Detection method, detection device and detection equipment

Abstract

The present application provides a detection method, a detection device and a detection equipment, which relate to the field of display screen detection technology. It is used to improve the disadvantage of the traditional mini LED detection method that image fusion and splicing will lead to reduced detection accuracy of the fusion area. The offset data of the first model relative to the display to be detected is determined by the first position information of the marking point of the display to be detected and the second position information of the marking point of the preset first model. The first model is corrected according to the offset data to obtain the second model. The corrected second model is aligned with the display to be detected. The second model is divided into multiple sub-models according to the position of the camera's shooting trigger point and the size of the camera's field of view. The sub-models correspond to the shooting trigger point one by one, and the image taken by the camera at the shooting trigger point is compared with the sub-model corresponding to the shooting trigger point for detection.

Description

Technical Field

[0001] The present application relates to the technical field of display screen detection, and in particular to a detection method, a detection device, and a detection equipment. Background Art

[0002] As an indispensable product in our daily lives, screens are constantly evolving with technological advancements. Higher brightness, wider color gamut, more detailed display effects, and improved power efficiency are the demands placed on screens. This has led to the emergence of sub-millimeter light-emitting diodes (miniLEDs).

[0003] MiniLEDs are a further refinement of fine-pitch LEDs, with chip sizes ranging from 50-200 microns. During miniLED panel production, due to the high precision of components, damage, foreign matter, and misalignment are prone to occur. MiniLED appearance inspection involves a gantry moving at a constant speed, with a camera mounted on the gantry taking pictures at a fixed frequency. After all images are taken, they are fused and stitched together, and the resulting image is processed to identify any issues.

[0004] However, image fusion and splicing will inevitably cause certain distortion to the image, and image fusion and splicing will lead to a decrease in the accuracy of fusion area detection. Summary of the Invention

[0005] SUMMARY OF THE INVENTION: The present invention provides a detection method, a detection apparatus, and a detection device to improve the problem that image fusion and splicing may reduce the accuracy of fusion zone detection.

[0006] To achieve the above objectives, this application adopts the following technical solutions:

[0007] In a first aspect, a detection method is provided, comprising: determining offset data of a first model relative to the display to be detected based on first position information of a marker point of the display to be detected and second position information of a marker point of a preset first model. The first model is used to indicate the distribution of chip particles in the display and the position of a camera's capture trigger point. The first model is corrected based on the offset data to obtain a second model, and the second model is aligned with the display to be detected. The second model is segmented to obtain multiple sub-models, and the sub-models correspond one-to-one with the camera's capture trigger point of the second model.

[0008] The rectified model is segmented to ensure a one-to-one correspondence between the images captured by the camera at each trigger point and the sub-models generated after the segmentation and correction. The size of each sub-model corresponds to the camera's field of view, so the chip particle information in the image should match the chip particle information in each sub-model. This segmentation process ensures that each photo corresponds to a sub-model, eliminating the need to stitch and fuse multiple images before detection. This improves the problem of inaccurate detection caused by distortion in the fused area during image processing and fusion of all photos after capture.

[0009] In a possible embodiment, the first position information includes the position coordinates of multiple marking points of the display to be detected, and the second position information includes the position coordinates of multiple marking points of the first model, and the multiple marking points of the display to be detected correspond one to one with the multiple marking points of the first model. The offset data includes the lateral offset and the longitudinal offset as well as the offset angle of the multiple marking points of the first model relative to the multiple marking points of the display to be detected, and the offset angle is the average value of the angular deviation between the boundary line of the first model and the corresponding boundary line of the display to be detected, and the boundary line is the line connecting two adjacent marking points. Based on the first position information of the marking points of the display to be detected and the second position information of the preset marking points of the first model, determining the offset data of the first model relative to the display to be detected includes: determining the lateral offset and the longitudinal offset, as well as the offset angle based on the position coordinates of the multiple marking points of the display to be detected and the position coordinates of the multiple marking points of the first model, and the number of marking points is greater than or equal to 3.

[0010] Due to screen tolerances and alignment accuracy, there may be an offset between the actual screen being inspected and the model. Compared to the high cost of a visual alignment platform, model correction is low-cost and more flexible. The horizontal and vertical offsets, as well as the angle of the offset, are calculated from the coordinates of the corresponding markers. This allows the position of the chip particles and the markers to be adjusted, ultimately achieving model correction.

[0011] In one possible implementation, determining the horizontal offset and the vertical offset includes determining the horizontal offset as the difference between the horizontal coordinate of any marking point of the first model and the horizontal coordinate of the marking point corresponding to the display to be detected. Determining the vertical offset includes determining the vertical offset as the difference between the vertical coordinate of any marking point of the first model and the vertical coordinate of the marking point corresponding to the display to be detected.

[0012] In one possible embodiment, determining the offset angle includes: determining a first slope of the first boundary line based on the position coordinates of two marking points of any first boundary line of the first model, where the first boundary line is a line connecting two adjacent marking points of the first model. Determining a second slope of the second boundary line of the display to be detected based on the position coordinates of two marking points of the second boundary line of the display to be detected, where the second boundary line is a line connecting two marking points of the display to be detected corresponding to the two marking points of the first boundary line. Determining an angular deviation between the first boundary line of the first model and the second boundary line of the display to be detected based on the first slope and the second slope. Determining an average value of the multiple angular deviations as the offset angle.

[0013] In a possible embodiment, before determining the offset data of the first model relative to the display to be detected based on the first position information of the marking point of the display to be detected and the second position information of the marking point of the preset first model, the method further includes: determining that the number of warping points in the marking points of the display to be detected is less than or equal to half of the number of marking points, and the warping point refers to the marking point where the angle between the two adjacent boundary lines is not a right angle. Eliminate the warping point in the marking point. The relative distance between the four marking points of the display in the mounting process is determined, and the existence of the warping point will destroy the relative distance between the two points adjacent to the warping point and other points. Therefore, after excluding the warping point, the calculation of the model offset will be more accurate.

[0014] In one possible embodiment, the offset data correction of the first model to obtain the second model includes: taking any marked point of the first model as a reference point, rotating the first model by an offset angle. Compensating the horizontal offset of the horizontal coordinate of the rotated first model and the vertical offset of the vertical coordinate to obtain the second model. During the model correction process, all chip particles and marked points must be corrected. First, a standard point is selected as the reference point. With the reference point as a reference, the other points are first subjected to angular deflection, and then horizontal and vertical compensations are performed. The corresponding correction is performed according to the offset data to achieve alignment of the second model with the display to be tested.

[0015] In one possible embodiment, correcting the first model based on the offset data to obtain the second model further includes: correcting the positions of the camera's capture trigger points included in the first model based on the offset data to obtain the positions of the capture trigger points of the second model. Since the second model is obtained by correcting the first model, if the capture trigger points before correction are not changed, the chip particles may not be imaged at the predetermined positions. Therefore, the positions of the capture trigger points of the first model need to be corrected. This allows the camera to appear more clearly and completely within the camera's field of view when taking pictures at the capture trigger points.

[0016] In one possible embodiment, segmenting the second model into multiple sub-models includes segmenting the model based on the positions of the capture trigger points of the second model and the camera's field of view to obtain the multiple sub-models. Segmenting the model based on the capture trigger points ensures that each segmented sub-model corresponds to a photo to be captured by the camera, facilitating subsequent detection steps.

[0017] In one possible embodiment, comparing the image captured by the camera at the capture trigger point with the corresponding sub-model includes obtaining chip particle distribution information in the image. Comparing the chip particle distribution information in the image with the chip particle distribution information in the sub-model. The quality of the display to be inspected is determined based on the degree of match between the chip particle position information in the sub-model and the chip particle position coordinate information shown in the actual image.

[0018] In a second aspect, a detection device is provided, which includes: a calculation module for determining the offset data of the first model relative to the display to be detected based on the first position information of the mark point of the display to be detected and the second position information of the mark point of the preset first model. The first position information includes the position coordinates of multiple mark points of the display to be detected, and the second position information includes the position coordinates of multiple mark points of the first model, and the multiple mark points of the display to be detected correspond one-to-one to the multiple mark points of the first model. The first model is used to indicate the chip particle distribution of the display and the position of the camera's shooting trigger point. The correction module is used to correct the first model according to the offset data to obtain a second model, and the second model is aligned with the display to be detected. The segmentation module is used to segment the second model to obtain multiple sub-models, and the sub-models correspond one-to-one to the shooting trigger points of the camera of the second model. The detection module is used to compare and detect the image taken by the camera at the shooting trigger point with the corresponding sub-model.

[0019] A third aspect provides a detection device, comprising a processor and a memory, wherein the processor is configured to execute computer instructions stored in the memory to implement the method provided in any embodiment of the first aspect.

[0020] For the technical effects of the second and third aspects, please refer to the relevant description of the first aspect above. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the structure of a detection device provided in an embodiment of the present application;

[0022] Figure 2 FIG2 is a flow chart of a detection method provided in an embodiment of the present application;

[0023] Figure 3a This is a schematic diagram of the model provided in the embodiment of the present application before correction;

[0024] Figure 3b A schematic diagram of a model correction provided in an embodiment of the present application;

[0025] Figure 3c Another model correction schematic diagram provided in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of another detection method provided in an embodiment of the present application;

[0027] Figure 5 A schematic diagram of another detection method provided in an embodiment of the present application;

[0028] Figure 6 A schematic diagram of a process for determining an offset angle is provided for an embodiment of the present application;

[0029] Figure 7 A schematic diagram of a model correction process is provided for an embodiment of the present application;

[0030] Figure 8a A schematic diagram of a shooting trigger point correction provided in an embodiment of the present application;

[0031] Figure 8b A schematic diagram of another shooting trigger point correction provided in an embodiment of the present application;

[0032] Figure 9 A schematic diagram of data processing provided in an embodiment of the present application;

[0033] Figure 10 A detection device is also provided for the embodiment of the present application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments.

[0035] In the following, the terms "first," "second," etc. are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the quantity of the technical features indicated. Therefore, a feature specified as "first," "second," etc. may explicitly or implicitly include one or more of the features.

[0036] In addition, in this application, directional terms such as "up", "down", "left", "right", "horizontal" and "vertical" are defined relative to the orientation of the components in the drawings. It should be understood that these directional terms are relative concepts. They are used for relative description and clarification, and they can change accordingly according to changes in the orientation of the components in the drawings.

[0037] The term "module" as used in this application refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic, or combination of hardware and / or software code that is capable of performing the functions associated with that element.

[0038] The detection method provided in the embodiments of the present application can be applied to a detection device, for example, Figure 1 , Figure 1 This is a schematic diagram of the structure of a detection device provided in an embodiment of the present application. Figure 1 As shown, the detection device 020 may include a processor 021 , a communication line 022 and a communication interface 023 .

[0039] Exemplarily, the detection device 020 may further include a memory 024. The processor 021, the memory 024, and the communication interface 023 may be connected via a communication line 022. The processor 021 may be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device, or any combination thereof. The processor 021 may also be other devices with processing capabilities, such as circuits, devices, or software modules.

[0040] The communication line 022 is used to transmit information between the various components of the detection device.

[0041] Communication interface 023 is used to communicate with other devices or communication networks. The communication network can be Ethernet, a radio access network (RAN), a wireless local area network (WLAN), etc. Communication interface 023 can be an interface circuit, a pin, a radio frequency module, a transceiver, or any other device capable of communication.

[0042] Memory 024 is used to store instructions. Wherein, the instruction can be a computer program for executing the detection method provided in the embodiment of the present application. Wherein, the memory 024 can be a read-only memory (ROM) or other types of static storage devices that can store static information and / or instructions, or a random access memory (RAM) or other types of dynamic storage devices that can store information and / or instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, optical disc storage, magnetic disk storage medium or other magnetic storage device, and optical disc storage includes a compact disc, a laser disc, an optical disc, a digital versatile disc, or a Blu-ray disc, etc.

[0043] It should be noted that memory 024 can exist independently of processor 021 or be integrated with processor 027. Memory 024 can be used to store instructions, program code, or data. Memory 024 can be located within or outside of detection device 020, without limitation. The processor is configured to execute the instructions stored in memory 024 to implement the detection methods provided in the following embodiments of this application.

[0044] In one example, the processor 021 may include one or more CPUs, such as Figure 1 CPU0 and CPU1 in.

[0045] As an optional implementation, the detection device 020 includes multiple processors, for example, Figure 1 In addition to the processor 021, the processor 027 may also be included.

[0046] As an optional implementation, the detection device 020 further includes an output device 025 and an input device 026. For example, the input device 026 can be a keyboard, a mouse, a microphone, a joystick, etc., and the output device 025 can be a display screen, a speaker, etc. It should be noted that the detection device 020 can be a desktop computer, a portable computer, a network server, a mobile phone, a tablet computer, a wireless terminal, an embedded device, a chip system, or a computer with a computer programmable logic controller. Figure 1 Detection equipment of similar structure 020. In addition, Figure 1 The composition structure shown in the figure does not constitute a limitation on the processing device, except Figure 1In addition to the components shown, the detection device 020 may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0047] Currently, sub-millimeter light-emitting diodes (miniLEDs) have become a key technological trend in the display industry. Due to their advantages of thin film, miniaturization, array-based design, high brightness, and low cost, miniLEDs are being widely used in various fields, including automotive panels, mobile phones, and large-size panels. As major panel manufacturers increase their investment in miniLEDs, the corresponding demand for testing is also growing.

[0048] The size of miniLED panels on the market is generally no less than 8 inches (diagonal length) and can reach up to 110 inches. This means that the total image acquisition format of the corresponding visual inspection system is relatively large. The miniLED chip particles, which serve as the main light source, are approximately 50-200 microns in size and are mounted on the printed circuit board through a mounting process. To clearly display chip particle defects, the support of a micro-optical system is required, that is, its optical resolution cannot be lower than 5μm / pixel. However, imaging systems with higher optical resolutions often have a relatively small field of view.

[0049] Before the actual photo capture process, a model is pre-established. By inputting the chip location information provided by the miniLED manufacturer, the model generates the trigger points and camera movement trajectory required for the camera to capture the photo. The starting position of the movement trajectory is called the initial position, and the ending position is called the final position.

[0050] After the model is established, a camera is used to photograph and identify the miniLED's data code. The data code includes a barcode for miniLED identification, and each miniLED has a unique data code. In actual testing, a printed circuit board (PCB) typically contains more than one miniLED. When multiple miniLEDs are present, the testing equipment plans a path by calculating the shortest motor travel required to capture all the data codes. The camera then photographs and identifies the data code, grouping and summarizing the information it reads. For example, if there are four miniLEDs on a PCB, the final summarized information would be MiniLED number one corresponding to MiniLED number one; MiniLED number two corresponding to MiniLED number two; and so on. Grouping the miniLEDs on an entire PCB allows each miniLED to be tested independently, making testing more convenient.

[0051] During the actual photography process, the miniLED is fixed in a fixed groove on the base and aligned, and each camera takes a picture of it. Each camera is installed on a gantry. The gantry is driven by a motor to move, and the cameras on the gantry also move with it, taking pictures at the corresponding shooting trigger points. During the entire photography process, a large number of images are collected. After the images are collected, they need to be fused and processed into a complete large image. This is then compared with the chip particle parameter information configured by the miniLED factory to detect whether the miniLED chip particle parameter information at the same position is consistent with the chip particle parameter information after image fusion processing. However, image fusion and splicing will inevitably lead to a certain degree of distortion in the spliced ​​image. Image fusion and splicing will cause image distortion in the fusion and splicing area, resulting in reduced detection accuracy.

[0052] In order to improve the problem of reduced detection accuracy caused by the need to splice images during the detection process, the present invention provides a detection method to improve the problem of reduced detection accuracy of the fusion area caused by image fusion and splicing. The detection method provided by the present invention can be applied to Figure 1 The detection equipment is shown.

[0053] For example, combined Figure 2 , Figure 2 The figure shows a flow chart of a detection method provided in an embodiment of the present application. The detection method provided in an embodiment of the present application includes:

[0054] S110: Determine offset data of the first model relative to the display to be detected based on first position information of a marking point of the display to be detected and second position information of a preset marking point of the first model.

[0055] The mark point can provide a reference point for the placement of all components in the display placement process. In the embodiment of the present application, a detection method provided in the embodiment of the present application is described in detail by taking the display to be detected as a miniLED as an example.

[0056] The first position information includes the position coordinates of multiple marking points of the miniLED, and the second position information includes the position coordinates of multiple marking points of the first model. The multiple marking points of the miniLED correspond one-to-one with the multiple marking points of the first model. The first model is a model established according to the design rules or circuits of the miniLED to indicate the position of each chip particle in the miniLED. A large number of chip particles are provided on the miniLED. The first model is a model established according to the design rules or circuits of the miniLED, and the distribution of its chip particles is ideal. However, due to process or other reasons during the production process of the actual miniLED, the distribution of the chip particles of the produced miniLED may differ from the ideal situation. In order to be able to detect the display to be detected, the first model also includes the movement path or movement trajectory of the camera and the position information of the camera's shooting trigger point. The camera can move according to the movement path or movement trajectory and capture an image of the display to be detected at the shooting trigger point for detecting the display to be detected.

[0057] During the production process, multiple marking points are set on the miniLED. Correspondingly, marking points are also set on the first model. The marking points of the first model correspond one-to-one to the multiple marking points of the display to be tested.

[0058] The offset data includes the lateral offset and longitudinal offset as well as the offset angle of multiple marking points of the first model relative to the multiple marking points of the miniLED, wherein the offset angle is the average value of the angular deviation between the first boundary line of the first model and the corresponding second boundary line of the miniLED.

[0059] The first boundary line is the line connecting two adjacent marking points in the first model, and the second boundary line is the line connecting two adjacent marking points in the miniLED. Since the marking points of the first model correspond one-to-one with the marking points of the miniLED, the second boundary line also has a one-to-one correspondence with the first boundary line.

[0060] The position coordinates of the multiple marking points of the miniLED in the first position information and the position coordinates of the multiple marking points of the first model in the second position information are determined by using the intersection of two adjacent boundary lines of the base as the coordinate origin and the two adjacent boundary lines as the horizontal axis and the vertical axis to establish a plane rectangular coordinate system. Figure 3a For example, Figure 3aThis is a schematic diagram of the model before correction provided in an embodiment of the present application. The dashed line represents the first model, and the solid line represents the miniLED. A rectangular coordinate system is established with point O of the base as the origin, OP as the horizontal axis, and OQ as the vertical axis. The four marking points of the miniLED are Marker Point A, Marker Point B, Marker Point C, and Marker Point D. The corresponding four marking points of the first model are Marker Point A', Marker Point B', Marker Point C', and Marker Point D'.

[0061] Taking the marking point A' and the marking point A as an example, the difference between the horizontal coordinates of the marking point A' of the first model and the marking point A of the miniLED in the plane rectangular coordinate system POQ is the horizontal offset. The difference between the vertical coordinates of the marking point A' of the first model and the marking point A of the miniLED in the plane rectangular coordinate system POQ is the vertical offset. The first boundary line is the edge line of the figure formed by the four marking points of the first model, and the second boundary line is the boundary line of the figure formed by connecting the four marking points of the miniLED. Figure 3a For example, the second boundary lines of miniLED include AB, AD, BC and CD. The first boundary lines of the miniLED corresponding to the first model are A'B', A'D', B'C' and C'D' respectively. Among them, the first boundary line A'B' corresponds to the second boundary line AB, the first boundary line A'D' corresponds to the second boundary line AD, the first boundary line B'C' corresponds to the second boundary line BC, and the first boundary line C'D' corresponds to the second boundary line CD. The offset angle is the average value of the angular deviations of AB and A'B', AD and A'D', BC and B'C', and CD and C'D'. The angular deviation refers to the angle between the first boundary line and the second boundary line, such as the angle between the first boundary line A'B' and the second boundary line AB, the angle between the first boundary line A'D' and the second boundary line AD, the angle between the first boundary line B'C' and the second boundary line BC, and the angle between the first boundary line C'D' and the second boundary line CD.

[0062] S130: Correcting the first model according to the offset data to obtain a second model, and aligning the second model with the display to be inspected.

[0063] The first model is an idealized model. In actual testing, due to inevitable errors during miniLED manufacturing, the length and width of the miniLED may differ from the ideal length and width shown in the design drawing. This error is called screen tolerance. This tolerance affects the alignment accuracy of the display under test when it is placed in the mounting groove of the base. This alignment accuracy is due to the fact that the position and angle of the miniLED in the groove may be offset due to the discrepancy in the length and width of the miniLED.

[0064] In order to avoid the influence of the mismatch between the model and the actual miniLED, the miniLED can be aligned to the position of the ideal model through the visual alignment platform. The visual alignment platform adjusts the actual position of the miniLED, but the visual alignment platform is large and costly. Model correction occupies a small volume and is low in cost. Unlike the physical adjustment of the visual alignment platform, the solution provided in the embodiment of the present application first uses a pre-set marking point as the reference point (as the center of the circle) to rotate the chip particles included in the first model by an offset angle to obtain the updated position of each chip particle. The horizontal coordinate of the updated position of the chip particle is then compensated according to the horizontal offset, and the vertical coordinate of the updated position of the chip particle is compensated according to the vertical offset. Finally, the horizontal coordinate of the reference point is compensated according to the horizontal offset, and the vertical coordinate is compensated according to the vertical offset.

[0065] For example, Figure 3b As shown, Figure 3b A schematic diagram of a model correction provided in an embodiment of the present application, wherein the dotted line represents the first model and the solid line represents the miniLED. Figure 3a In comparison, the first model selected the lower right corner mark point C' as the reference point to correct the offset angle. Figure 3c Another model correction schematic diagram provided in an embodiment of the present application, wherein the light dotted line in the figure represents the second model obtained after the first model is corrected, and the dark dotted line represents the model obtained after the first model is rotated and offset. Figure 3c based on Figure 3b Based on the above, the horizontal offset of the horizontal coordinate of the first model is compensated, and the vertical offset of the vertical coordinate of the first model is compensated. When the first model is corrected, the camera movement trajectory and the camera shooting trigger point in the first model are also corrected accordingly.

[0066] S150: Segment the second model to obtain a plurality of sub-models, where the sub-models correspond one-to-one to the shooting trigger points of the camera of the second model.

[0067] The second model includes the shooting trigger point of the camera, and the camera shoots an image of the display to be tested when it moves to the shooting trigger point, so that the shot image can be compared with the second model to detect and judge the quality of the display. In order to be able to compare and detect the chip particle distribution of the display to be tested based on the second model, for example, the solution provided in the embodiment of the present application divides the second model into multiple sub-models based on the shooting trigger point of the camera, and the size of each sub-model is the same as the size of the image field of the display to be tested taken by the camera. In this way, the image of the display to be tested taken by the camera at a certain shooting trigger point can be compared with the sub-model of the second model corresponding to the shooting trigger point. When the second model includes the distribution of the chip particles of the display, the divided sub-model also includes the chip particle distribution of the area of ​​the display to be tested corresponding to the camera field of view at the shooting trigger point, for example, including the position coordinate information of the chip particles.

[0068] S170: Compare and detect the image captured by the camera at the shooting trigger point with the corresponding sub-model.

[0069] The image captured by the camera at the trigger point shows the chip particle position coordinates. This image is then compared with the corresponding chip particle position coordinates in the sub-model. The quality of the miniLED is determined by the degree of match between the chip particle position coordinates in the sub-model and the actual image. For example, a 100% match indicates excellent quality; a 90% match indicates fair quality; and an 80% match indicates substandard quality.

[0070] Compared to the high cost of visual alignment platforms, model correction is low-cost and more flexible. The corrected model is segmented to ensure a one-to-one correspondence between the images captured by the camera at each trigger point and the sub-models generated after correction and segmentation. The size of each sub-model corresponds to the camera's field of view, so the chip particle information in the photo should match the chip particle information in each sub-model. This segmentation of the model ensures that each image corresponds to a sub-model, improving the problem of reduced inspection accuracy caused by distortion during the image stitching and fusion process after capture.

[0071] However, in some cases, if the error of the miniLED is too large, for example, if the deformation of a certain area is large, in this case, the miniLED is bound to be a substandard product and no further inspection is required. Therefore, before conducting actual inspection of the miniLED, a preliminary screening can be performed by determining whether there are warping points among the marked points.

[0072] Among them, the warping point refers to a marking point in the miniLED, and the angle between the two second boundary lines where it is located is not 90 degrees. The area where the marking point of the surface miniLED is located is greatly deformed.

[0073] For example, if the miniLED is set with 4 marking points, then if there is more than one warping point, the miniLED is determined to be an unqualified product.

[0074] For example, Figure 4 The flowchart of a detection method provided by an embodiment of the present application is shown. Before determining the offset data of the first model relative to the display to be detected based on the first position information of the marking point of the display to be detected and the second position information of the marking point of the preset first model, the method further includes:

[0075] S101: Determine that the number of warping points among the marking points of the display to be inspected is less than or equal to half the number of marking points, where the warping points refer to marking points where the angle between two adjacent boundary lines is not a right angle.

[0076] During the mounting process, the display is affixed with four marking points. For example, if the miniLED display is to be tested, there are four marking points A, B, C, and D. The position coordinates of the marking point A are known to be (x1, y1), the position coordinates of the marking point D are known to be (x2, y2), and the position coordinates of the marking point B are known to be (x3, y3). The slope k3 of the miniLED boundary line AD is (y2-y1) / (x2-x1), and the slope k4 of the miniLED boundary line AB is known to be (y3-y1) / (x3-x1). If the tangent value of the angle formed by the boundary lines AD and AB is a valid value, that is, (k3-k4) / (1+k3*k4) is a valid value, then A is determined to be a warping point.

[0077] S102: Eliminate warping points in the marking points.

[0078] When there is a warping point, the above offset angle should be determined as the average of the two angle deviations. Taking the mark point A as the warping point, the angle deviation α3 between the miniLED boundary line BC and the first model boundary line B'C' and the angle deviation α4 between the miniLED boundary line CD and the first model boundary line C'D' are calculated. The final offset angle is (α 3+ α4) / 2.

[0079] When there is a warping point, the angle formed by the warping point and the boundary line connecting the two adjacent points is not perpendicular. However, the relative distances between the four marking points in the display mounting process are fixed. That is to say, if the warping point is selected as the calculation point for the lateral and longitudinal offsets, it will be inconsistent with the offsets of the other three points, so it is discarded.

[0080] After removing the warping points, the offset data of the first model relative to the miniLED is determined based on the remaining marking points of the miniLED and the corresponding marking points of the first model. For example, Figure 5 A schematic diagram of another detection method provided in the embodiment of the present application is shown. Figure 5 , S110 includes:

[0081] S110 - a : Determine a lateral offset according to a difference between a horizontal coordinate of any marking point of the first model and a horizontal coordinate of a marking point corresponding to the display to be detected.

[0082] The difference between the horizontal coordinates of the marking point of the first model and the horizontal coordinates of the marking point corresponding to the miniLED is determined as the lateral offset. For example, the position coordinates of the miniLED marking point A are (x1, y1), and the corresponding position coordinates of the marking point A' of the first model are (x1', y1'). The lateral offset is x1'-x1. The relative distances between the four marking points in the display mounting process are fixed, that is, after eliminating the warping point, the offsets of the remaining three marking points are consistent. Therefore, the lateral offset of all marking points can be determined by selecting the coordinates of any marking point for calculation.

[0083] S110 - b: Determine a lateral offset according to a difference between a vertical coordinate of any marking point of the first model and a vertical coordinate of a marking point corresponding to the display to be detected.

[0084] The longitudinal offset is determined as the difference between the vertical coordinates of the marker point on the first model and the corresponding marker point on the miniLED. For example, if the coordinates of marker point A on the miniLED are (x1, y1), and the coordinates of marker point A' on the first model are (x1', y1'), then the longitudinal offset is y1'-y1.

[0085] S110 - c : Determine an offset angle according to the position coordinates of the plurality of marking points of the first model and the position coordinates of the corresponding plurality of marking points of the display to be detected.

[0086] For example, the four marking points of the miniLED are marked as A, B, C, and D. The corresponding four marking points of the first model are marked as A', B', C', and D'; the boundary lines of the miniLED are AB, AD, BC, and CD. For example, the boundary lines of the first model corresponding to the miniLED are A'B', A'D', B'C', and C'D'. The offset angle is the average of the angular deviations between AB and A'B', AD and A'D', BC and B'C', and CD and C'D'.

[0087] For example, the embodiment of the present application provides an example of determining the offset angle. Figure 6 , S110-c includes:

[0088] S110-c-1: Determine a first slope of any first boundary line according to the position coordinates of two marking points of the first model, where the first boundary line is a line connecting two adjacent marking points of the first model.

[0089] S110-c-2: Determine the second slope of the second boundary line of the miniLED based on the position coordinates of the two marking points of the second boundary line of the miniLED, where the second boundary line is the line connecting the two marking points of the miniLED corresponding to the two marking points of the first boundary line.

[0090] S110-c-3: Determine an angular deviation between a first boundary line of the first model and a second boundary line of the miniLED based on the first slope and the second slope.

[0091] S110-c-4: Determine an average value of the multiple angle deviations as the offset angle.

[0092] Take the adjacent markers A and D of a miniLED as an example. The coordinates of marker A are (x1, y1), and the coordinates of marker D are (x2, y2). The second boundary line of the miniLED formed by markers A and D is AD.

[0093] A first slope of the first boundary line A'D' is determined, wherein the first slope k1 is (y2'-y1') / (x2'-x1').

[0094] The coordinates of the first model's marker point A', corresponding to adjacent marker points A and D of the miniLED, are (x1', y1'). The coordinates of the first model's marker point D' are (x2', y2'). The first boundary line of the first model formed by marker points A' and D' is A'D'. The first boundary line A'D' of the first model corresponds to the second boundary line AD of the display.

[0095] The second slope k2 of the second boundary line AD is determined to be (y2-y1) / (x2-x1).

[0096] The tangent value tanα of the angle deviation α1 is equal to (k1−k2) / (1+k1*k2), and the value of the angle deviation α1 is obtained by an inverse tangent function.

[0097] The above angle deviation is only the angle deviation between the miniLED boundary line AD and the first model boundary line A'D'. The same calculation method can also be used to obtain the angle deviation α2 between the miniLED boundary line AB and the first model boundary line A'B', the angle deviation α3 between the miniLED boundary line BC and the first model boundary line B'C', and the angle deviation α4 between the miniLED boundary line CD and the first model boundary line C'D'. The final offset angle α is (α 1+ α 2+ α 3+ α4) / 4.

[0098] After obtaining the offset data, the first model is corrected so that the corrected second model is aligned with the miniLED. The following is an example of how to correct the first model. Figure 7 As shown, S130 includes:

[0099] S130-a: Using any marked point of the first model as a reference point, rotate the first model by an offset angle.

[0100] S130 - b : Compensating the horizontal offset of the horizontal coordinate and the vertical offset of the vertical coordinate of the rotated first model to obtain a second model.

[0101] The first model is corrected based on the offset data, and the corrected positions of the various marking points and chip particles of the first model are calculated based on the offset data to obtain a second model. The second model has the same structure as the first model. In the embodiment of the present application, the offset amount of the first model relative to the display to be tested is determined as the offset data, and then the first model is corrected based on the offset data to obtain a corrected first model, and the corrected first model is aligned with the display to be tested. For ease of distinction, in the embodiment of the present application, the corrected first model is referred to as the second model.

[0102] Offset data includes lateral offset, longitudinal offset, and offset angle. To calibrate the first model based on this offset data, a reference point is selected, each chip is rotated by the offset angle, and then the lateral and longitudinal offsets are compensated. After calculating the offset data and adjusting the chip position, the detection equipment will evaluate the offset data to see if the lateral and longitudinal offsets of the chip are consistent with the calculations. If they are not consistent, the correction has failed, and the chip needs to be repositioned and moved again.

[0103] For example, combined with Figure 3b and Figure 3c First, rotate the first model using any marked point as a reference point and offset the angle. The rotated model will align with the miniLED at the same angle, but there will still be a positional deviation. Therefore, after the offset rotation, the lateral and vertical offsets are compensated. The following is a calculation method for model correction.

[0104] For example, if the position of a chip particle in the first model is (x i ,y i ), the reference point is (x s ,y s ), after correction, the corresponding chip particle position in the second model is (x o ,y o )

[0105] Then, the calculation method of the chip particle position in the second model is as follows: the coordinates of the initial point (x i ,y i ), coordinates of the reference point (x s ,y s ), output point coordinates (x o ,y o ); Taking the rotation angle α as an example, the horizontal offset is x1 and the vertical offset is y1, then:

[0106] x o =[(x s -x i )×cos(α)-(y s -y i )×sin(α)]+x1

[0107] y o =[(x s -x i )×sin(α)+(y s -y i )×cos(α)]+y1

[0108] After calibrating the positions of each chip particle and the reference point, the reference point needs to be compensated horizontally and vertically.

[0109] x s =x s +x1

[0110] y s =y s +y1

[0111] For example, the position of the shooting trigger point can be determined based on the corrected second model. Figure 8a As shown, the triangles in the figure represent the chip particles in miniLED. Figure 8a The triangle m and triangle n in the figure represent the chip particle positions under the first model. Figure 8b The triangle m and triangle n in the figure represent the chip particle positions under the second model. Since the position of the shooting trigger point is also adjusted after the model is corrected, as shown in the figure below: Figure 8a , the first camera and the second camera both take pictures at the trigger point T1. But after correction, Figure 8b As shown, the first camera takes a picture at the trigger point T3, and the second camera takes a picture at the trigger point T2. At this time, the first camera and the second camera are not in a vertical line, so the position of the camera needs to be adjusted.

[0112] For example, if the first camera has coordinates (a1, b1) and the second camera has coordinates (a2, b2), the cameras only need to be offset laterally. The required displacement for the first camera is a1 - ((a1 + a2) / 2); the required displacement for the second camera is a2 - ((a1 + a2) / 2).

[0113] The second model is segmented according to the camera's field of view at the shooting trigger point, and the second model is segmented into multiple sub-models. The sub-models correspond one-to-one to the photos taken by the camera at the shooting trigger point.

[0114] For example, the camera shooting trigger point is (ai, bi) and the camera's field of view is d*d. Then the second model is divided according to the shooting trigger point and the field of view, and the area of ​​the obtained sub-model is: the area of ​​the second model defined by the shape surrounded by the four points (ai-d, bi+d), (ai+d, bi+d), (ai+d, bi-d), and (ai-d, bi-d).

[0115] For example, taking the camera field of view size of 10*10 as an example, the size of the sub-model corresponding to the camera trigger point (ai,bi) is the area surrounded by the four points (ai-10,bi+10), (ai+10,bi+10), (ai+10,bi-10), and (ai-10,bi-10).

[0116] After the model is split, the camera can move according to the movement trajectory planned by the second model. Before starting to move, the shooting trigger points required by each camera are sorted and sent to the motion control. After the motion control is successfully set up, the motor drives the gantry, and the camera on the gantry moves accordingly to take pictures at each shooting trigger point until it moves to the target position of the movement trajectory. After reaching the target position, the number of images taken by each camera is checked. If the number of pictures taken by the camera is incorrect, an alarm is issued and the data is discarded. If the number of pictures taken by the camera is correct, it is determined whether the target position is the end position of the movement trajectory planned by the second model. If not, the number of pictures taken at the next target position is calculated until it moves to the end position. After moving to the end position, the chip particle distribution of the photo is compared with the chip particle distribution of the sub-model. After waiting for the test results to be completed, upload them to the cache database.

[0117] Before comparing and testing the image captured by the camera at the trigger point with the corresponding sub-model, a filtering step is also included. Since the chip particles in the camera's field of view may overlap when the camera is moving, for example, two photos capture the same chip particle, but the imaging position of the chip particle in the two photos is different. The chip particle may be completely captured in the first photo, but since the camera moves with the movement of the gantry, the same chip particle in the second photo may move to the edge of the camera's field of view, or even incomplete chip particles may be captured. The clarity varies during mobile photography. Through the data filtering method provided in the embodiment of the present application, the test result data can be filtered to obtain high-quality pictures, making the test results more accurate.

[0118] For example, a grayscale image of the detected chip particle image in the red channel is used. The image is first filtered using the second-order derivative of the Laplace calculation to calculate the variance value of the image (this value is first converted into a percentage score based on the maximum and minimum values, and then the score is weighted with a weight of 0.6). Then, the Sobel operator is used to calculate the horizontal and vertical gradients of the chip particle edge (the processing method is the same as above, and the weight of this value is 0.2 for both horizontal and vertical directions). The total score is 58, 69, and 80 in the ranges of 0-50, 51-85, and 86-100, respectively.

[0119] The imaging position is calculated as follows Figure 9As shown, the camera's field of view is divided into three zones: edge, middle, and center, with scores of 2, 12, and 20, respectively. The inspection results are obtained based on the defect type (OK, other defects except those not identified (NG), and unidentified) fed back by the visual inspection algorithm. The coordinates of the chip particles are calculated based on the coordinates of the chip particles actually identified in the camera's field of view using the formula (x2-x1). 2 +(y2-y1) 2 Calculate the square of the distance from the center and then use it to filter the data.

[0120] Exemplarily, a possible data filtering method is as follows: first, delete all detection results that are unrecognized; then calculate the score of the imaging position and the image clarity. The two scores are summed to form the final score of the position; traverse all detection results and store the chip particle information, including the chip particle name, particle position coordinates and final score, in the hash table in turn. In the case where the chip particle name is the same but the final score is different, the one with the higher final score is overwritten in the hash table. If the final scores are the same, calculate the straight-line distance from the center point, use the data with the shorter distance and overwrite it. In the end, a hash table without any duplication and with the optimal shooting position can be obtained. The hash table can quickly retrieve the information corresponding to the chip particle name by looking up the chip particle name, including the coordinates of the chip particle, and quickly locate the camera and the shooting trigger point, thereby facilitating the sub-model comparison test.

[0121] After data filtering, the highest-quality photos are selected. The detection equipment identifies the chip particle locations in the image and compares them with the chip particle locations in the corresponding sub-model to determine the degree of match. The quality of the miniLED is determined based on the degree of match between the chip particle location information in the sub-model and the chip particle location coordinates shown in the actual image. For example, a 100% match indicates excellent quality; a 90% match indicates average quality; and an 80% match indicates substandard quality.

[0122] The present application also provides a detection device 200, such as Figure 10 As shown, the apparatus 200 includes a calculation module 201 , a correction module 202 , a segmentation module 203 and a detection module 204 .

[0123] Calculation module 201 is configured to determine offset data of the first model relative to the display to be inspected based on first position information of marker points on the display to be inspected and second position information of marker points of a preset first model. The first position information includes the position coordinates of multiple marker points on the display to be inspected, and the second position information includes the position coordinates of multiple marker points on the first model. The multiple marker points on the display to be inspected correspond one-to-one with the multiple marker points on the first model. The first model is used to indicate the distribution of chip particles on the display and the location of the camera's capture trigger point.

[0124] The correction module 202 is configured to correct the first model according to the offset data to obtain a second model, and align the second model with the display to be inspected.

[0125] The segmentation module 203 is used to segment the second model into multiple sub-models, and the sub-models correspond one-to-one to the shooting trigger points of the camera of the second model.

[0126] The detection module 204 is used to compare and detect the image captured by the camera at the capture trigger point with the corresponding sub-model.

[0127] In the above embodiments, all or part of the embodiments may be implemented using software, hardware, firmware, or any combination thereof. When implemented using a software program, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0128] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of this application and include common knowledge or customary techniques in the art that are not disclosed herein.

[0129] Finally, it should be noted that the above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A detection method, characterized in that: The method comprises: Determining offset data of the first model relative to the display to be inspected based on first position information of a marking point on the display to be inspected and second position information of a marking point on a preset first model; the first model is used to indicate chip particle distribution of the display to be inspected and the position of a camera shooting trigger point; Correcting the first model according to the offset data to obtain a second model, wherein the second model is aligned with the display to be inspected; Splitting the second model into a plurality of sub-models, wherein the sub-models correspond one-to-one to the shooting trigger points of the camera of the second model; The image captured by the camera at the shooting trigger point is compared with the corresponding sub-model for detection: specifically, the chip particle distribution of the image is compared with the chip particle distribution of the sub-model.

2. The detection method according to claim 1, wherein The first position information includes the position coordinates of the multiple marking points of the display to be detected, and the second position information includes the position coordinates of the multiple marking points of the first model, and the multiple marking points of the display to be detected correspond one-to-one to the multiple marking points of the first model; the offset data includes the lateral offset and the longitudinal offset of the multiple marking points of the first model relative to the multiple marking points of the display to be detected, and the offset angle is the average value of the angular deviation between the boundary line of the first model and the corresponding boundary line of the display to be detected, and the boundary line is the line connecting two adjacent marking points; The determining, based on the first position information of the marking point of the display to be detected and the second position information of the marking point of the preset first model, the offset data of the first model relative to the display to be detected includes: The lateral offset, the longitudinal offset, and the offset angle are determined according to the position coordinates of the multiple marking points of the display to be detected and the position coordinates of the multiple marking points of the first model, and the number of the marking points is greater than or equal to 3.

3. The detection method according to claim 2, characterized in that Determining the lateral offset and the longitudinal offset includes: Determine the difference between the horizontal coordinate of any marking point of the first model and the horizontal coordinate of the marking point corresponding to the display to be detected as the horizontal offset; The difference between the longitudinal coordinate of any marking point of the first model and the longitudinal coordinate of the marking point corresponding to the display to be detected is determined as the longitudinal offset.

4. The detection method according to claim 2, characterized in that Determining the offset angle includes: determining a first slope of any first boundary line of the first model according to the position coordinates of two marking points of the first boundary line, where the first boundary line is a line connecting two adjacent marking points of the first model; determining a second slope of a second boundary line of the display to be detected based on the position coordinates of the two marking points of the second boundary line of the display to be detected, where the second boundary line is a line connecting the two marking points of the display to be detected corresponding to the two marking points of the first boundary line; Determine an angular deviation between a first boundary line of the first model and a second boundary line of the display to be detected according to the first slope and the second slope; An average value of a plurality of the angular deviations is determined as the offset angle.

5. The detection method according to any one of claims 1 to 4, characterized in that Before determining the offset data of the first model relative to the display to be detected based on the first position information of the marking point of the display to be detected and the second position information of the marking point of the preset first model, the method further includes: Determining that the number of warping points among the marking points of the display to be inspected is less than or equal to one-half of the number of the marking points, wherein the warping points refer to marking points where the angle between two adjacent boundary lines is not a right angle; The warping points in the marking points are removed.

6. The detection method according to any one of claims 1 to 5, characterized in that Correcting the first model according to the offset data to obtain a second model includes: Taking any of the marking points of the first model as a reference point, rotating the first model by the offset angle; The second model is obtained by compensating the horizontal offset of the horizontal coordinate of the rotated first model and compensating the vertical offset of the vertical coordinate.

7. The detection method according to claim 6, characterized in that Correcting the first model according to the offset data to obtain a second model further includes: The positions of the shooting trigger points of the cameras included in the first model are corrected based on the offset data to obtain the positions of the shooting trigger points of the second model.

8. The detection method according to any one of claims 6 to 7, characterized in that: Splitting the second model into a plurality of sub-models comprises: The second model is divided according to the positions of the shooting trigger points of the second model and the field of view of the camera to obtain a plurality of sub-models.

9. A detection device, characterized in that: The device comprises: a calculation module, configured to determine offset data of the first model relative to the display to be inspected based on first position information of a marking point of the display to be inspected and second position information of a marking point of a preset first model; the first model is used to indicate chip particle distribution of the display and the position of a camera shooting trigger point; a correction module, configured to correct the first model according to the offset data to obtain a second model, wherein the second model is aligned with the display to be inspected; a segmentation module, configured to segment the second model into a plurality of sub-models, wherein the sub-models correspond one-to-one to the shooting trigger points of the camera of the second model; A detection module, configured to compare and detect the image captured by the camera at the capture trigger point with the corresponding sub-model; The detection module is specifically used to obtain the particle distribution in the image, and compare the chip particle distribution in the image with the chip particle distribution of the sub-model.

10. A detection device, characterized in that: The detection equipment includes: processor and memory; The processor is configured to execute computer instructions stored in the memory to implement the method according to any one of claims 1 to 8.

Images