface fitting and estimation. This adaptively fits the lens distortion curve, accurately estimating the true image plane position of each pixel even with significant image edge curvature. This effectively reduces measurement errors introduced by distortion and expands the compatibility of imaging colorimeters with low-cost, high-distortion lenses.

[0019] As an optional implementation, after locating the remaining pixel digits according to the auxiliary row coordinate axis and the auxiliary column coordinate axis, the method further includes: Within the detection area of ​​each digital detector, the imaging pixel with the highest brightness value is searched, and the coordinates of the pixel digital map are corrected to the location coordinates of the imaging pixel with the highest brightness value, so as to correct the coordinates of the pixel digital map.

[0020] In OLED panels, the actual light-emitting center of a physical pixel can exhibit T-shaped distortion or halo shift under tilted viewing or specific angles, causing a misalignment between the geometric center and the brightness centroid of the physical pixel. This embodiment introduces local peak brightness search and center correction to dynamically adjust the positioning reference of the digital detector from the geometrically calculated center to the point of highest brightness. This effectively eliminates measurement deviations caused by optical asymmetry during large-view detection (such as a 40° side view), ensuring that the digital detector always captures the area with the largest light-emitting contribution from the physical pixel, thereby significantly improving the accuracy and repeatability of colorimetric measurements.

[0021] As an optional implementation, the OLED panel optical inspection method based on an imaging colorimeter is used to detect one or more of the following detection scenarios: (a) The detected digital image has a rotation angle; (b) The detected digital image has a side-view T-shaped distortion and supports a maximum side-view angle of 40°; (c) The detected digital image exhibits barrel distortion or pincushion distortion; (d) The OLED panel under test has a punch-hole area, a notch area, or a rounded corner area; (e) The physical pixel arrangement of the OLED panel under test is either Uniform or SPR.

[0022] Existing technologies are often only effective for ideal, vertical, and defect-free rectangular panels. When faced with irregularly shaped screens, such as those with notches, punch holes, or special arrangements, custom development is required. This embodiment relies on row and column interpolation and coordinate correction, making it insensitive to image rotation and distortion. Furthermore, the row and column positioning mechanism can identify the physical boundaries of irregularly shaped areas and automatically skip punch hole areas. Naturally, it can cover the optical inspection needs of all types of display panels, from irregularly shaped mobile phone screens to large-size curved screens, greatly reducing the debugging costs and algorithm switching time when changing product models on the production line.

[0023] As an optional implementation, the OLED panel optical inspection method based on an imaging colorimeter further includes: When the physical pixel arrangement of the OLED panel under test is SPR arrangement, the OLED panel under test is made to present an equivalent display of Uniform arrangement by lighting up every other row and every other column. The equivalent display of all pixel digital images is located by interpolation positioning or combined positioning, and the following steps are performed: setting a corresponding digital detector for each pixel digital image, using the average brightness and average chromaticity values ​​obtained by each digital detector as the brightness and chromaticity measurement values ​​of the corresponding pixel digital image, generating a brightness and chromaticity composite image of the same size as the resolution of the OLED panel under test, and saving it as a brightness and chromaticity data file.

[0024] SPR array panels suffer from ambiguous physical pixel definitions due to sub-pixel sharing, making it difficult for traditional algorithms to directly locate and measure them. This embodiment uses specific interlaced row and column drive signals to transform the complex SPR logical arrangement into a regular Uniform logical arrangement at the moment of optical measurement. Subsequently, the above-mentioned auxiliary row and column positioning method can be directly applied for full pixel positioning, fundamentally avoiding optical crosstalk caused by SPR sub-pixel borrowing. This enables the luminance and chrominance extraction accuracy of SPR panels to reach the same level as standard RGB panels, filling the gap in existing technology for high-precision optical detection of SPR arrays.

[0025] Secondly, embodiments of the present invention provide an optical inspection device for OLED panels based on an imaging colorimeter, comprising: An imaging colorimeter is used to acquire digital images of the OLED panel under test. The lens is mounted on the imaging colorimeter, and its imaging circle size is larger than the CMOS target size of the imaging colorimeter, and the relative illuminance at the edge is not less than 80%. An adjustment mechanism is used to support the imaging colorimeter and / or the OLED panel under test, and to realize one or more combinations of translation along the X-axis, translation along the Y-axis, translation along the Z-axis, rotation around the X-axis, rotation around the Y-axis, and rotation around the Z-axis; A control and image processing module is electrically connected to the imaging colorimeter and the adjustment mechanism, respectively, and the control and image processing module is configured to perform any embodiment of the first aspect.

[0026] The lens design, with an imaging circle larger than the CMOS target surface and an edge relative illumination of no less than 80%, physically suppresses the vignetting effect at the edges of the imaging colorimeter image, ensuring the consistency of the dynamic range measurement across the entire image. Combined with the adjustment mechanism and the control and image processing module incorporating the optical detection algorithm mentioned in the first aspect, this device achieves full automation from mechanical orientation adjustment and optical imaging to data processing. This equipment can be deployed directly as a standardized machine on OLED mass production lines, or in OLED panel quality inspection facilities. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0028] Figure 1 This is a schematic flowchart of the OLED panel optical detection method based on an imaging colorimeter in an embodiment of the present invention; Figure 2 This is a schematic diagram of the calibration image in one embodiment of the present invention; Figure 3 This is a schematic diagram of the calibration image recognition window in one embodiment of the present invention; Figure 4 This is a schematic diagram of the detected image in one embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the establishment of reference coordinates in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the establishment of reference coordinates for the OLED panel under test in an embodiment of the present invention; Figure 7 This is a schematic diagram illustrating the establishment of reference coordinates for barrel distortion or pincushion distortion in one embodiment of the present invention; Figure 8 This is a schematic diagram illustrating the establishment of reference coordinates for barrel or pincushion distortion in another embodiment of the present invention; Figure 9This is a schematic diagram illustrating the establishment of reference coordinates for T-shaped distortion or halo shift from a side viewpoint in one embodiment of the present invention; Figure 10 This is a schematic diagram illustrating the establishment of reference coordinates for T-shaped distortion or halo shift from a side view in another embodiment of the present invention; Figure 11 This is a schematic diagram illustrating the establishment of reference coordinates for the SPR arrangement panel in an embodiment of the present invention; Figure 12 This is a schematic diagram of the structure of the OLED panel optical inspection device based on an imaging colorimeter in an embodiment of the present invention; Figure 13 This is a structural diagram of an OLED panel optical inspection device based on an imaging colorimeter in one embodiment of the present invention. Detailed Implementation

[0029] The features and exemplary embodiments of various aspects of the present invention will now be described in detail. To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present invention and are not configured to limit the present invention. For those skilled in the art, the present invention can be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present invention by illustrating examples of the invention.

[0030] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0031] In a first aspect, embodiments of the present invention provide an optical detection method for an OLED panel based on an imaging colorimeter, used to detect the brightness and color of an OLED panel under test. The OLED panel under test includes multiple physical pixels, which are pixel units that actually emit light on the OLED panel and include R / G / B sub-pixels or other arranged pixels.

[0032] Please see below. Figure 1As shown, the optical inspection method for OLED panels based on an imaging colorimeter includes the following steps S101 to S105: Step S101: Control the display of the calibration image on the OLED panel under test, and control the movement of the imaging colorimeter and / or the OLED panel under test through the adjustment mechanism so that the position parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements.

[0033] Specifically, an imaging colorimeter is an optical instrument capable of simultaneously measuring the luminance and chromaticity of points in a two-dimensional image. It typically consists of a high-resolution CMOS (Complementary Metal Oxide Semiconductor) sensor and filters matched to the CIE color matching function. The calibration image can be set according to the actual detection configuration requirements. In one optional implementation, the calibration image can be referenced as follows: Figure 2 As shown, the calibration image has a pure black background (i.e., a grayscale value of 0), and five white circular spots with preset coordinate positions and two sets of orthogonal line pairs are set on the background. The five white circular spots are located at the center of the image and at the preset coordinates of the four corner areas. The horizontal line pairs consist of multiple sets, each consisting of multiple white stripes of equal width arranged at equal intervals along the horizontal direction; the vertical line pairs also consist of multiple sets, each consisting of multiple white stripes of equal width arranged at equal intervals along the vertical direction. Specifically, ensuring that the positional parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements can include: Identify at least three white circles in the calibration image and obtain the calibration coordinates of each white circle. Calculate the center offset between the center of the calibration image and the field of view center of the imaging colorimeter, the horizontal tilt angle, the vertical tilt angle, and the rotation angle based on the calibration coordinates. Calculate the spatial modulation based on the horizontal and vertical line pairs of the calibration image. Using center offset, imaging horizontal tilt angle, vertical tilt angle, rotation angle, and spatial modulation as feedback parameters, the adjustment mechanism drives the imaging colorimeter and / or the OLED panel under test to perform one or more combinations of X-axis translation, Y-axis translation, Z-axis translation, rotation around the X-axis, rotation around the Y-axis, and rotation around the Z-axis. The criteria are as follows: center offset less than 0.2 mm, horizontal tilt angle, vertical tilt angle, and rotation angle all less than 0.2°, and spatial modulation degree between 0.8 and 0.99. Here, center offset is the pixel deviation between the center of the calibrated image and the center of the imaging colorimeter's field of view; horizontal tilt angle is the tilt angle of the imaging colorimeter's optical axis about the X-axis (horizontal axis) relative to the panel normal; vertical tilt angle is the tilt angle of the imaging colorimeter's optical axis about the Y-axis (vertical axis); rotation angle is the rotation angle of the imaging colorimeter about the optical axis (Z-axis) relative to the panel; and spatial modulation degree is a modulation transfer function used to evaluate the imaging system's ability to transfer contrast at spatial frequencies, with values ​​closer to 1 indicating greater sharpness.

[0034] In practical implementation, the aforementioned adjustment mechanism can be a motorized six-axis stage. By calculating the spatial modulation and feeding back the target imaging distance to the adjustment mechanism, the adjustment mechanism changes the actual distance between the imaging colorimeter and the OLED panel under test, effectively ensuring the perpendicularity of the imaging colorimeter's optical axis and enabling the imaging colorimeter to have the optimal focusing plane. Furthermore, by controlling the center offset within 0.2mm, the deviations of the imaging horizontal tilt angle, vertical tilt angle, and rotation angle are all controlled within 0.2°, eliminating systematic errors caused by imaging blur, field of view eccentricity, and perspective distortion from the hardware source, providing stable and high-quality raw image input for subsequent pixel-level high-precision positioning.

[0035] More specifically, the aforementioned center offset, imaging horizontal tilt angle, vertical tilt angle, rotation angle, and spatial modulation can be quickly confirmed using image recognition algorithms. Specifically, the calculation of the center offset, imaging horizontal tilt angle, vertical tilt angle, and rotation angle can be achieved by extracting the centroid coordinates of at least three white circular spots in the image plane coordinate system of the imaging colorimeter, and establishing a homography transformation matrix with the preset coordinates of the white circular spots in the physical coordinate system of the OLED panel under test. The homography transformation matrix is ​​then decomposed to obtain the center offset, imaging horizontal tilt angle, vertical tilt angle, and rotation angle. Specifically, the calculation of the spatial modulation can be achieved by extracting the regions of interest (ROIs) for both horizontal and vertical lines in the calibration image; calculating the maximum and minimum gray values ​​within the ROIs of the horizontal lines to obtain the horizontal contrast; calculating the maximum and minimum gray values ​​within the ROIs of the vertical lines to obtain the vertical contrast; and calculating the spatial modulation based on the horizontal and / or vertical contrast.

[0036] During the calibration of the positional relationship between the imaging colorimeter and the OLED panel under test, a calibration image recognition window is generated by acquiring a calibration image. In one optional implementation, the calibration image recognition window can refer to, for example,... Figure 3 As shown, Figure 3The left side of the image displays the positional image quality evaluation metrics of the OLED panel under test, while the right side shows the real-time image. The yellow-outlined ellipse and closed quadrilateral indicate the current position between the imaging colorimeter and the OLED panel under test; the green dashed rectangle represents the adjustment target.

[0037] The calibration image recognition window calculates the center offset, imaging horizontal tilt angle, vertical tilt angle, rotation angle, and spatial modulation, as well as the corresponding adjustment values ​​for these parameters. These adjustment values ​​are then sent to the adjustment mechanism, which performs automated and closed-loop adjustment to ensure that the position parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements.

[0038] In practical implementation, the calculation of center offset, imaging horizontal tilt angle, vertical tilt angle, and rotation angle can refer to the following steps A1 to A3: Step A1: Perform binarization and connected component analysis on the acquired digital image to identify at least three regions of white circular spots. For each identified white circular spot region, calculate its image plane coordinates using the gray-scale centroid method to extract the centroid of the white spot. Let the image plane coordinates of a certain white circular spot within the field of view of the imaging colorimeter be (x... i ,y i The known preset physical coordinates of the OLED panel under test in the coordinate system are (X... i ,Y i ), where i = 1, 2, 3, ..., n; and n ≥ 3.

[0039] Step A2: Based on the centroid matching relationship of at least three extracted white circular spots, i.e. (x i ,y i ) and (X i ,Y i Construct the homography transformation matrix between the image plane coordinate system and the physical coordinate system. The homography transformation matrix satisfies the following relationship: Image plane homogeneous coordinate vector [x i ,y i The transpose of [1] and the physical homogeneous coordinate vector [X] i ,Y i The transpose of [1], after being subjected to the homography transformation matrix, is proportional, denoted as:

[0040] Here, H is a 3x3 homography transformation matrix, where the symbol "·" represents matrix multiplication and the symbol "~" represents a proportional relationship. The homography transformation matrix contains information on rotation, translation, scaling, and perspective projection.

[0041] Using the known focal length and working distance parameters of the imaging colorimeter lens, a matrix decomposition algorithm, such as RQ decomposition or the Levenberg-Marquardt iterative optimization algorithm, is employed to decompose the homography matrix into geometric motion components: center offset, imaging horizontal tilt angle, vertical tilt angle, and rotation angle. This enables the calculation of the homography matrix and the decomposition of position parameters. Specifically, the center offset includes offsets along the X-axis and Y-axis of the panel; the imaging horizontal tilt angle is the angle by which the imaging colorimeter's optical axis rotates around the X-axis of the OLED panel under test; the vertical tilt angle is the angle by which the imaging colorimeter's optical axis rotates around the Y-axis of the OLED panel under test; and the rotation angle is the angle by which the imaging colorimeter's sensor plane rotates around the Z-axis of the OLED panel under test.

[0042] Step A3: Use the calculated center offset, imaging horizontal tilt angle, vertical tilt angle, and rotation angle as real-time feedback parameters, and compare these feedback parameters with the preset allowable deviation range: If the absolute value of the center offset is greater than or equal to 0.2mm, a corresponding drive command is generated to control the adjustment mechanism to perform X-axis translation and / or Y-axis translation. If the absolute value of the horizontal tilt angle is greater than or equal to 0.2°, a corresponding drive command is generated to control the adjustment mechanism to rotate around the X-axis; if the absolute value of the vertical tilt angle is greater than or equal to 0.2°, a corresponding drive command is generated to control the adjustment mechanism to rotate around the Y-axis; if the absolute value of the rotation angle is greater than or equal to 0.2°, a corresponding drive command is generated to control the adjustment mechanism to rotate around the Z-axis.

[0043] Repeat steps A1 to A3 until the absolute value of the calculated center offset is less than 0.2 mm and the absolute value of each angle deviation is less than 0.2°.

[0044] After confirming that the center offset and angular deviation meet the preset conditions, the spatial modulation is further calculated. In specific implementation, the calculation of the spatial modulation can refer to the following steps B1 to B3: Step B1: Based on the pre-set physical coordinates of the horizontal and vertical line pairs in the calibration image, and combined with the homography matrix calculated in step A2 above, accurately map and extract the regions of interest for the horizontal and vertical line pairs on the acquired digital image.

[0045] Step B2: For the region of interest of the horizontal line pair, calculate the gray-level distribution curve along the direction perpendicular to the line, and find the gray-level maxima and minima. Calculate the horizontal spatial modulation; the formula can be found here: MTF H =(I max H-I min ,H) / (I maxH+I min ,H) Among them, MTF H For the horizontal spatial regulation, I max H represents the maximum grayscale value, and I represents... min H represents the minimum gray level.

[0046] Similarly, for the region of interest along the vertical line, calculate the gray-level maxima and minima, and then calculate the vertical spatial modulation. The calculation formula can be found here: MTF V =(I max V-I min ,V) / (I max V+I min ,V) Among them, MTF V For vertical spatial modulation, I max V represents the maximum grayscale value, and I represents the maximum grayscale value. min V represents the minimum gray level.

[0047] The arithmetic mean of the horizontal spatial regulation and the vertical spatial regulation is used as the comprehensive spatial regulation.

[0048] Step B3: Compare the calculated spatial modulation with the preset threshold range [0.8, 0.99] as follows: If the spatial modulation degree is less than 0.8, it indicates that the current system is out of focus or the image is blurry. It is necessary to drive the adjustment of the focal length of the imaging colorimeter lens and return to re-execute the image acquisition and calculation process. If the spatial modulation index is greater than 0.99, it indicates that there may be overexposure leading to contrast saturation distortion, and the exposure time of the imaging colorimeter or the brightness of the light source should be reduced. If the spatial modulation is greater than or equal to 0.8 and less than or equal to 0.99, it indicates that the current focus and imaging contrast meet the testing requirements, and the calibration process ends.

[0049] Step S102: Display the detection image on the OLED panel under test, and use an imaging colorimeter to image the display area to obtain a detection digital image.

[0050] The detection image can be set according to the actual luminescence detection requirements. The detection image can be pure green or pure white. In one optional implementation, the detection image can refer to... Figure 4 As shown. The detected digital image includes multiple pixel digital maps, with each physical pixel corresponding to one pixel digital map, and one pixel digital map including multiple imaging pixels.

[0051] It's necessary to distinguish between the imaging pixel and the physical pixels of the OLED panel under test. An imaging pixel is the smallest photosensitive unit of the CMOS sensor inside the imaging colorimeter, similar to a pixel in a digital camera. Correspondingly, the detection digital image is the raw digital image acquired by the imaging colorimeter, composed of the grayscale or color values ​​of multiple imaging pixels, while the pixel digital image is a small area within the detection digital image corresponding to a single physical pixel, typically containing multiple imaging pixels. For example, a pixel digital image can include 17×17 imaging pixels.

[0052] Step S103: Determine a pixel digital image from the detected digital image as a reference point and set the reference coordinates; locate all pixel digital images along the row direction of the reference point with the reference point as the origin to determine the auxiliary row coordinate axis, and locate all pixel digital images along the column direction of the reference point with the reference point as the origin to determine the auxiliary column coordinate axis.

[0053] In the actual production of OLED panels, there may be cases where pixels are completely damaged or a large number of pixels do not emit light when detecting low grayscale / black images. These non-emitting (dead pixels or black patterns) physical pixel digital image areas are called non-illuminated pixel digital images, which cannot be directly located. Traditional centroid positioning algorithms fail when dealing with non-illuminated areas of OLED panels. However, this invention determines a pixel digital image from the detected digital image as a reference point and sets reference coordinates. The reference point is generally selected as the center of the detected digital image, and the row and column of the reference point's coordinates preferably do not pass through areas without physical pixels, such as punch holes, rounded corners, or notches. In an optional implementation, the selection of the reference point can follow these rules: 1. Preferably use pixel digit maps that are easily identifiable and located near the center of the OLED panel to be tested. This is because the distortion in the center area is relatively small, resulting in higher positioning reliability.

[0054] 2. The physical pixel corresponding to the reference point must be lit to ensure a sufficient signal-to-noise ratio.

[0055] 3. If the panel displays a uniformly bright field image, you can select the pixel image closest to the geometric center of the image; if a specific positioning pattern is displayed, such as... Figure 5 As shown, the feature points of the pattern (such as crosses or isolated bright spots) are used as reference points.

[0056] After determining the reference point, obtain the precise coordinates of the pixel digital image corresponding to the reference point, denoted as (X0, Y0). For determining the auxiliary row and column coordinate axes, in practice, start with the pixel digital image corresponding to the reference point and perform the first progressive search along the row direction of the physical pixels corresponding to the reference point. Within the preset first search window, identify the pixel digital images corresponding to adjacent physical pixels and record their centroid coordinates until all identifiable pixel digital images in that row direction are located. Arrange the location results sequentially to form the auxiliary row coordinate axes.

[0057] A second-step search is then performed along the column direction of the physical pixels corresponding to the reference point. Within a preset second search window, the pixel digital images corresponding to adjacent physical pixels are identified and their centroid coordinates are recorded until all identifiable pixel digital images along the column direction are located. The location results are then arranged sequentially to form an auxiliary column coordinate axis. The center of the first search window in the first-step search is determined based on the known physical pixel row spacing and the estimated current magnification of the imaging colorimeter. The center of the second search window in the second-step search is determined based on the known physical pixel column spacing and the estimated current magnification of the imaging colorimeter.

[0058] Regarding the determination of the auxiliary row coordinate axis, for example, starting from the reference point coordinates (X0, Y0), the first step of the search is performed to the left and right along the row direction of the physical pixels (i.e., the horizontal pixel arrangement direction of the OLED panel under test). The single-step execution logic of the search process is as follows: Based on the known physical pixel pitch of the OLED panel under test X Given the current imaging system's magnification M, calculate the estimated row step size D1=Pitch for adjacent physical pixels on the image plane. X ×M. The center of the first search window is set at the estimated adjacent position of the current pixel, and the size of the first search window is set to [D1×(1-α),D1×(1+α)], where α is the search tolerance coefficient, which is usually taken as 0.2 to 0.5 to tolerate lens distortion and panel manufacturing tolerances.

[0059] Within the first search window, the centroid coordinates of candidate pixel digit images are extracted using connected component analysis combined with the gray-scale centroid method. If the area and brightness value of the region corresponding to the extracted centroid coordinates meet the preset physical pixel feature threshold, then the location is confirmed as a valid pixel digit image, and its centroid coordinates are recorded as P. k =(X k ,Y k ), where k is the index number along the row direction, k=0 corresponds to the reference point, increasing to the right in a positive direction and decreasing to the left in a negative direction.

[0060] With the newly confirmed P kStarting from a new point, repeat the above steps to continue searching outwards until the image boundary region is reached, or until several consecutive searches fail to find a valid pixel image. All valid coordinate sequences {P} obtained along the row direction are then processed. k Arranged in sequence, these form an auxiliary row coordinate axis. This auxiliary row coordinate axis actually depicts the actual image position trajectory of the physical pixels in the row containing the reference point under the conditions of lens distortion and rotation.

[0061] Regarding the determination of the auxiliary column coordinate axis, for example, a completely symmetrical method is adopted, starting from the reference point (X0,Y0), and performing the second step search upward and downward along the column direction of the physical pixels (i.e. the vertical pixel arrangement direction of the OLED panel under test).

[0062] The single-step execution logic of the search process is the same as that of the row-direction search, except that the estimated row step size is replaced with the estimated column step size, where the estimated column step size D2 = Pitch. Y ×M, Pitch Y The physical pixel column spacing. The centroid coordinates of the effective pixel digital map during the search process in the second search window are set as P. l =(X l ,Y l ), where l is the index number along the row direction, l=0 corresponds to the reference point, upward is the positive direction and downward is the negative direction.

[0063] All valid coordinate sequences {P} obtained by searching along the column direction l Arranged in sequence, these form an auxiliary column coordinate axis. This auxiliary column coordinate axis depicts the actual position trajectory of the physical pixels in the column containing the reference point on the image.

[0064] Furthermore, during the first or second step search, if a valid pixel image cannot be identified in the current search window, the current position is skipped, and the corresponding auxiliary row or column coordinates are marked as missing positions. After the search is completed, the missing position is linearly interpolated using the two valid positioning coordinates adjacent to the missing position to obtain the interpolated coordinates corresponding to the missing position.

[0065] In actual testing, the following reasons may cause a step search window to fail to find a valid pixel image: 1. The physical pixel at this location is a dead pixel; the physical pixel itself is not lit. 2. The current image being detected is a low grayscale image, and the brightness of some pixels is below the detection threshold; 3. The search path passes through the edge of the punch-hole area or notch area of ​​the OLED panel under test.

[0066] In view of the above situation, if a valid pixel digital image cannot be confirmed in a certain step search, the coordinate of the index position is recorded as a null value mark, and the next step search is continued. When the search in this direction is completed, traverse the recorded coordinate sequence. For each position marked as a null value, extract two adjacent valid positioning coordinates before and after it, and use the linear interpolation formula to calculate the interpolated coordinate of the missing position.

[0067] For example, if the valid coordinate in front of the missing position is P 前 =(X 前 ,Y 前 ), the valid coordinate behind is P 后 =(X 后, Y 后 ), and the relative position proportion of the missing position in the index sequence is t (0<t<1), then the interpolation coordinate satisfies the following linear relationship: X 缺 =X 前 +t×(X 后 -X 前 ), Y 缺 =Y 前 +t×(Y 后 -Y 前 ) Write the calculated interpolation coordinates into the missing positions to ensure that the index sequences of the auxiliary row coordinate axis and the auxiliary column coordinate axis are continuous and complete. In this way, even under the condition of local pixel defects or dark field of the image, a complete reference grid of auxiliary row / column coordinate axes that covers the entire panel area and can truly reflect the image distortion and rotation state can still be obtained. The coordinate of non-bright physical pixels is filled by coordinate interpolation of two adjacent bright physical pixels, which ensures the integrity of the physical pixel array mapping grid and improves the robustness of generating a complete resolution luminance-chrominance data file.

[0068] Step S104: Perform coordinate positioning on the remaining pixel digital images according to the auxiliary row coordinate axis and the auxiliary column coordinate axis.

[0069] Specifically, as shown in Figure 6 , there can be an auxiliary column coordinate axis value corresponding along the row direction of the remaining pixel digital image, and an auxiliary row coordinate axis value corresponding along the column direction of the remaining pixel digital image, and the coordinates of the remaining pixel digital image are calculated by using a vector operation method based on the auxiliary column coordinate axis value and the auxiliary row coordinate axis value. In the process of calculating all pixel digital images, a vertical synchronization sequential replication positioning method can be adopted.

[0070] For the remaining pixel digital image in the i-th row and j-th column to be located, associate its row index i with the corresponding value in the auxiliary column coordinate axis, and associate its column index j with the corresponding value in the auxiliary row coordinate axis. Calculate the preliminary estimated coordinates of the remaining pixel digital image by two-dimensional surface fitting or bilinear interpolation algorithm. Perform precise centroid localization within the local neighborhood of the initially estimated coordinates, and use the centroid coordinates as the final localization coordinates of the remaining pixel digital image.

[0071] For example, let's take locating pixel A in the digital image as an example: Suppose we determine a pixel D as the reference point in the detected digital image, B is a pixel on the auxiliary column coordinate axis, and C is a pixel on the auxiliary row coordinate axis. Since the pixel coordinates of B, C, and D in the detected digital image are known, let... + = Thus, the coordinates of pixel A in the digital image can be calculated.

[0072] When inspecting large-size OLED panels, barrel or pincushion distortion, commonly found in imaging colorimeter lenses, leads to severe cumulative errors in traditional methods relying on fixed-step search, resulting in misaligned pixel positioning in edge regions. However, this invention utilizes auxiliary row and column coordinates established through prior scanning and employs a vector algorithm for two-dimensional surface fitting and calculation, enabling adaptive fitting of the lens distortion curve. Figure 7 and Figure 8 As shown, even if there is obvious curvature at the image edges, the true image plane position of each pixel can be accurately calculated, effectively reducing the measurement error introduced by distortion and expanding the compatibility of the imaging colorimeter with low-cost, high-distortion lenses.

[0073] Furthermore, because the actual light-emitting center of the physical pixels in an OLED panel exhibits T-shaped distortion or halo shift under tilted viewing or specific viewing angles, the geometric center of the physical pixel does not align with the brightness centroid. Figure 9 and Figure 10 As shown. After locating the coordinates of the remaining pixel digital map according to the auxiliary row coordinate axis and the auxiliary column coordinate axis, the imaging pixel with the highest brightness value can be searched within the detection area of ​​each digital detector, and the coordinates of the pixel digital map can be corrected to the position coordinates of the imaging pixel with the highest brightness value, so as to correct the coordinates of the pixel digital map.

[0074] By dynamically adjusting the positioning reference of the digital detector from the geometric calculation center to the point of highest brightness, the measurement deviation caused by optical asymmetry during large-view detection (such as a 40° side view) is effectively eliminated. This ensures that the digital detector always captures the area where the physical pixel contributes the most light emission, thereby significantly improving the accuracy and repeatability of colorimetric measurements.

[0075] Step S105: Set up a corresponding digital detector for each pixel digital image, use the average brightness and average chromaticity values ​​obtained by each digital detector as the brightness and chromaticity measurement values ​​of the corresponding pixel digital image, generate a brightness and chromaticity composite image of the same size as the resolution of the OLED panel under test, and save it as a brightness and chromaticity data file.

[0076] Regarding digital detectors, it's important to note that a digital detector is a virtual photosensitive area used to statistically analyze multiple imaging pixels within a pixel digital image, calculating the average brightness and chromaticity. A digital detector is either a rectangular area with sides equal to the distance between physical pixels, or a circular area with a diameter equal to the distance between physical pixels; the size of the digital detector matches the distance between the physical pixels. For example, a digital detector could be a rectangular area consisting of 17×17 imaging pixels or a circular area with a diameter of 17 imaging pixels.

[0077] The luminance mean is the arithmetic mean of the luminance values ​​of all imaging pixels within the digital detector, and the chrominance mean is the arithmetic mean of the chrominance coordinates of all imaging pixels within the digital detector. The luminance-chrominance composite image is a pseudo-color image generated by mapping the luminance and chrominance values ​​of each physical pixel to a two-dimensional matrix, with the same resolution as the panel. The luminance-chrominance data file stores the luminance and chrominance measurement results, commonly in CSV (Comma-Separated Values) format, with rows and columns corresponding to the panel rows and columns.

[0078] The rectangular digital detector is well-suited for the regular light-emitting areas of the Uniform arrangement, while the circular digital detector is better suited for the SPR arrangement with Gaussian diffusion of the light-emitting halo or specific OLED light-emitting characteristics. By strictly matching the spacing between physical pixels, crosstalk between adjacent pixels caused by an excessively large digital detector area is effectively avoided, as well as the loss of effective light emission due to an excessively small detection area. This ensures that the calculated average luminance and chrominance values ​​can truly reflect the independent optical characteristics of the physical pixel, rather than a mixture of values ​​from multiple pixels, thus improving the signal-to-noise ratio of luminance and chrominance data extraction.

[0079] like Figure 11 As shown, when the physical pixel arrangement of the OLED panel under test is SPR arrangement, the OLED panel under test can be made to present an equivalent display of Uniform arrangement by lighting up every other row and every other column. Then, the digital image of all pixels in the equivalent display is located by interpolation positioning or combination positioning, and the step of generating a luminance and color composite image with the same size as the resolution of the OLED panel under test is performed.

[0080] SPR array panels suffer from ambiguous physical pixel definitions due to sub-pixel sharing, making it difficult for traditional algorithms to directly locate and measure them. This embodiment uses specific interlaced row and column drive signals to transform the complex SPR logical arrangement into a regular Uniform logical arrangement at the moment of optical measurement. Subsequently, the above-mentioned auxiliary row and column positioning method can be directly applied for full pixel positioning, fundamentally avoiding optical crosstalk caused by SPR sub-pixel borrowing. This enables the luminance and chrominance extraction accuracy of SPR panels to reach the same level as standard RGB panels, filling the gap in existing technology for high-precision optical detection of SPR arrays.

[0081] Therefore, the above-described optical inspection method for OLED panels based on imaging colorimeters can be used to inspect one or more of the following inspection scenarios: 1. In one optional embodiment, when the detected digital image has a rotation angle, for example, after the imaging colorimeter is assembled or changed, there is a certain angle of deflection relative to the OLED panel under test around the optical axis, causing the pixel row and column directions in the acquired detected digital image to no longer be parallel to the horizontal and vertical axes of the image coordinate system. In this case, this embodiment does not directly search for pixels according to the fixed horizontal and vertical step sizes of the image coordinate system. Instead, it first selects a pixel digital image that is lit up and easily identifiable from the detected digital image as a reference point, and performs a first step search along the actual row direction of the physical pixels with the reference point as the origin to locate multiple pixel digital images in the row where the reference point is located, thereby forming an auxiliary row coordinate axis; at the same time, it performs a second step search along the actual column direction of the physical pixels to locate multiple pixel digital images in the column where the reference point is located, thereby forming an auxiliary column coordinate axis. Since the auxiliary row and column coordinate axes are generated based on the centroid coordinates of the actual pixel digital image, their directions can rotate synchronously with the rotation of the detected digital image. Therefore, when calculating the coordinates of the remaining pixel digital image, the positioning can be based on the actual row and column directions of the physical pixel array, rather than being limited by the fixed horizontal and vertical directions of the image coordinate system. In this way, even if the detected digital image as a whole is rotated, the one-to-one correspondence between physical pixels and pixel digital images can be maintained, avoiding misalignment of row and column numbers or incorrect writing of luminance and chrominance data due to image rotation.

[0082] 2. In another optional embodiment, when a T-shaped distortion is detected in the digital image at a side viewing angle, for example, to detect the brightness and chromaticity changes of an OLED panel at a wide viewing angle, the imaging colorimeter is controlled to acquire the digital image at an angle relative to the panel normal, with a maximum side viewing angle of 40°. In this scenario, due to the angle between the viewing direction and the panel's light emission direction, the imaging area of ​​some physical pixels may exhibit a "T"-shaped stretching, offset, or halo asymmetry, resulting in the geometric center calculated from the row and column coordinates not being entirely consistent with the area where the actual brightness contribution of the physical pixel is greatest. To address this, this embodiment, after completing the initial coordinate positioning based on the auxiliary row and column coordinate axes, further searches for the imaging pixel with the highest brightness value within the detection area of ​​each digital detector, and corrects the coordinates of the corresponding pixel digital image to the location coordinates of the imaging pixel with the highest brightness value. In other words, the row and column auxiliary coordinates are used to ensure which physical pixel the pixel digital image belongs to, and the local brightness peak search is used to further correct the actual light emission center of the physical pixel under side viewing conditions. This processing allows the digital detector to extract the average luminance and chromaticity values ​​around the region that contributes the most to luminance even if the geometric center of the physical pixel deviates from the luminance centroid, thereby improving the measurement accuracy and repeatability under wide-view detection conditions.

[0083] 3. In another optional embodiment, when barrel or pincushion distortion is detected in the digital image, for example, when an imaging colorimeter uses a large field-of-view lens to image the entire OLED panel in one pass, the pixel spacing in the central region of the image is not completely consistent with the pixel spacing in the edge region, causing the pixel rows or columns near the image edge to be curved. In this case, if a fixed pixel spacing or fixed step size is still used to directly push outward from the center, the positioning error will gradually accumulate with the increase of distance, eventually causing the pixel digital image in the edge region to be located as an adjacent physical pixel. To address this, this embodiment first establishes auxiliary row coordinate axes and auxiliary column coordinate axes along the row and column where the reference point is located, and uses these coordinate sequences obtained from actual sampling to describe the curvature trend of the pixel array in the detected digital image; subsequently, for the remaining pixel digital image in the i-th row and j-th column, its row index can be associated with the corresponding position in the auxiliary column coordinate axis, and its column index can be associated with the corresponding position in the auxiliary row coordinate axis, and the coordinates of the pixel digital image can be calculated by vector operation, two-dimensional surface fitting, or bilinear interpolation. Because the calculation process references the actual coordinate change trend of the located pixel digital image, it can adapt to the curved grid shape under barrel or pincushion distortion, and compensate for the pixel rows and columns that are curved at the image edge, thereby improving the pixel-level positioning accuracy under conditions of large-size panels or lenses with low distortion control requirements.

[0084] 4. In another optional embodiment, when the OLED panel under test has a punch-hole area, a notch area, or a rounded corner area, such as a mobile phone OLED panel with a camera punch-hole, earpiece notch, or rounded corners at the top, some theoretical row and column positions do not have actual physical pixels. In this case, when selecting the reference point, this embodiment prioritizes selecting pixel digital maps located near the center area of ​​the panel, in an illuminated state, and whose row and column do not pass through the punch-hole area, notch area, or rounded corner area, to ensure good continuity in the establishment process of the auxiliary row coordinate axis and auxiliary column coordinate axis. For localized missing locations encountered during the search process, if the location corresponds to a non-lit pixel or cannot be identified in the effective pixel digital map due to irregular cutting, the location can be marked as a missing position. For non-lit pixels located within the effective display area, the interpolated coordinates of the non-lit pixel digital map can be obtained by interpolating the pixel digital map coordinates corresponding to the two adjacent lit physical pixels. For locations without physical pixels located in punch-hole areas, notch areas, or rounded corner areas, they are not considered as effective luminance and chrominance measurement objects and can be set to null values, mask values, or invalid marks when generating the luminance and chrominance composite map. In this way, the row and column positioning mechanism can maintain the continuity of the pixel coordinate sequence within the effective display area and identify and avoid non-display areas formed by irregular cutting, avoiding misjudging punch-hole or rounded corner areas as dark spots or low-brightness defects.

[0085] 5. In another optional embodiment, when the physical pixel arrangement of the OLED panel under test is a Uniform arrangement, since each physical pixel is distributed in a regular row and column array, the row spacing and column spacing between pixel digital images are relatively stable. Therefore, a rectangular digital detector can be used to extract the luminance and chrominance of each pixel digital image. Specifically, after completing the positioning of the auxiliary row coordinate axis, auxiliary column coordinate axis, and the coordinates of the remaining pixel digital images, a rectangular digital detector matching the physical pixel spacing is set with the positioning coordinates of each pixel digital image as the center. The average luminance and chrominance values ​​of multiple imaging pixels within the coverage area of ​​the rectangular digital detector are counted as the luminance and chrominance measurement values ​​of the corresponding physical pixels. Since the size of the digital detector matches the physical pixel spacing, crosstalk between adjacent pixel emitting areas and the current pixel measurement value can be reduced, and the generated luminance and chrominance composite image can be ensured to maintain the same size correspondence with the physical resolution of the OLED panel under test.

[0086] 6. In another optional embodiment, when the physical pixel arrangement of the OLED panel under test is an SPR arrangement, since there may be shared or misaligned sub-pixels between adjacent physical pixels, if positioning and measurement are performed directly according to the complete illuminated image, optical crosstalk between adjacent pixels is likely to occur, causing the luminance and chromaticity measurement value of a certain pixel's digital image to be affected by the emission of surrounding sub-pixels. To address this, this embodiment can drive the OLED panel under test using an alternating row and column illumination method, so that the OLED panel under test presents an equivalent display of approximately a Uniform arrangement in the current detection frame. Subsequently, in this equivalent display state, all pixel digital images are positioned using interpolation positioning or combined positioning methods, and the corresponding digital detector is set according to the positioning results. For SPR pixels with Gaussian diffusion of the emission halo or incompletely regular emission areas, a circular digital detector can be selected to reduce the truncation effect of the rectangular boundary on the irregular emission area. By combining the above-mentioned alternating row and column illumination, equivalent Uniform positioning, and digital detector matching settings, the complex SPR pixel arrangement can be transformed into a more easily identifiable and measurable regular detection state, thereby improving the pixel-level luminance and chrominance extraction accuracy of the SPR arrangement panel.

[0087] 7. In actual testing, the above-mentioned multiple testing scenarios can coexist. For example, the OLED panel under test can be an SPR array mobile phone screen with rounded corners and punch-hole areas. The imaging colorimeter acquires the test digital image at a side viewing angle, while the lens also has a certain degree of barrel distortion. For this complex scenario, the positional parameters of the imaging colorimeter and the OLED panel under test can be adjusted to meet the detection requirements by first calibrating the image and adjusting the mechanism. Then, the SPR (Surface Reflector) array panel is made into a Uniform equivalent display by lighting the pixels in alternating rows and columns. Subsequently, the illuminated pixel digital map, avoiding the punch-hole area and rounded corner area, is selected as the reference point, and auxiliary row coordinate axes and auxiliary column coordinate axes are established along the row and column directions of the reference point, respectively. For the curved pixel rows and columns caused by barrel distortion in the edge area, the coordinates of the remaining pixel digital map are calculated by vector operation or two-dimensional surface fitting. For the offset of the light emission center caused by T-shaped distortion of the side view, the imaging pixel with the highest brightness value is searched within the detection area of ​​the digital detector and the coordinates are corrected. Finally, a corresponding digital detector is set for each effective pixel digital map, the average brightness and average chromaticity are calculated, and a luminance and chromaticity composite map corresponding to the effective physical resolution of the OLED panel under test is generated. Thus, this embodiment can still achieve accurate correspondence between physical pixels and luminance and chromaticity measurements under complex detection conditions where rotation, side view distortion, lens distortion, irregular cutting, and special pixel arrangement coexist.

[0088] By establishing a one-to-one mapping between pixel digital images and physical pixels, and utilizing a gridded positioning mechanism with bidirectional row and column auxiliary coordinates, this method effectively overcomes image geometric distortion and pixel misalignment problems caused by imaging colorimeter tilt, rotation, or lens distortion. Compared to traditional global rigid matching algorithms, this scheme can perform independent digital detection and data acquisition for each pixel digital image, significantly improving the spatial correspondence accuracy of luminance and chrominance extraction. The final generated luminance and chrominance composite image is strictly the same size as the panel's physical resolution, and the data file can be directly used for subsequent Demura compensation without conversion errors, thus solving the technical problems of luminance and chrominance data misalignment and afterimages caused by misalignment.

[0089] As an optional implementation, the OLED panel optical inspection method based on an imaging colorimeter further includes: controlling the imaging colorimeter to perform at least three non-overlapping translations relative to the OLED panel under test, with each translation distance being a multiple of the size of the imaging pixel, to achieve at least one of the following operating modes: First mode: Improve the resolution of the imaging colorimeter to enhance the spatial resolution of the detected digital image; The second mode involves measuring each physical pixel of the OLED panel under test multiple times and sampling it at least once by the imaging pixel to obtain a true-color detection digital image; the imaging pixel includes at least three types of pixels: red, green, and blue.

[0090] This embodiment utilizes a pixel interleaving algorithm to effectively increase the spatial sampling frequency of the imaging colorimeter by employing non-overlapping imaging pixel translation, enabling the imaging colorimeter to detect even minute pixel-level brightness unevenness. Furthermore, by ensuring that each physical pixel is independently measured by imaging pixels with R / G / B three-color responses, colorimetric measurement artifacts introduced by spatial deviations of the Bayer array or filters are eliminated, achieving true-color, high-fidelity luminance and chromaticity data acquisition.

[0091] For example, this embodiment uses an OptiCam C151 high-resolution imaging colorimeter equipped with an LM90 high-definition lens. This imaging colorimeter supports pixel shift, a technique that precisely translates the imaging colorimeter, shifting one or more pixel sizes at a time, and then synthesizing a high-resolution or true-color image from multiple images. Assume the OLED panel under test is a 6.7-inch AMOLED mobile phone screen with a resolution of 3200×1440, using an SPR (diamond) array. The detection target is to measure the uniformity of brightness and color across the entire screen.

[0092] The OLED panel under test is fixed on the stage of a motorized six-axis platform, and an imaging colorimeter is mounted above the platform. An industrial computer is connected to the imaging colorimeter, the six-axis platform controller, and the panel drive signal generator.

[0093] The calibration procedure was run to display a calibration image on the OLED panel under test: five white circular spots and four sets of line pairs horizontally and vertically on a black background. An imaging colorimeter acquired one frame of image, identified the coordinates of the five circular spots, and calculated the center offset ΔX = 0.35 mm, ΔY = 0.28 mm, horizontal tilt angle 0.5°, vertical tilt angle 0.3°, and rotation angle 0.1°. Simultaneously, the spatial modulation of the fourth horizontal line pair was calculated to be 0.65, and the spatial modulation of the fourth vertical line pair was calculated to be 0.62.

[0094] The industrial computer issues adjustment commands, and the electric six-axis platform first performs X and Y translational compensation for the offset, then adjusts the Z-axis height to achieve clear focus, and finally fine-tunes the tilt angles around the X and Y axes. After three iterations, the center offset is reduced to 0.05mm, all tilt angles are less than 0.1°, and the spatial modulation is improved to 0.92, meeting the positional parameters of the imaging colorimeter and the OLED panel under test to meet the detection requirements.

[0095] To obtain true-color information for each physical pixel, the pixel shift mode of the imaging colorimeter was enabled. Four images were acquired sequentially, shifting the image by one pixel each time, with non-overlapping shift trajectories to ensure the acquisition of red, blue, and green colors, resulting in four sub-images. The software then combined these four sub-images into a single high-resolution true-color image, with each physical pixel containing R, G, and B color response data.

[0096] The synthesized digital image has a resolution of 6400×2880, which corresponds to the physical pixel count of the OLED panel under test (3200×1440). Each physical pixel corresponds to approximately 2×2 imaging pixels. The initial pixel coordinates are set to the position of a bright pixel near the center of the digital image. Positioning is performed sequentially to the right along the row containing that pixel. When a dark pixel is encountered (such as a sub-pixel in the corresponding SPR pixel arrangement that does not participate in the current color), its coordinates are obtained by interpolation from the bright pixels on either side. After completing the auxiliary row and column coordinate axis positioning, the coordinates of all pixels in the digital image are calculated using a vector method. Then, for each pixel, the brightest pixel is searched within a 3×3 pixel range around its theoretical coordinates, and the coordinate positioning center is moved to that brightest pixel.

[0097] The digital detector is configured as a square with sides equal to the physical pixel pitch, approximately equal to two imaging pixels. The average luminance and CIE 1931 color coordinates are calculated for each detector. The luminance and chromaticity data of 3200×1440 pixels are saved as a CSV file, and a pseudo-color mapping image is generated.

[0098] If the OLED panel under test in this embodiment of the invention uses an SPR (Surface Reflection Per Pixel) arrangement, conventional methods cannot directly locate each sub-pixel. This embodiment employs an alternating row and column illumination method, illuminating only all red sub-pixels to form a uniform grid in space for positioning and brightness measurement; then, green and blue sub-pixels are illuminated sequentially for positioning measurements. Finally, all sub-pixel data are combined into full-pixel brightness and color data.

[0099] This invention successfully detected three OLED panels with brightness <10 cd / m². 2 There are dark spots and two uneven areas with a brightness deviation of >5%.

[0100] Secondly, embodiments of the present invention provide an optical inspection device for OLED panels based on an imaging colorimeter, such as... Figure 12 As shown, the device may include: Imaging colorimeter 201 is used to acquire digital images of the OLED panel under test. Lens 202 is mounted on imaging colorimeter 201. Its imaging circle size is larger than the CMOS target surface size of imaging colorimeter 201, and its edge relative illuminance is not less than 80%. The adjustment mechanism 203 is used to carry the imaging colorimeter 201 and / or the OLED panel to be tested, and to realize one or more combinations of translation along the X-axis, translation along the Y-axis, translation along the Z-axis, rotation around the X-axis, rotation around the Y-axis and rotation around the Z-axis; The control and image processing module 204 is electrically connected to the imaging colorimeter 201 and the adjustment mechanism 203, respectively, and the control and image processing module 204 is configured to perform any embodiment of the first aspect.

[0101] In one embodiment, the OLED panel optical inspection device based on an imaging colorimeter can refer to, for example... Figure 13 As shown, lens 202 employs an imaging circle larger than the CMOS target surface and an edge relative illumination of no less than 80%, which physically suppresses the vignetting effect at the image edges of the imaging colorimeter 201, ensuring the consistency of the dynamic range measurement across the entire image. Combined with adjustment mechanism 203 and control and image processing module 204, which incorporates the optical detection algorithm described in the first aspect, this device achieves full automation from mechanical orientation adjustment and optical imaging to data processing. This device can be directly deployed as a standardized machine on OLED mass production lines, or it can be deployed in OLED panel quality inspection facilities.

[0102] The adjustment mechanism 203 can be an electrically operated six-axis adjustment platform, and the control and image processing module 204 can be an industrial computer or industrial control computer, or other machine with control and data processing capabilities. The CMOS target surface size is the diagonal length of the effective imaging area of ​​the CMOS sensor within the imaging colorimeter. Edge relative illuminance is the ratio of the brightness of the image edge region to the brightness of the center region, usually expressed as a percentage; a higher value indicates better brightness uniformity.

[0103] In summary, the OLED panel optical inspection method and device based on an imaging colorimeter provided in this invention effectively overcomes the problems of image geometric deformation and pixel misalignment caused by the tilt, rotation, or lens distortion of the imaging colorimeter by establishing a one-to-one correspondence between pixel digital images and physical pixels, and by utilizing a gridded positioning mechanism with bidirectional auxiliary coordinates. Compared to traditional global rigid matching algorithms, this scheme can perform independent digital detection and data acquisition for each pixel digital image, significantly improving the spatial correspondence accuracy of luminance and chrominance extraction. The final generated luminance and chrominance composite image is strictly the same size as the physical resolution of the panel, and the data file can be directly used for subsequent Demura compensation without conversion errors, thereby solving the technical problems of luminance and chrominance data misalignment and afterimage retention caused by misalignment.

[0104] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0105] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0106] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0107] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0108] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0109] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0110] The above are merely specific embodiments of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. An optical inspection method for OLED panels based on an imaging colorimeter, characterized in that, The method is used to detect the brightness and color gamut of an OLED panel under test, wherein the OLED panel under test comprises multiple physical pixels; the method includes: Controlling the display of a calibration image on the OLED panel under test, and controlling the movement of the imaging colorimeter and / or the OLED panel under test through an adjustment mechanism, so that the position parameters of the imaging colorimeter and the OLED panel under test meet the detection requirements, including: Identify at least three white circular spots from the calibration image and obtain the calibration coordinates of each white circular spot. Calculate the center offset, horizontal tilt angle, vertical tilt angle, and rotation angle of the center of the calibration image relative to the field of view center of the imaging colorimeter based on the calibration coordinates. Calculate the spatial modulation based on the horizontal and vertical line pairs of the calibration image. Using the center offset, the imaging horizontal tilt angle, the vertical tilt angle, the rotation angle, and the spatial modulation degree as feedback parameters, the adjustment mechanism is driven to move the imaging colorimeter and / or the OLED panel under test; until the center offset is less than 0.2 mm, the deviations of the imaging horizontal tilt angle, the vertical tilt angle, and the rotation angle are all less than 0.2°, and the spatial modulation degree is between 0.8 and 0.99; The test image is displayed on the OLED panel under test, and the display area is imaged using the imaging colorimeter to obtain a test digital image. The test digital image includes multiple pixel digital images, each physical pixel corresponds to one pixel digital image, and one pixel digital image includes multiple imaging pixels. A pixel digital image is determined from the detected digital image as a reference point and reference coordinates are set. All pixel digital images are located along the row direction of the reference point, with the reference point as the origin. The centroid coordinates of all pixel digital images located along the row direction are arranged sequentially to form an auxiliary row coordinate axis. The auxiliary row coordinate axis is a sequence of centroid coordinates used to characterize the actual position trajectory of the physical pixels in the row of the reference point in the detected digital image. All pixel digital images are located along the column direction of the reference point, with the reference point as the origin. The centroid coordinates of the pixel digital image are arranged sequentially to form an auxiliary column coordinate axis. The auxiliary column coordinate axis is a sequence of centroid coordinates used to characterize the actual position trajectory of the physical pixel in the column where the reference point is located in the detected digital image. For the remaining pixel digital image in the i-th row and j-th column to be located, its row index i is associated with the corresponding value in the auxiliary column coordinate axis, and its column index j is associated with the corresponding value in the auxiliary row coordinate axis. Based on the corresponding values ​​in the auxiliary column coordinate axis and the corresponding values ​​in the auxiliary row coordinate axis, the coordinates of the remaining pixel digital image are calculated using a vector operation algorithm. For each pixel digital image, a corresponding digital detector is set. The average brightness and average chromaticity values ​​obtained by each digital detector are used as the brightness and chromaticity measurement values ​​of the corresponding pixel digital image. A brightness and chromaticity composite image with the same size as the resolution of the OLED panel under test is generated and saved as a brightness and chromaticity data file.

2. The optical inspection method for OLED panels based on an imaging colorimeter according to claim 1, characterized in that, Also includes: The imaging colorimeter is controlled to perform at least three non-overlapping translations relative to the OLED panel under test, with each translation distance being a multiple of the size of the imaging pixel, to achieve at least one of the following operating modes: First mode: Increase the resolution of the imaging colorimeter to improve the spatial resolution of the detected digital image; Second mode: Each physical pixel of the OLED panel under test is measured multiple times and sampled at least once by an imaging pixel to obtain a true color detection digital image; the imaging pixel includes at least three types of pixels: red, green and blue.

3. The OLED panel optical inspection method based on an imaging colorimeter according to claim 1, characterized in that, The step of locating all the pixel digital images along the row direction of the reference point with the reference point as the origin, and determining the auxiliary row coordinate axis, includes: Within a preset first search window, identify the pixel digital image corresponding to the reference point and perform a first-step search along the row direction of the physical pixels; use the first search window to identify adjacent pixel digital images along the row direction and record the centroid coordinates of the adjacent pixel digital images; using the adjacent pixel digital images as new starting points, repeat the first-step search until all identifiable pixel digital images in the row direction are located; arrange the centroid coordinates of all located pixel digital images in sequence to form the auxiliary row coordinate axis; The step of locating all the pixel digital images along the column direction of the reference point with the reference point as the origin, and determining the auxiliary column coordinate axis, includes: Within a preset second search window, the pixel digital image corresponding to the reference point is identified, and a second step search is performed along the column direction of the physical pixel; the second search window is used to identify adjacent pixel digital images along the column direction, and the centroid coordinates of the adjacent pixel digital images are recorded; the adjacent pixel digital images are used as new starting points, and the second step search is repeated until all identifiable pixel digital images in the column direction are located; the centroid coordinates of all located pixel digital images are arranged in sequence to form the auxiliary column coordinate axis.

4. The OLED panel optical inspection method based on an imaging colorimeter according to claim 3, characterized in that, The process of locating all the pixel digital images along the row direction of the reference point with the reference point as the origin to determine the auxiliary row coordinate axis, and locating all the pixel digital images along the column direction of the reference point with the reference point as the origin to determine the auxiliary column coordinate axis, includes: If a non-lit pixel digital image is encountered, the interpolated coordinates of the non-lit pixel digital image are obtained by interpolating the coordinates of the pixel digital images corresponding to two adjacent lit physical pixels.

5. The optical inspection method for OLED panels based on an imaging colorimeter according to claim 4, characterized in that, The step of locating the coordinates of the remaining pixel digital image based on the auxiliary row coordinate axis and the auxiliary column coordinate axis includes: There is an auxiliary column coordinate axis value corresponding to the row direction of the remaining pixel digital image, and there is an auxiliary row coordinate axis value corresponding to the column direction of the remaining pixel digital image. The coordinates of the remaining pixel digital image are calculated by a vector operation algorithm based on the auxiliary column coordinate axis value and the auxiliary row coordinate axis value.

6. The optical inspection method for OLED panels based on an imaging colorimeter according to claim 5, characterized in that, After locating the remaining pixel digital image according to the auxiliary row coordinate axis and the auxiliary column coordinate axis, the method further includes: Within the detection area of ​​each digital detector, the imaging pixel with the highest brightness value is searched, and the coordinates of the pixel digital map are corrected to the location coordinates of the imaging pixel with the highest brightness value, so as to correct the coordinates of the pixel digital map.

7. The OLED panel optical inspection method based on an imaging colorimeter according to claim 6, characterized in that, The method is used to detect one or more of the following detection scenarios: (a) The detected digital image has a rotation angle; (b) The detected digital image has a side-view T-shaped distortion and supports a maximum side-view angle of 40°; (c) The detected digital image exhibits barrel distortion or pincushion distortion; (d) The OLED panel under test has a punch-hole area, a notch area, or a rounded corner area; (e) The physical pixel arrangement of the OLED panel under test is either a Uniform pixel arrangement or an SPR subpixel rendering arrangement.

8. The optical inspection method for OLED panels based on an imaging colorimeter according to claim 7, characterized in that, The method further includes: When the physical pixel arrangement of the OLED panel under test is SPR arrangement, the OLED panel under test is made to present an equivalent display of Uniform arrangement by lighting up every other row and every other column. The equivalent display of all pixel digital images is located by interpolation positioning or combined positioning, and the following steps are performed: setting a corresponding digital detector for each pixel digital image, using the average brightness and average chromaticity values ​​obtained by each digital detector as the brightness and chromaticity measurement values ​​of the corresponding pixel digital image, generating a brightness and chromaticity composite image of the same size as the resolution of the OLED panel under test, and saving it as a brightness and chromaticity data file.

9. An optical inspection device for OLED panels based on an imaging colorimeter, characterized in that, include: An imaging colorimeter is used to acquire digital images of the OLED panel under test. The lens is mounted on the imaging colorimeter, and its imaging circle size is larger than the CMOS target size of the imaging colorimeter, and the relative illuminance at the edge is not less than 80%. An adjustment mechanism is used to support the imaging colorimeter and / or the OLED panel under test, and to realize one or more combinations of translation along the X-axis, translation along the Y-axis, translation along the Z-axis, rotation around the X-axis, rotation around the Y-axis, and rotation around the Z-axis; A control and image processing module is electrically connected to the imaging colorimeter and the adjustment mechanism, respectively, and the control and image processing module is configured to perform the method of any one of claims 1 to 8.

Images