WO2019037221A1 - 自适应对位方法 - Google Patents

自适应对位方法 Download PDF

Info

Publication number
WO2019037221A1
WO2019037221A1 PCT/CN2017/106931 CN2017106931W WO2019037221A1 WO 2019037221 A1 WO2019037221 A1 WO 2019037221A1 CN 2017106931 W CN2017106931 W CN 2017106931W WO 2019037221 A1 WO2019037221 A1 WO 2019037221A1
Authority
WO
WIPO (PCT)
Prior art keywords
mask
substrate
mark
image
region
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/106931
Other languages
English (en)
French (fr)
Inventor
徐湘伦
邹新
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan China Star Optoelectronics Semiconductor Display Technology Co Ltd
Original Assignee
Wuhan China Star Optoelectronics Semiconductor Display Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Wuhan China Star Optoelectronics Semiconductor Display Technology Co Ltd filed Critical Wuhan China Star Optoelectronics Semiconductor Display Technology Co Ltd
Publication of WO2019037221A1 publication Critical patent/WO2019037221A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10K—ORGANIC ELECTRIC SOLID-STATE DEVICES
    • H10K71/00—Manufacture or treatment specially adapted for the organic devices covered by this subclass
    • H10K71/10—Deposition of organic active material
    • H10K71/16—Deposition of organic active material using physical vapour deposition [PVD], e.g. vacuum deposition or sputtering
    • H10K71/166—Deposition of organic active material using physical vapour deposition [PVD], e.g. vacuum deposition or sputtering using selective deposition, e.g. using a mask
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10K—ORGANIC ELECTRIC SOLID-STATE DEVICES
    • H10K71/00—Manufacture or treatment specially adapted for the organic devices covered by this subclass
    • H10K71/10—Deposition of organic active material
    • H10K71/191—Deposition of organic active material characterised by provisions for the orientation or alignment of the layer to be deposited
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10P—GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
    • H10P72/00—Handling or holding of wafers, substrates or devices during manufacture or treatment thereof
    • H10P72/50—Handling or holding of wafers, substrates or devices during manufacture or treatment thereof for positioning, orientation or alignment
    • H10P72/53—Handling or holding of wafers, substrates or devices during manufacture or treatment thereof for positioning, orientation or alignment using optical controlling means
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10P—GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10
    • H10P72/00—Handling or holding of wafers, substrates or devices during manufacture or treatment thereof
    • H10P72/50—Handling or holding of wafers, substrates or devices during manufacture or treatment thereof for positioning, orientation or alignment
    • H10P72/57—Mask-wafer alignment
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10K—ORGANIC ELECTRIC SOLID-STATE DEVICES
    • H10K71/00—Manufacture or treatment specially adapted for the organic devices covered by this subclass

Definitions

  • the present invention relates to the field of OLED display device process technologies, and in particular, to an adaptive alignment method.
  • OLED Organic Light Emitting Diode
  • OLED Organic Light Emitting Diode
  • the OLED display device generally includes a substrate, an anode disposed on the substrate, a hole injection layer disposed on the anode, a hole transport layer disposed on the hole injection layer, and a light-emitting layer disposed on the hole transport layer.
  • the fabrication process of the OLED display device is vapor deposition and inkjet printing, and the OLED display device with medium and small size and high pixel density usually adopts an evaporation process.
  • the evaporation process requires the use of a precision mask, and the alignment mark on the mask is aligned with the alignment mark on the substrate, so that the substrate and the mask are accurately fitted.
  • the position of the target is generally obtained by a charge-coupled device (CCD), also known as a CCD image sensor.
  • CCD charge-coupled device
  • the invention patent application with the publication number of CN 106054543 A provides a aligning method and a aligning system.
  • the technical solution of the invention claims that the outer shape of the target is the same, the size is the same, the contour lines are the same, and the alignment is performed. The requirements are very high.
  • the substrate has a series of exposure, development, etching and other processes before entering the vapor deposition device, which may cause differences in the color and contour definition of the alignment mark on the substrate, and the alignment mark on the mask is processed. There will also be differences in size, brightness, and darkness.
  • the process differences between the substrate and the mask itself may cause differences in the size, brightness, and contrast of the target image captured by the CCD image sensor ( For example, the contrast between the alignment mark on the flexible substrate and the target on the rigid substrate is different, which affects the alignment accuracy, which may cause a series of problems such as alignment failure, process stop, product scrap, equipment alarm, and capacity drop.
  • the object of the present invention is to provide an adaptive alignment method, which can effectively improve the alignment accuracy, improve the compatibility of the alignment mechanism with the substrate and the mask, reduce the requirements for the alignment, and reduce the alignment fault. Improve equipment utilization rate.
  • the present invention provides an adaptive alignment method comprising the following steps:
  • Step S1 providing a plurality of substrates, and a plurality of mask plates, and using an CCD image sensor to acquire an image of the region where the substrate is located and a region of the region where the mask is located;
  • Step S2 preprocessing the image of the region where the target is located on the substrate obtained by the CCD image sensor, and the image of the region where the mask is located in the mask;
  • Step S3 respectively, the image of the pre-processed substrate to the area where the target is located, and the image of the area where the mask is located in the mask are identified by the position mark;
  • Step S4 performing dimension conversion processing and gray level normalization processing on the identified substrate pair mark and the mask to the position mark respectively;
  • Step S5 selecting a substrate for performing the substrate size conversion processing and the gray normalization processing, and a mask for completing the mask size conversion processing and the gray normalization processing, and calculating the corresponding substrate. Deviation of the center position of the position mark and the mask to the mark;
  • Step S6 The alignment mechanism moves the substrate to the position mark and the mask to the center position deviation of the mask according to the step S5, so that the substrate alignment mark and the mask plate are aligned and overlapped, and the corresponding The position of the substrate and the mask is matched.
  • the step S5 further includes calculating a torsion angle between the substrate pair mark connection line on the selected substrate and the corresponding mask plate pair mark connection line on the mask plate; the step S6 further includes a alignment mechanism Rotating according to the torsion angle calculated in step S5, the substrate pair mark connection line on the substrate and the corresponding mask plate on the mask plate are aligned with the mark line.
  • the step S2 preprocesses the image of the region where the target is located on the substrate obtained by the CCD image sensor and the image of the region where the mask is located, including the illumination compensation process.
  • the step S2 performs preprocessing on the image of the region where the target is located in the substrate obtained by the CCD image sensor and the image of the region where the mask is located, and further includes denoising processing.
  • the denoising process is a median filtering process on the image using a median filter.
  • Step S31 respectively establishing a first shape statistical model of the shape change rule of the target substrate of the reaction substrate, and a first partial gray scale model of the gray scale distribution law of the target substrate; and establishing a shape of the target mask of the reaction mask respectively a second shape statistical model of the variation law, and a second partial gray scale model of the gray scale distribution law of the response mask;
  • Step S32 Searching for a substrate pair mark by using the first partial gray scale model, and using the first shape statistical model to approximate the shape of the searched substrate to the target mark; and using the second partial gray scale
  • the model search mask is paired with the target, and the shape of the searched mask is approximated by the second shape statistical model;
  • Step S33 determining the rationality of the shape of the substrate and the rationality of the shape of the mask, and if the determination is unreasonable, repeating steps S31 and S32; if the determination is reasonable, executing step S4.
  • step S33 the upper limit number of steps S31 and S32 is repeated, and if the upper limit is reached, an alarm is issued.
  • the step S4 takes the median size when performing the size conversion process.
  • the step S4 performs grayscale normalization processing using a histogram equalization method.
  • the invention also provides an adaptive alignment method, comprising the following steps:
  • Step S1 providing a plurality of substrates, and a plurality of mask plates, and using an CCD image sensor to acquire an image of the region where the substrate is located and a region of the region where the mask is located;
  • Step S2 preprocessing the image of the region where the target is located on the substrate obtained by the CCD image sensor, and the image of the region where the mask is located in the mask;
  • Step S3 respectively, the image of the pre-processed substrate to the area where the target is located, and the image of the area where the mask is located in the mask are identified by the position mark;
  • Step S4 performing dimension conversion processing and gray level normalization processing on the identified substrate pair mark and the mask to the position mark respectively;
  • Step S5 selecting a substrate for performing the substrate size conversion processing and the gray normalization processing, and a mask for completing the mask size conversion processing and the gray normalization processing, and calculating the corresponding substrate. Deviation of the center position of the position mark and the mask to the mark;
  • Step S6 The alignment mechanism moves the substrate to the position mark and the mask to the center position deviation of the mask according to the step S5, so that the substrate alignment mark and the mask plate are aligned and overlapped, and the corresponding The position of the substrate and the mask is matched;
  • the step S5 further includes calculating a torsion angle between the substrate pair mark connection line on the selected substrate and the corresponding mask plate pair mark connection line on the mask; the step S6 further includes alignment The mechanism rotates according to the torsion angle calculated in step S5, so that the substrate pair mark connection line on the substrate and the corresponding mask plate on the mask plate are aligned with the mark line;
  • the step S2 performs pre-processing on the image of the region where the target is located on the substrate obtained by the CCD image sensor and the image of the region where the mask is located, including the illumination compensation process;
  • step S2 the image obtained by the CCD image sensor and the image of the region where the target is located, and the image of the region where the mask is located are further processed by the denoising process;
  • the denoising process is a median filtering process on the image using a median filter.
  • An adaptive alignment method provided by the present invention for a CCD image
  • the image obtained by the sensor is used to pre-process the image of the region in which the target is located and the image of the mask in the region where the target is located, and then identify the alignment mark, and then respectively perform the identification of the identified substrate on the target and the mask.
  • Dimensional transformation processing and grayscale normalization processing so that the size of the target is uniform, the center is unchanged, and the edges are clear.
  • the dimensions of the masks are uniform, the center is unchanged, the edges are clear, and then the corresponding substrate pairs are calculated.
  • the position mark and the mask plate are offset from the center position of the target mark, and finally the alignment mechanism moves according to the center position deviation, so that the substrate alignment mark and the mask plate are aligned and overlapped to realize the corresponding substrate and
  • the position matching of the mask can not only improve the alignment accuracy, but also improve the compatibility of the alignment mechanism with the substrate and the mask.
  • the requirements for the dimensional deviation and color deviation of the alignment mark are reduced, and the alignment rate is improved. Improvement, the alignment fault is reduced, and the equipment utilization rate is improved.
  • 1 is a flow chart of an adaptive alignment method of the present invention
  • step S2 is a schematic diagram of step S2 in the adaptive alignment method of the present invention.
  • step S4 in the adaptive alignment method of the present invention
  • step S5 is a schematic diagram of step S5 in the adaptive alignment method of the present invention.
  • FIG. 5 is a schematic diagram of step S6 in the adaptive alignment method of the present invention.
  • the present invention provides an adaptive alignment method applied to an evaporation process of an OLED display device, comprising the following steps:
  • Step S1 providing a plurality of substrates 1 (usually glass substrates) and a plurality of masks 2, and using the CCD image sensor to acquire the substrate pairs during the process in which the substrate 1 and the mask 2 enter the platform of the vapor deposition process chamber The image of the area where the bit 11 is located and the image of the area where the mask 21 is located.
  • Step S2 preprocesses the image of the region where the substrate 11 is located and the image of the region where the mask 21 is located with the mask obtained by the CCD image sensor.
  • the preprocessing includes an illumination compensation process and a denoising process.
  • the CCD image sensor acquires the light and dark images of the target 11 of the substrate. Inconsistent, the grayscale characteristics of the feedback to the computer are different.
  • the illumination compensation process can weaken the instability of the etching process and the influence of the existence of other layers on the subsequent identification of the target. Further, the illumination compensation process can take the existing The SSR (Single Scale Retines) algorithm, the MSR (Multi-Scale Retines) algorithm, or other illumination compensation algorithms are used.
  • the median filter is used to mediate the image to achieve the denoising effect.
  • step S2 illumination compensation processing and the denoising processing the influence of the substrate difference and the noise on the subsequent step identification of the subsequent steps can be substantially eliminated.
  • step S3 the image of the area where the pre-processed substrate is in the position mark 11 and the image of the area where the mask 21 is located are respectively identified by the target mark.
  • step S3 the specific process of identifying the location mark in the step S3 is:
  • Step S31 respectively establishing a first shape statistical model of the shape change rule of the target substrate 11 and a first partial gray scale model of the gray scale distribution rule of the target substrate 11; and establishing a reaction mask alignment a second shape statistical model of the shape change rule of the standard 21; and a second partial gray scale model of the gray scale distribution rule of the target mask 21;
  • the first partial gray scale model and the second partial gray scale model are obtained through training, that is, statistics and analysis are performed on the gray scale data of the position mark 11 and the gray scale data of the mask mark 21 by a large number of acquisition substrates; get.
  • Step S32 Searching the substrate alignment target 11 by using the first partial gray scale model, and using the first shape statistical model to approximate the shape of the searched substrate pair target 11; and searching the mask panel by using the second partial gray scale model.
  • the shape of the searched mask is aligned with the second shape statistical model;
  • Step S33 determining the rationality of the shape of the substrate 11 and the rationality of the shape of the mask 21, that is, adaptively searching for and identifying the substrate alignment mark 11 and different shapes of different shapes and sizes. , the mask of different size and size is opposite to the target 21;
  • step S31 and step S32 are repeated, and the upper limit times of step S31 and step S32 are repeated (if the upper limit is set to 20 times), and once the upper limit is reached, the control system is controlled to be alarmed;
  • step S4 is performed downward.
  • Step S4 as shown in FIG. 3, respectively, the identified substrate pair target 11 and the mask pair target 21 are subjected to size conversion processing and grayscale normalization processing, so that the size of each substrate alignment mark 11 is unified.
  • the center is unchanged and the edges are clear.
  • the size of the mask 21 is uniform, the center is unchanged, and the edges are clear.
  • the median size of each of the substrate pairs 11 and the median size of each of the masks 21 are obtained.
  • the step S4 performs gray level normalization processing using a histogram equalization method, and the calculation formula is as follows:
  • N1 represents the sum of the pixels in the original image
  • r j1 represents a gray level in the original image
  • n1 j1 represents the number of pixels in the original image with a gray level of r j1
  • L is the total number of gray levels present in the original image.
  • P r (r j1 ) represents the ratio of the pixels of the original image with the gray level r j1 to the total pixels
  • S k1 represents the gray level of each pixel mapped to the output image by the gray level r j1 in the original image.
  • the histogram equalization method maps each pixel of the original image with a gray level of r j1 to a corresponding pixel of the output image with a gray level of S k1 according to the transformation formula.
  • Step S5 selecting a substrate 1 for performing size conversion processing and grayscale normalization processing on the substrate alignment mark 11 and a completion mask size alignment processing and grayscale normalization processing
  • the mask 2 calculates the center position deviation ⁇ X and ⁇ Y of the corresponding substrate alignment mark 11 and the mask alignment mark 21, where ⁇ X represents the lateral deviation and ⁇ Y represents the longitudinal deviation.
  • the placement direction of the substrate 1 and the mask 2 is the same, and the center position deviation ⁇ X, ⁇ Y of the corresponding substrate alignment mark 11 and the mask alignment mark 21 can be calculated as the subsequent step S6.
  • the middle alignment mechanism provides a basis for how to move; however, there are also cases where the placement direction of the substrate 1 and the mask 2 are inconsistent, and this step S5 is required to calculate the connection of the substrate alignment mark 11 on the selected substrate 1.
  • Step S6 as shown in FIG. 5, the alignment mechanism moves the substrate position index 11 calculated by the step S5 and the center position deviation ⁇ X, ⁇ Y of the mask 21 with the mask 21 to make the substrate alignment target 11 and The mask is aligned and coincided with the position mark 21. If the torsion angle ⁇ is present, the alignment mechanism is further rotated according to the torsion angle ⁇ calculated in the above step S5, so that the substrate is aligned on the substrate 1. The 11 connection line and the corresponding mask on the mask 2 are aligned with the alignment mark 21.
  • the adaptive alignment method of the present invention it is possible to overcome the alignment between the substrate 1 and the mask 2 Due to the difference between the substrates 1 and the misalignment caused by the manufacturing variations of the mask 2, the flexible substrate and the rigid substrate are compatible, and the requirements for the dimensional deviation and color deviation of the alignment mark are lowered, and the alignment accuracy and the alignment are passed. The rate is improved, the misalignment is reduced, and the equipment utilization rate is improved.
  • the adaptive alignment method of the present invention pre-processes the image of the region where the substrate is located in the CCD image sensor and the image of the region where the mask is located, and then identifies the alignment mark, and then Dimensional transformation processing and grayscale normalization processing are performed on the identified substrate pair mark and mask, respectively, so that the dimensions of the substrate are uniform, the center is unchanged, the edges are clear, and the masks are aligned.
  • the size of the target is uniform, the center is unchanged, and the edge is clear.
  • the center position deviation of the corresponding substrate alignment mark and the mask is calculated, and finally the alignment mechanism moves according to the center position deviation, so that the substrate is aligned.
  • the position and the alignment of the mask with the mask are achieved, and the position matching of the corresponding substrate and the mask is realized, which can effectively improve the alignment accuracy and improve the compatibility of the alignment mechanism with the substrate and the mask.
  • the requirements for dimensional deviation, color deviation, etc. of the position mark are reduced, the alignment pass rate is improved, the alignment failure is reduced, and the equipment utilization rate is improved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Exposure And Positioning Against Photoresist Photosensitive Materials (AREA)
  • Electroluminescent Light Sources (AREA)

Abstract

一种自适应对位方法，对CCD图像传感器获得的基板对位标(11)所在区域的图像、及掩膜板对位标(21)所在区域的图像进行预处理后识别对位标(S3)，再分别对识别到的基板对位标(11)、掩膜板对位标(21)进行尺寸变换处理与灰度归一化处理(S4)，使各基板对位标(11)的尺寸统一、中心不变、边缘清晰，各掩膜板对位标(21)的尺寸统一、中心不变、边缘清晰，然后计算相应的基板对位标(11)与掩膜板对位标(21)的中心位置偏差(S5)，最后对位机构根据所述中心位置偏差进行移动，使所述基板对位标(11)与掩膜板对位标(21)对位、重合，实现相应的基板(1)和掩膜板(2)的位置匹配(S6)，既能够有效提高对位精度，又能够提升对位机构对基板(1)与掩膜板(2)的兼容性，降低对对位标的制作要求，减少对位故障，改善设备稼动率。

Description

自适应对位方法 技术领域
本发明涉及OLED显示器件制程技术领域，尤其涉及一种自适应对位方法。
背景技术
有机发光二极管(Organic Light Emitting Diode，OLED)显示器件具有自发光、驱动电压低、发光效率高、响应时间短、清晰度与对比度高、近180°视角、使用温度范围宽、可实现柔性显示与大面积全色显示等诸多优点，被业界公认为是最有发展潜力的显示装置。
OLED显示器件通常包括：基板、设于基板上的阳极、设于阳极上的空穴注入层、设于空穴注入层上的空穴传输层、设于空穴传输层上的发光层、设于发光层上的电子传输层、设于电子传输层上的电子注入层、及设于电子注入层上的阴极。
OLED显示器件的制作工艺有蒸镀及喷墨打印，中、小尺寸高像素密度的OLED显示器件通常采用蒸镀工艺。蒸镀工艺需要用到精密掩膜板，并将掩膜板上的对位标与基板上的对位标进行对位，使得基板与掩膜板精准套合。为了达到对位精度，目前一般通过电荷耦合元件(Charge-coupled Device，CCD)，又称为CCD图像传感器获得对位标的位置来实现。
申请公布号为CN 106054543 A的发明专利申请提供了一种对位方法及对位系统，此发明申请的技术方案要求对位标的外围形状一样、尺寸大小一样、轮廓线条一样，对对位标的制作要求非常高。
但实际情况是，基板在进入蒸镀设备之前有一系列曝光、显影、蚀刻等工序会造成基板上的对位标的颜色、轮廓清晰度等有差异，掩膜板上的对位标在加工制作时也会出现尺寸大小、轮廓亮暗上的差异，加上基板、掩膜板本身的工艺差别，会导致CCD图像传感器捕获到的对位标图像在大小、亮暗、对比度等各方面存在差异(例如柔性基板上的对位标与刚性基板上对位标的图像对比度便不一样)，影响对位精度，可能造成对位失败、制程中止、产品报废、设备报警、产能下降等一系列问题。
因此，有必要设计一种新的对位方法来解决上述问题。
发明内容
本发明的目的在于提供一种自适应对位方法，既能够有效提高对位精度，又能够提升对位机构对基板与掩膜板的兼容性，降低对对位标的制作要求，减少对位故障，改善设备稼动率。
为实现上述目的，本发明提供一种自适应对位方法，包括如下步骤：
步骤S1、提供多个基板、及多个掩膜板，使用CCD图像传感器获取基板对位标所在区域的图像、及掩膜板对位标所在区域的图像；
步骤S2、对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理；
步骤S3、分别对预处理后的基板对位标所在区域的图像、掩膜板对位标所在区域的图像进行对位标识别；
步骤S4、分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理；
步骤S5、选取一完成基板对位标尺寸变换处理与灰度归一化处理的基板及一完成掩膜板对位标尺寸变换处理与灰度归一化处理的掩膜板，计算相应的基板对位标与掩膜板对位标的中心位置偏差；
步骤S6、对位机构根据步骤S5计算得到的基板对位标与掩膜板对位标的中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配。
所述步骤S5还包括计算被选取的基板上的基板对位标连线与掩膜板上的的相应掩膜板对位标连线之间的扭转角；所述步骤S6还包括对位机构根据步骤S5计算得到的扭转角进行转动，使所述基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线重合。
所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理包括光照补偿处理。
所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理还包括去噪处理。
所述去噪处理为使用中值滤波器对图像进行中值滤波处理。
所述步骤S3进行对位标识别的具体过程为：
步骤S31、分别建立起反应基板对位标形状变化规律的第一形状统计模型、反应基板对位标灰度分布规律的第一局部灰度模型；及分别建立起反应掩膜板对位标形状变化规律的第二形状统计模型、反应掩膜板对位标灰度分布规律的第二局部灰度模型；
步骤S32、利用第一局部灰度模型搜索基板对位标，利用第一形状统计模型对搜索到的基板对位标的形状进行近似表达；同时利用第二局部灰度 模型搜索掩膜板对位标，利用第二形状统计模型对搜索到的掩膜板对位标的形状进行近似表达；
步骤S33、对基板对位标形状的合理性、掩膜板对位标形状的合理性进行判定，若判定不合理则重复步骤S31与步骤S32；若判定合理则执行步骤S4。
所述步骤S33中设定重复步骤S31与步骤S32的上限次数，若达到上限次数则进行报警。
所述步骤S4进行尺寸变换处理时取中值尺寸。
所述步骤S4进行灰度归一化处理采用直方图均衡法。
本发明还提供一种自适应对位方法，包括如下步骤：
步骤S1、提供多个基板、及多个掩膜板，使用CCD图像传感器获取基板对位标所在区域的图像、及掩膜板对位标所在区域的图像；
步骤S2、对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理；
步骤S3、分别对预处理后的基板对位标所在区域的图像、掩膜板对位标所在区域的图像进行对位标识别；
步骤S4、分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理；
步骤S5、选取一完成基板对位标尺寸变换处理与灰度归一化处理的基板及一完成掩膜板对位标尺寸变换处理与灰度归一化处理的掩膜板，计算相应的基板对位标与掩膜板对位标的中心位置偏差；
步骤S6、对位机构根据步骤S5计算得到的基板对位标与掩膜板对位标的中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配；
其中，所述步骤S5还包括计算被选取的基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线之间的扭转角；所述步骤S6还包括对位机构根据步骤S5计算得到的扭转角进行转动，使所述基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线重合；
其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理包括光照补偿处理；
其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理还包括去噪处理；
其中，所述去噪处理为使用中值滤波器对图像进行中值滤波处理。
本发明的有益效果：本发明提供的一种自适应对位方法，对CCD图像 传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理后识别对位标，再分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理，使各基板对位标的尺寸统一、中心不变、边缘清晰，各掩膜板对位标的尺寸统一、中心不变、边缘清晰，然后计算相应的基板对位标与掩膜板对位标的中心位置偏差，最后对位机构根据所述中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配，既能够有效提高对位精度，又能够提升对位机构对基板与掩膜板的兼容性，对对位标的尺寸偏差、颜色偏差等的要求降低，对位通过率得以提高，对位故障得以减少，设备稼动率得以改善。
附图说明
为了能更进一步了解本发明的特征以及技术内容，请参阅以下有关本发明的详细说明与附图，然而附图仅提供参考与说明用，并非用来对本发明加以限制。
附图中，
图1为本发明的自适应对位方法的流程图；
图2为本发明的自适应对位方法中步骤S2的示意图；
图3为本发明的自适应对位方法中步骤S4的示意图；
图4为本发明的自适应对位方法中步骤S5的示意图；
图5为本发明的自适应对位方法中步骤S6的示意图。
具体实施方式
为更进一步阐述本发明所采取的技术手段及其效果，以下结合本发明的优选实施例及其附图进行详细描述。
请参阅图1，本发明提供一种应用于OLED显示器件蒸镀制程的自适应对位方法，包括如下步骤：
步骤S1、提供多个基板1(通常为玻璃基板)、及多个掩膜板2，在所述基板1与掩膜板2进入蒸镀工艺腔室内平台的过程中使用CCD图像传感器获取基板对位标11所在区域的图像、及掩膜板对位标21所在区域的图像。
步骤S2、如图2所示，对CCD图像传感器获得的基板对位标11所在区域的图像、及掩膜板对位标21所在区域的图像进行预处理。
具体地，所述预处理包括光照补偿处理与去噪处理。
由于基板对位标11的图像会受蚀刻工艺不稳定、基板1上有无聚酰亚胺或其它膜层等因素的影响，导致CCD图像传感器获取到基板对位标11的图像的明、暗不一致，反馈给计算机的灰度特征就存在差异，采用光照补偿处理能够弱化蚀刻工艺不稳定、其它膜层的存在对后续对位标识别的影响，进一步地，光照补偿处理可以采取现有的、应用较广泛的SSR(Single Scale Retines)算法、MSR(Multi-Scale Retines)算法、或其它光照补偿算法来进行。
此外，在CCD图像传感器获取图像的实际过程中会携带随机电子噪声、与环境干扰噪声，如果对噪声不处理直接进行对位标识别，会影响位标识别过程，导致最终的对位结果存在较大的误差，因此在进行对位标识别之前应先进行去噪处理，优选地，选用中值滤波器对图像进行中值滤波处理，达到去噪效果。
经该步骤S2光照补偿处理与去噪处理，能够基本消除基板差异及噪声对后续步骤做对位标识别的影响。
步骤S3、分别对预处理后的基板对位标11所在区域的图像、掩膜板对位标21所在区域的图像进行对位标识别。
具体地，该步骤S3进行对位标识别的具体过程为：
步骤S31、分别建立起反应基板对位标11形状变化规律的第一形状统计模型、反应基板对位标11灰度分布规律的第一局部灰度模型；及分别建立起反应掩膜板对位标21形状变化规律的第二形状统计模型、反应掩膜板对位标21灰度分布规律的第二局部灰度模型；
其中，所述第一局部灰度模型与第二局部灰度模型通过训练得到，即通过大量采集基板对位标11的灰度数据与掩膜板对位标21的灰度数据做统计、分析得到。
步骤S32、利用第一局部灰度模型搜索基板对位标11，利用第一形状统计模型对搜索到的基板对位标11的形状进行近似表达；同时利用第二局部灰度模型搜索掩膜板对位标21，利用第二形状统计模型对搜索到的掩膜板对位标21的形状进行近似表达；
步骤S33、对基板对位标11形状的合理性、掩膜板对位标21形状的合理性进行判定，即自适应寻找并识别不同形状、不同尺寸大小的基板对位标11、与不同形状、不同尺寸大小的掩膜板对位标21；
若判定不合理则重复步骤S31与步骤S32，并设定重复步骤S31与步骤S32的上限次数(如将上限次数设定为20次)，一旦达到上限次数则控制对位系统进行报警；
若判定合理则向下执行步骤S4。
步骤S4、如图3所示，分别对识别到的基板对位标11、掩膜板对位标21进行尺寸变换处理与灰度归一化处理，使各基板对位标11的尺寸统一、中心不变、边缘清晰，各掩膜板对位标21的尺寸统一、中心不变、边缘清晰。
具体地，该步骤S4做尺寸变换处理时取的是各基板对位标11的中值尺寸、及各掩膜板对位标21的中值尺寸。
该步骤S4进行灰度归一化处理采用直方图均衡法，其计算公式如下：
Figure PCTCN2017106931-appb-000001
其中：k1＝0，1，2，3，……L-1
n1表示原始图像中像素的总和，rj1表示原始图像中的一灰度级，n1j1表示原始图像中灰度级为rj1的像素个数，L为原始图像中存在的灰度级总数，Pr(rj1)表示原始图像中灰度级为rj1的像素占总的像素的比例，Sk1表示由原始图像中灰度级为rj1的各像素映射到输出图像中的灰度级。
直方图均衡方法就是根据该变换公式将原始图像中灰度级为rj1的各像素映射到输出图像中灰度级为Sk1的对应像素。
步骤S5、如图4所示，选取一完成基板对位标11尺寸变换处理与灰度归一化处理的基板1及一完成掩膜板对位标21尺寸变换处理与灰度归一化处理的掩膜板2，计算相应的基板对位标11与掩膜板对位标21的中心位置偏差ΔX与ΔY，其中ΔX表示横向偏差，ΔY表示纵向偏差。
在大多数情况下，基板1与掩膜板2的放置方向是一致的，计算出相应的基板对位标11与掩膜板对位标21的中心位置偏差ΔX、ΔY即可为后续步骤S6中对位机构如何运动提供依据；但也存在少数基板1与掩膜板2的放置方向不一致的情况，这就需要该步骤S5计算出被选取的基板1上的基板对位标11连线与掩膜板2上的相应掩膜板对位标21连线之间的扭转角θ。
步骤S6、如图5所示，对位机构根据步骤S5计算得到的基板对位标11与掩膜板对位标21的中心位置偏差ΔX、ΔY进行移动，使所述基板对位标11与掩膜板对位标21对位、重合，若存在所述扭转角θ，则对位机构还要根据上述步骤S5计算得到的扭转角θ进行转动，使所述基板1上的基板对位标11连线与掩膜板2上的相应掩膜板对位标21连线重合。
采用本发明的自适应对位方法，能够克服在基板1与掩膜板2对位过 程中由于基板1间的差异、掩膜板2的制作偏差造成的对位困难，兼容柔性基板与刚性基板，对对位标的尺寸偏差、颜色偏差等的要求降低，对位精度与对位通过率得以提高，对位故障得以减少，设备稼动率得以改善。
综上所述，本发明的自适应对位方法，对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理后识别对位标，再分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理，使各基板对位标的尺寸统一、中心不变、边缘清晰，各掩膜板对位标的尺寸统一、中心不变、边缘清晰，然后计算相应的基板对位标与掩膜板对位标的中心位置偏差，最后对位机构根据所述中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配，既能够有效提高对位精度，又能够提升对位机构对基板与掩膜板的兼容性，对对位标的尺寸偏差、颜色偏差等的要求降低，对位通过率得以提高，对位故障得以减少，设备稼动率得以改善。
以上所述，对于本领域的普通技术人员来说，可以根据本发明的技术方案和技术构思作出其他各种相应的改变和变形，而所有这些改变和变形都应属于本发明的权利要求的保护范围。

Claims (14)

  1. 一种自适应对位方法，包括如下步骤：
    步骤S1、提供多个基板、及多个掩膜板，使用CCD图像传感器获取基板对位标所在区域的图像、及掩膜板对位标所在区域的图像；
    步骤S2、对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理；
    步骤S3、分别对预处理后的基板对位标所在区域的图像、掩膜板对位标所在区域的图像进行对位标识别；
    步骤S4、分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理；
    步骤S5、选取一完成基板对位标尺寸变换处理与灰度归一化处理的基板及一完成掩膜板对位标尺寸变换处理与灰度归一化处理的掩膜板，计算相应的基板对位标与掩膜板对位标的中心位置偏差；
    步骤S6、对位机构根据步骤S5计算得到的基板对位标与掩膜板对位标的中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配。
  2. 如权利要求1所述的自适应对位方法，其中，所述步骤S5还包括计算被选取的基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线之间的扭转角；所述步骤S6还包括对位机构根据步骤S5计算得到的扭转角进行转动，使所述基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线重合。
  3. 如权利要求1所述的自适应对位方法，其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理包括光照补偿处理。
  4. 如权利要求3所述的自适应对位方法，其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理还包括去噪处理。
  5. 如权利要求4所述的自适应对位方法，其中，所述去噪处理为使用中值滤波器对图像进行中值滤波处理。
  6. 如权利要求1所述的自适应对位方法，其中，所述步骤S3进行对位标识别的具体过程为：
    步骤S31、分别建立起反应基板对位标形状变化规律的第一形状统计模 型、反应基板对位标灰度分布规律的第一局部灰度模型；及分别建立起反应掩膜板对位标形状变化规律的第二形状统计模型、反应掩膜板对位标灰度分布规律的第二局部灰度模型；
    步骤S32、利用第一局部灰度模型搜索基板对位标，利用第一形状统计模型对搜索到的基板对位标的形状进行近似表达；同时利用第二局部灰度模型搜索掩膜板对位标，利用第二形状统计模型对搜索到的掩膜板对位标的形状进行近似表达；
    步骤S33、对基板对位标形状的合理性、掩膜板对位标形状的合理性进行判定，若判定不合理则重复步骤S31与步骤S32；若判定合理则执行步骤S4。
  7. 如权利要求6所述的自适应对位方法，其中，所述步骤S33中设定重复步骤S31与步骤S32的上限次数，若达到上限次数则进行报警。
  8. 权利要求1所述的自适应对位方法，其中，所述步骤S4进行尺寸变换处理时取中值尺寸。
  9. 权利要求1所述的自适应对位方法，其中，所述步骤S4进行灰度归一化处理采用直方图均衡法。
  10. 一种自适应对位方法，包括如下步骤：
    步骤S1、提供多个基板、及多个掩膜板，使用CCD图像传感器获取基板对位标所在区域的图像、及掩膜板对位标所在区域的图像；
    步骤S2、对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理；
    步骤S3、分别对预处理后的基板对位标所在区域的图像、掩膜板对位标所在区域的图像进行对位标识别；
    步骤S4、分别对识别到的基板对位标、掩膜板对位标进行尺寸变换处理与灰度归一化处理；
    步骤S5、选取一完成基板对位标尺寸变换处理与灰度归一化处理的基板及一完成掩膜板对位标尺寸变换处理与灰度归一化处理的掩膜板，计算相应的基板对位标与掩膜板对位标的中心位置偏差；
    步骤S6、对位机构根据步骤S5计算得到的基板对位标与掩膜板对位标的中心位置偏差进行移动，使所述基板对位标与掩膜板对位标对位、重合，实现相应的基板和掩膜板的位置匹配；
    其中，所述步骤S5还包括计算被选取的基板上的基板对位标连线与掩膜板上的相应掩膜板对位标连线之间的扭转角；所述步骤S6还包括对位机构根据步骤S5计算得到的扭转角进行转动，使所述基板上的基板对位标连 线与掩膜板上的相应掩膜板对位标连线重合；
    其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理包括光照补偿处理；
    其中，所述步骤S2对CCD图像传感器获得的基板对位标所在区域的图像、及掩膜板对位标所在区域的图像进行预处理还包括去噪处理；
    其中，所述去噪处理为使用中值滤波器对图像进行中值滤波处理。
  11. 如权利要求10所述的自适应对位方法，其中，所述步骤S3进行对位标识别的具体过程为：
    步骤S31、分别建立起反应基板对位标形状变化规律的第一形状统计模型、反应基板对位标灰度分布规律的第一局部灰度模型；及分别建立起反应掩膜板对位标形状变化规律的第二形状统计模型、反应掩膜板对位标灰度分布规律的第二局部灰度模型；
    步骤S32、利用第一局部灰度模型搜索基板对位标，利用第一形状统计模型对搜索到的基板对位标的形状进行近似表达；同时利用第二局部灰度模型搜索掩膜板对位标，利用第二形状统计模型对搜索到的掩膜板对位标的形状进行近似表达；
    步骤S33、对基板对位标形状的合理性、掩膜板对位标形状的合理性进行判定，若判定不合理则重复步骤S31与步骤S32；若判定合理则执行步骤S4。
  12. 如权利要求11所述的自适应对位方法，其中，所述步骤S33中设定重复步骤S31与步骤S32的上限次数，若达到上限次数则进行报警。
  13. 权利要求10所述的自适应对位方法，其中，所述步骤S4进行尺寸变换处理时取中值尺寸。
  14. 权利要求10所述的自适应对位方法，其中，所述步骤S4进行灰度归一化处理采用直方图均衡法。
PCT/CN2017/106931 2017-08-21 2017-10-19 自适应对位方法 Ceased WO2019037221A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710720174.9 2017-08-21
CN201710720174.9A CN107437592B (zh) 2017-08-21 2017-08-21 自适应对位方法

Publications (1)

Publication Number Publication Date
WO2019037221A1 true WO2019037221A1 (zh) 2019-02-28

Family

ID=60461055

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/106931 Ceased WO2019037221A1 (zh) 2017-08-21 2017-10-19 自适应对位方法

Country Status (2)

Country Link
CN (1) CN107437592B (zh)
WO (1) WO2019037221A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114250438A (zh) * 2022-01-26 2022-03-29 福建华佳彩有限公司 一种掩膜板及其圆形开孔偏移量测方法
CN116560180A (zh) * 2023-05-29 2023-08-08 江苏影速集成电路装备股份有限公司 一种晶圆对位标记的快速检测系统及方法

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111406310A (zh) * 2017-12-29 2020-07-10 深圳市柔宇科技有限公司 用于基板的对位方法及系统
CN108846863B (zh) * 2018-05-08 2021-12-17 信利（惠州）智能显示有限公司 定位标位置检测方法、装置、计算机和存储介质
CN111106055B (zh) * 2019-12-31 2020-12-04 上海精测半导体技术有限公司 一种晶圆位置确定方法
CN111982075A (zh) * 2020-08-18 2020-11-24 上海振华重工(集团)股份有限公司 一种集卡集装箱对箱自动检测系统及其检测方法
US11393698B2 (en) * 2020-12-18 2022-07-19 STATS ChipPAC Pte. Ltd. Mask design for improved attach position

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101117043A (zh) * 2006-08-04 2008-02-06 株式会社日立工业设备技术 丝网印刷装置及图像识别对位方法
US20080112609A1 (en) * 2006-11-09 2008-05-15 Masao Inoue Position detecting method and device, patterning device, and subject to be detected
CN106054543A (zh) * 2016-08-17 2016-10-26 京东方科技集团股份有限公司 对位方法及对位系统
CN206282833U (zh) * 2016-12-28 2017-06-27 上海天马微电子有限公司 一种掩膜板对位系统

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100523350B1 (ko) * 2003-08-21 2005-10-24 이곤철 백-사이드 마스크 정렬이 가능한 노광장치
KR20050074192A (ko) * 2004-01-13 2005-07-18 삼성에스디아이 주식회사 레이저 패터닝 장비의 정렬 방법 및 그에 적합한 패턴마스크
KR20060131276A (ko) * 2005-06-15 2006-12-20 주성엔지니어링(주) 쉐도우 마스크 및 이를 포함한 박막 증착 장치
CN103268028B (zh) * 2013-05-20 2015-06-17 京东方科技集团股份有限公司 自动调节灰度的检出系及方法
CN105549320A (zh) * 2016-01-05 2016-05-04 京东方科技集团股份有限公司 一种对位标记结构、掩模板、基板以及对位方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101117043A (zh) * 2006-08-04 2008-02-06 株式会社日立工业设备技术 丝网印刷装置及图像识别对位方法
US20080112609A1 (en) * 2006-11-09 2008-05-15 Masao Inoue Position detecting method and device, patterning device, and subject to be detected
CN106054543A (zh) * 2016-08-17 2016-10-26 京东方科技集团股份有限公司 对位方法及对位系统
CN206282833U (zh) * 2016-12-28 2017-06-27 上海天马微电子有限公司 一种掩膜板对位系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WANG, XIANMEI ET AL.: "Pupil Localization for Multi-view Eyeballs", JOURNAL OF COMPUTER-AIDED DESIGN & COMPUTER GRAPHICS, vol. 23, no. 8, 31 August 2011 (2011-08-31), China, pages 1427 - 1432, XP055578722 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114250438A (zh) * 2022-01-26 2022-03-29 福建华佳彩有限公司 一种掩膜板及其圆形开孔偏移量测方法
CN114250438B (zh) * 2022-01-26 2023-07-14 福建华佳彩有限公司 一种掩膜板及其圆形开孔偏移量测方法
CN116560180A (zh) * 2023-05-29 2023-08-08 江苏影速集成电路装备股份有限公司 一种晶圆对位标记的快速检测系统及方法
CN116560180B (zh) * 2023-05-29 2024-04-02 江苏影速集成电路装备股份有限公司 一种晶圆对位标记的快速检测系统及方法

Also Published As

Publication number Publication date
CN107437592B (zh) 2019-04-30
CN107437592A (zh) 2017-12-05

Similar Documents

Publication Publication Date Title
WO2019037221A1 (zh) 自适应对位方法
CN115791822B (zh) 晶圆表面缺陷的视觉检测算法及检测系统
CN113298810B (zh) 结合图像增强和深度卷积神经网络的道线检测方法
CN105046271B (zh) 基于模板匹配的melf元件定位与检测方法
US9842890B2 (en) Method of manufacturing display device
CN114494045A (zh) 一种基于机器视觉的大型直齿轮几何参数测量系统及方法
CN107808366B (zh) 一种基于块匹配的自适应光转移单幅阴影去除方法
CN111932504A (zh) 一种基于边缘轮廓信息的亚像素定位方法及装置
CN111444948A (zh) 一种图像特征提取与匹配方法
Liu et al. A novel subpixel industrial chip detection method based on the dual-edge model for surface mount equipment
CN112115948A (zh) 一种基于深度学习的芯片表面字符识别方法
CN114926671A (zh) 基于模板匹配的nled/qled/oled像素缺陷定位方法及系统
CN109697957B (zh) 图像像素校正方法及系统
CN118644479B (zh) 一种纳米压印过程的刻蚀缺陷定位方法
CN109191429B (zh) 一种基于机器视觉的3d打印喷嘴检测方法
US9620567B2 (en) OLED backboard, method of manufacturing the same, alignment system and alignment method thereof
CN117095055B (zh) 一种基于混合标签的鲁棒性标志块及其定位方法
CN112347950A (zh) 基于深度学习的pcb板镭射标靶识别方法及系统
CN115205356A (zh) 一种基于双目立体视觉的实训平台快速调试方法
CN115439327A (zh) 一种改进surf无人机影像快速拼接方法
CN115601297A (zh) 一种柔印标签的在线缺陷检测方法
CN114298977A (zh) 一种基于频域模板匹配的复杂背景图像缺陷检测方法
CN120259629A (zh) 图像处理方法、装置、设备、可读存储介质和程序产品
CN118229795A (zh) 一种长串绝缘子紫外放电图像优化方法、设备及存储介质
CN115410013B (zh) 快速模板匹配方法、设备及计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17922146

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17922146

Country of ref document: EP

Kind code of ref document: A1