KR20230060250A - Ink-jet type automatic defect repair method for display panel using artificial intelligence - Google Patents

Abstract

A first step of providing a display panel for inspecting the presence or absence of defects; a second step of photographing the upper surface of the panel and generating an input image; A third step of constructing a learning model for repair based on artificial intelligence (AI) using the input image; If there is a defect, a fourth step of determining the defect by analyzing information about the defect; If it is determined that the defect is repairable, a fifth step of repairing the panel by applying a repair recipe corresponding to the defect; an automatic defect repair method for an inkjet display panel using artificial intelligence (AI). provides

Description

Automatic defect repair method for inkjet type display panel using artificial intelligence (AI)

The present invention relates to a defect repair method for a display panel, and more particularly, to an inkjet method for automatically correcting defects using an artificial intelligence-based algorithm.

Defects detected in a display panel inspection process are repaired through a repair process. To this end, the operator manually classifies the defect detected in the inspection process, determines a repair method suitable for the defect, and proceeds with the repair. However, this method has problems in that the accuracy of repair varies depending on the operator's proficiency and is inefficient in terms of cost and productivity.

In addition, machine vision inspection systems such as automatic optical inspection have been introduced to compensate for the problems of manual inspection. However, there was a limitation in that the performance depended on how well the engineer made the rules according to the characteristics of the process. In addition, when the shape of the defect is rather complex, there is a problem in that the detection accuracy is lowered.

In addition, among display panels, especially OLED, μ-LED, etc., there is a problem in that it is not easy to apply inkjet repair. In particular, there was a difficulty in realizing the line width of the ink due to problems such as misalignment of the inkjet nozzle.

Republic of Korea Patent Registration No. 10-2043664 (2019. 11. 06. Registration) Auto repair system and method for detecting and repairing defects in a board Korean Patent Registration No. 10-1681947 (registered on November 28, 2016) Method and apparatus for classifying defects of TTF board

Embodiments of the present invention have been made to solve the above problems, and are intended to provide an automatic defect repair method of a display panel having an artificial intelligence-based inkjet method.

In addition, the object of the present invention is to automatically analyze the type of defect, select a suitable repair recipe, and then automatically repair it.

In addition, based on artificial intelligence, errors in the position of inkjet nozzles are to be corrected. In addition, we want to provide a series of processes that can effectively repair various defects.

Embodiments of the present invention to solve the above problems, a first step of providing a display panel for inspecting the existence of defects; a second step of photographing the upper surface of the panel and generating an input image; A third step of constructing a learning model for repair based on artificial intelligence (AI) using the input image; If there is a defect, a fourth step of determining the defect by analyzing information about the defect; If it is determined that the defect is repairable, a fifth step of repairing the panel by applying a repair recipe corresponding to the defect; an automatic defect repair method for an inkjet display panel using artificial intelligence (AI). provides

Between the second step and the third step,

The method may further include grouping the first feature group of defects detected in the panel and building an inference module for repair based on a rule for analyzing the first feature group.

In the fourth step, it is preferable to determine the final type of defect using the analysis data obtained from the repair learning model and the repair inference module.

It is preferable that the learning model for repair uses, as learning data, a second feature group including 1) the location of a layer with a defect, 2) a type of defect, and 3) a region with a defect detected from a defect.

In the second step, it is preferable to divide the input image into at least one unit area image having a predetermined size.

Preferably, in the fourth step, information on the defect is detected by comparing each pattern of the unit area image and the reference image.

The fourth step may include a 4-1 step of transmitting the defect to a data server unit when the defect is detected; It is preferable to include; a 4-2 step of determining the defect as one of repairable, non-repairable and unknown.

When the learning model learns the second feature group for the defect, it is preferable to update the repair recipe.

In the fifth step, it is preferable to automatically move the tip of an inkjet nozzle for ejecting repair ink in response to the repair recipe.

According to the problem solving means of the present invention as described above, various effects including the following can be expected. However, the present invention is not established when all of the following effects are exerted.

An embodiment of the present invention may provide an automatic defect repair method for a display panel having an artificial intelligence based inkjet method.

In addition, this embodiment automatically analyzes the type of defect, selects a suitable repair recipe, and then automatically repairs it.

Also, according to the present embodiment, an error in the position of an inkjet nozzle can be corrected based on artificial intelligence. In addition, this embodiment can provide a series of processes capable of effectively repairing various defects.

1 is a flowchart illustrating an automatic defect repair method for a display panel according to an embodiment of the present invention.
FIG. 2A is a diagram explaining the input image and unit area image of FIG. 1, and FIG. 2B is a diagram showing a plurality of regions and defects formed in the unit area image (unit pixel).
In relation to the embodiment of FIG. 1 , FIG. 3A is an exemplary diagram for extracting a region where a defect exists, FIG. 3B is an example diagram showing a location of a layer where a defect exists, and FIG. 3C is an example showing types of defects. floor plan.
FIG. 4 is a flowchart illustrating a process of selecting a repair recipe for a defect occurring in the unit region image of FIG. 1;
5A is a view for correcting the X-axis and Y-axis positions of the tip of the inkjet nozzle of FIG. 1 to the correct position;
FIG. 5B is a view showing a process for correcting the Z-axis position of the front end of the inkjet nozzle of FIG. 1 to the correct position.

In order to fully understand the configuration and effects of the present disclosure, preferred embodiments of the present disclosure will be described with reference to the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below, and may be implemented in various forms and various changes may be made. Hereinafter, in the description of the present invention, detailed descriptions thereof will be omitted if it is determined that the related known functions may unnecessarily obscure the subject matter of the present invention as obvious matters to those skilled in the art.

Terms such as first and second may be used to describe various components, but the components should not be limited by the terms. The terms may only be used for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present disclosure.

In this application, the terms "include" or "have" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Terms used in this application are only used to describe specific embodiments and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. Terms used in the embodiments of the present disclosure may be interpreted as meanings commonly known to those skilled in the art unless otherwise defined.

Hereinafter, specific embodiments of the present invention will be described in detail with reference to the drawings.

1 is a flowchart illustrating an automatic defect repair method for a display panel according to an embodiment of the present invention, FIG. 2A is a diagram explaining an input image and a unit area image of FIG. 1, and FIG. 2B is a unit area image ( It is a diagram showing a plurality of regions and defects formed in a unit pixel). In relation to the embodiment of FIG. 1 , FIG. 3A is an exemplary view for extracting a region where a defect exists, FIG. 3B is an example view showing a location of a layer where a defect exists, and FIG. 3C is an example view showing the type of defect. This is an example drawing. FIG. 4 is a flowchart showing a process of selecting a repair recipe for a defect occurring in the unit region image of FIG. 1 .

Referring to FIGS. 1 to 4 , an automatic defect repair method for an inkjet display panel using artificial intelligence (AI) according to an embodiment of the present invention may include first to fifth steps. This embodiment can detect and classify defects generated in the display panel using artificial intelligence. In addition, the present embodiment can automatically repair a defect by providing a repair recipe corresponding to the defect. On the other hand, since artificial intelligence technology can be implemented through already known artificial intelligence-related machine learning algorithms, a detailed description thereof will be omitted.

First, the first step is a step of providing a display panel for inspecting whether or not there is a defect. (s10) The present embodiment is characterized in that artificial intelligence (AI) is applied to defect repair of the display panel. Here, the display panel may be any panel including electrode wiring, such as LCD, OLED, Mini-LED, and μ-LED. Meanwhile, the panel includes a plurality of unit pixels. In addition, each unit pixel may be composed of a plurality of layers. The layer may be, for example, an active (ACT) layer, a gate layer, a source and drain (SD) layer, and the like.

Meanwhile, different zones according to a plurality of patterns may be formed in a unit pixel. For example, a first zone (zone ①) to a fourth zone (zone ④) may be preset in the unit pixel. In this case, the first region may be, for example, a region where a gate line is formed.

A defect may occur in at least one or more regions of unit pixels. A defect may occur in a region where a plurality of regions intersect. These defects include, for example, 1) an open defect caused by the metal layer not being formed on some of the metal layer lines, 2) a short defect caused by the metal layer being connected between the metal layer lines, and 3) a defect caused by particles depositing on the metal layer. Particle defects and the like may be included. On the other hand, particles exist in a three-dimensional shape, but the area of a particle detected in a unit pixel is measured on the plane and can be defined as the size of the plane area.

Also, defects can be formed on any particular layer. For example, a defect may occur when a particle settles on an active layer of any one unit pixel.

Next, the second step is to photograph the top surface of the panel and generate an input image (s20). To this end, the panel can be transferred to a predetermined location (preset location). Then, the camera module acquires an image by photographing the panel for a preset time. At this time, the image may be formed in a size enlarged by a certain magnification compared to the original. Then, the image generator may automatically detect the panel in the image and automatically generate an input image for the panel area. The input image may be extracted based on a specific time and a specific frame from the video.

Also, in the second step, it is preferable to divide the input image into at least one unit area image having a predetermined size. The unit area recognizing unit 100 recognizes the input image as each unit area image. Here, the unit area image means an image for a unit pixel. This makes it possible to specify in which unit pixel a defect exists.

Meanwhile, the unit area recognizing unit 100 may recognize each layer existing in the unit area image. In this case, the layers may be distinguished from each other by, for example, different colors. This makes it possible to detect in which layer the defect exists, so that the location can be specified more specifically.

In addition, in the process of analyzing a defect on a panel, an input image and a unit area image are stored in a memory unit. In addition, a defect-free reference image serving as a criterion for defect determination is also pre-stored in the memory unit.

Next, the third step is to build an artificial intelligence (AI)-based learning model 210 for repair using the input image. For selection, etc., first, a process of building a learning model 210 for repair using learning data is required. The learning model 210 for repair may use a conventional deep learning technique. Deep learning refers to a network composed of two to three or more hidden layers, and is an artificial intelligence technology in which the connections between neurons are very deep by imitating the human brain.

The learning data may be a data set of input images accumulated from the past. The input image, which was the subject of judgment, is incorporated into the learning data after the judgment on the defect. Training data continuously increases due to the addition of input images. In addition, the learning model 210 for repair may be continuously updated through real-time learning. As a result, the learning model according to an embodiment may update a recipe for repair when learning the second feature group for the defect.

The learning model 210 for repair may use the second feature group including 1) the location of a layer where a defect exists, 2) the type of defect, and 3) a region where a defect exists as training data. In this embodiment, among the characteristics of defects, in particular, the second feature group is determined through the learning model 210 for repair. Here, the region where the defect exists may be indicated by the maximum size or average size of the defect size. The present area of a defect can be expressed as a standard deviation of the defect size. In addition, the second feature group may further include a material of a layer having defects.

Meanwhile, in the present embodiment, between the second and third steps, a rule-based repair inference module 220 that analyzes the first feature group by grouping the first feature group of defects detected in the panel A step of building a may be further included. Here, the first feature group may include the brightness of the defect, the thickness of the defect, and the like. The brightness of defects can be displayed as minimum brightness, maximum brightness, and average brightness. Also, the brightness of a defect can be expressed as a standard deviation of brightness. To this end, information about the brightness of the defect and the thickness of the defect according to the shape of the defect may be set in advance. Meanwhile, the unit area recognizing unit 100 may infer the brightness of defects and the thickness of defects in the unit area image.

In this embodiment, the reasoning module 220 for repair may determine the type of defect by using distance conversion information obtained by converting brightness information of a defect into a distance unit as an element. For example, the reasoning module 220 for repair may classify a defect as a deep and thin defect as the distance conversion value constituting the distance conversion information is low.

Next, the fourth step is a step of determining the defect by analyzing information about the defect, if there is a defect. (S40) In this embodiment, the fourth step is the repair learning model 210 and the repair reasoning module. The type of the final defect may be determined using each analysis data obtained in step 220 . That is, defects according to one embodiment are characterized in that they are finally classified by integrated information included in the first feature group and the second feature group. In the final decision process, each application weight for the repair learning model 210 and the repair inference module 220 may be set in advance.

To this end, in the fourth step, defect information is detected by comparing each pattern of the unit area image and the reference image. To this end, each image may undergo a pre-processing process in advance. Areas according to a plurality of patterns may be respectively set in the unit area image. For example, each region where a gate line, data line, etc. are located can be set as a pattern. Through this, the present embodiment can provide a recipe for repair suitable for the type of defect present in each pattern.

The defect determination unit 200 first compares the unit area image and the reference image for each pattern corresponding to each other to detect the existence of a defect. At the same time, the defect determination unit 200 may determine the type of defect through the repair learning model 210 and the repair inference model. In contrast, the repair learning model 210 may determine the location of the defect. More specifically, the learning model 210 for repair may determine which unit area image of the defect exists, which area the defect exists in, which layer the defect exists in, and the like.

In the fourth step, when a defect is detected, the 4-1 step of transmitting the defect to the data server unit and the fourth step of determining the defect as one of 1) repairable determination, 2) unrepairable determination, and 3) unknown determination. -2 steps may be included. The defect determination unit 200 may determine the type of defect as one of a repairable determination, an unrepairable determination, and an unknown determination.

Step 5 is to repair the panel by applying a repair recipe corresponding to the defect if it is determined that the defect can be repaired. (S50) The repair recipe includes at least the strength of the repair laser beam and the irradiation of the repair laser beam. Each set value for the position, the irradiation time of the laser beam for repair, and the injection amount of the repair inkjet may be included. The recipe for repair may vary depending on the type of defect. That is, the recipe for repair may vary according to patterns, types of defects, locations of defects, and the like. A laser beam for repair may be irradiated through a laser unit. A laser unit according to an embodiment is operated by a plurality of organically coupled components. On the other hand, the laser unit according to an embodiment can be formed by combining known technologies, and detailed descriptions thereof will be omitted below.

If the defect is judged to be repairable, the reasoning module for repay sets a boundary for the processing area including the defect. The defect may exist in any one area (pattern) of the unit area image. Alternatively, defects may span multiple zones. At this time, the defect and the vicinity where the defect exists are exposed using a mask.

Unlike this, the present embodiment may further include a step 4-3 of transmitting the defect to the data server unit when the defect is determined to be unknown. This is to reclassify the defect. In the case of a defect that is first judged as an ignorance decision, it can be re-determined as either a fatal defect or a non-fatal defect. And, in the case of a fatal defect, a numerical value for the degree of reliability may be additionally calculated.

On the other hand, in the case of a defect of ignorance, when the degree of reliability is too low, an operator's decision may intervene. That is, information on defects can be specified with data input by the operator. At this time, the information related to the second feature group may be updated as learning data of a learning module for repay.

Further, the fifth step is characterized in that the front end of the inkjet nozzle 300 for ejecting the repair ink is automatically moved in response to the repair recipe. 5A is a diagram for correcting the X-axis and Y-axis positions of the tip of the inkjet nozzle 300 of FIG. 1 to the correct position, and FIG. 5B is a process for correcting the Z-axis position of the tip of the inkjet nozzle 300 of FIG. 1 to the correct position. is a drawing showing

Referring to FIGS. 5A and 5B , here, the tip position of the inkjet nozzle 300 may be precisely controlled by a motor (stepper motor, linear motor, etc.). The front end of the inkjet nozzle 300 may be adjusted in the X-axis, Y-axis, and Z-axis directions. Meanwhile, the camera unit for photographing the front end of the inkjet nozzle 300 may also be precisely controlled by a separate motor. However, if automatic repair for a certain defect is terminated due to errors in the motor itself, the position of the tip of the inkjet nozzle 300 is changed after repair, etc., the tip of the inkjet nozzle 300 is out of position in the next repair process. A problem arises.

Meanwhile, the learning model 210 for repair may control a motor so that the front end of the inkjet nozzle 300 can be corrected back to the correct position after repair. To this end, the learning model 210 for repair uses the image of the previous inkjet nozzle 300 as a reference image to correct the tip position of the next inkjet nozzle 300 . However, the learning model 210 for repair may use the images of the plurality of previous inkjet nozzles 300 as reference images by applying different weights to previous images used as training data.

Specifically, the image generator may create a nozzle image in which only the inkjet nozzle 300 exists by subtracting an image in which the inkjet nozzle 300 does not exist from an image in which the inkjet nozzle 300 exists.

For example, the repair learning model 210 gives the highest weight to first information extracted from the image of the immediately preceding inkjet nozzle 300, and second information extracted from a plurality of previously set images of the inkjet nozzle 300. An error in the position of the tip of the inkjet nozzle 300 may be corrected by assigning weights according to ranks to information to n-th information.

In the above, preferred embodiments of the present invention have been described by way of example, but the scope of the present invention is not limited only to these specific embodiments, and can be appropriately changed within the scope described in the claims.

s10: first step s20: second step
s30: 3rd step s40: 4th step
s50: 5th step

100: unit area recognition unit
200: defect determination unit 210: learning model for repair
220: reasoning module for repair
300: inkjet nozzle

Claims (9)

A first step of providing a display panel for inspecting the presence or absence of defects;
a second step of photographing the upper surface of the panel and generating an input image;
A third step of constructing a learning model for repair based on artificial intelligence (AI) using the input image;
If there is a defect, a fourth step of determining the defect by analyzing information about the defect;
If it is determined that the defect is repairable, a fifth step of repairing the panel by applying a repair recipe corresponding to the defect; an automatic defect repair method for an inkjet display panel using artificial intelligence (AI). .
According to claim 1,
Between the second step and the third step,
Grouping the first feature group of defects detected in the panel and building a rule-based repair reasoning module for analyzing the first feature group; inkjet using artificial intelligence (AI) further comprising: Automated fault repair method for display panels of the type.
According to claim 2,
In the fourth step,
Automatic defect repair for an inkjet-type display panel using artificial intelligence (AI), characterized in that the final form of defect is determined using the repair learning model and each analysis data obtained from the repair inference module. method.
According to claim 1,
The learning model for repair is artificial, characterized in that the second feature group including 1) the location of the layer where the defect exists, 2) the type of defect, and 3) the region where the defect exists detected from the defect is used as learning data. Automatic defect repair method for inkjet type display panel using intelligence (AI).
According to claim 1,
In the second step,
An automatic defect repair method for an inkjet type display panel using artificial intelligence (AI), characterized in that the input image is divided into at least one unit area image of a preset size.
The method of claim 5, wherein the fourth step
An automatic defect repair method for an inkjet display panel using artificial intelligence (AI), characterized in that by comparing each pattern of the unit area image and the reference image to detect information on the defect.
The method of claim 1, wherein the fourth step,
a 4-1 step of transmitting the defect to a data server unit when the defect is detected;
An automatic defect repair method for an inkjet type display panel using artificial intelligence (AI), including a 4-2 step of determining the defect as one of repairable, non-repairable and unknown.
According to claim 4,
The automatic defect repair method for an inkjet type display panel using artificial intelligence (AI), characterized in that the learning model updates the repair recipe when learning the second feature group for the defect.
The method of claim 1, wherein the fifth step,
An automatic defect repair method for an inkjet-type display panel using artificial intelligence (AI), characterized in that automatically moving the tip of an inkjet nozzle for injecting repair ink in response to the repair recipe.

Images