WO2015158024A1 - 图像处理方法、装置和自动光学检测机 - Google Patents
图像处理方法、装置和自动光学检测机 Download PDFInfo
- Publication number
- WO2015158024A1 WO2015158024A1 PCT/CN2014/077579 CN2014077579W WO2015158024A1 WO 2015158024 A1 WO2015158024 A1 WO 2015158024A1 CN 2014077579 W CN2014077579 W CN 2014077579W WO 2015158024 A1 WO2015158024 A1 WO 2015158024A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- product
- product image
- value
- detection area
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
Definitions
- the present invention relates to the field of automatic optical measurement technology, and in particular to an image processing method, apparatus and automatic optical inspection machine.
- TFT-LCD Thin Film Transistor Liquid Crystal Display
- AOI Automatic Optical Inspection
- vibrations usually occur during the operation of the AOI device, resulting in a decrease in the sharpness of the image captured by the camera, which affects the detection result.
- FIG. 1 is a schematic diagram of RGB sub-pixel distribution of a color filter in the prior art.
- the black matrix BM (Black Matrix) 12 is arranged on the substrate 11, and the red sub-pixel 13a, the green sub-pixel 13b, and the blue sub-pixel 13c are spaced apart between the black matrixes 12.
- the width D of the black matrix is taken as the line width CD (Critical Demention).
- CD Crohn's the RGB sub-pixels overlap with the black matrix 12 to form an overlap 14 .
- AOI equipment is used to measure the line width CD (Critical Demention) and overlap width of TFT-LCD products to detect product quality.
- FIG. 2 is a schematic diagram showing the blurring of overlapping areas in the scanned image caused by vibration.
- the measured overlap region width L2 is smaller than the normal overlap region width L1.
- the method for reducing the vibration of the AOI equipment is to perform anti-shock treatment on the floor below the machine table, or to transform the rolling device of the machine base substrate carrier into a platform, and further, an air floating function can be added to the platform base to reduce environmental vibration. Caused by substrate vibration.
- the present invention is directed to the above problems in the prior art, and provides an image processing method for use in an automatic optical detection process, including:
- a comparing step of comparing a product image sharpness with the image sharpness threshold when the product image sharpness is greater than the image sharpness threshold, performing a product image selection step; when the product image sharpness is less than the image sharpness threshold Performing the product image sharpness determining step;
- the product image selection step selects the current product image as the image to be detected.
- the resolution threshold determining step comprises:
- a sample scanning step of performing multiple scans on the same detection sample to obtain a plurality of sample images matching the standard slice; a threshold calculation step of respectively acquiring gray values of each pixel point of the plurality of sample images in the detection area, and calculating Image sharpness threshold.
- the product image definition determining step comprises:
- the image sharpness calculating step acquires a gray value of each pixel in the detection area in the product image, and calculates a product image sharpness.
- the threshold calculation step comprises:
- the minimum value of the resolution of all samples is selected as the image sharpness threshold.
- the image sharpness calculation step comprises: Obtaining a gray value of a pixel of the product image in the detection area; respectively calculating an absolute value of a difference value of a gray value of each adjacent pixel in the detection area in the product image, and selecting a maximum value of the absolute value as a product The gray level difference of the image line; the maximum value of the gray level difference of the product image line is selected as the product image sharpness.
- an image processing apparatus for use in an automatic optical inspection machine, comprising: a detection area determining unit for determining a rectangular detection area in a detection image; and a sharpness threshold determining unit for a grayscale value of each pixel point of the sample image in the detection area, calculating an image sharpness threshold; a product image sharpness determining unit for graying each pixel point in the detection area according to the product image a value, calculating a product image sharpness; a comparison unit for comparing the product image sharpness with the image sharpness threshold; when the product image sharpness is greater than the image sharpness threshold, calling the product image selection unit; When the image sharpness is less than the image sharpness threshold, calling the product image sharpness determining unit;
- the product image selection unit is configured to select the current product image as the image to be detected.
- the definition threshold determining unit further includes:
- a sample scanning subunit for performing multiple scans on the same detection sample to obtain a plurality of sample images matching the standard slice
- a threshold calculation subunit configured to acquire a gray value of a pixel in the detection area in each sample image; and respectively calculate a difference value of a gray value of each adjacent pixel in the detection area in each sample image Absolute value, the maximum value of the absolute value is selected as the gray level difference of the sample image line; the maximum value of the gray level difference of the sample image line is selected as the sample sharpness of the current sample image;
- the minimum value of the resolution of all samples is selected as the image sharpness threshold.
- the product image definition determining unit further includes: a product scanning subunit, configured to scan a product to obtain a product image;
- An image sharpness calculation subunit configured to obtain a gray value of a pixel of the product image in the detection area; and respectively calculate an absolute value of a difference value of a gray value of each adjacent pixel in the detection area in the product image , select the maximum value of the absolute value as the gray line difference of the product image line; select the maximum value of the gray line difference of the product image line for the product image clear Clearness.
- an automatic optical detecting machine comprising the above image processing apparatus, further comprising detecting means for performing optical detection based on the image to be detected.
- the definition threshold determination unit can calculate the definition threshold corresponding to the product by multiple scans, thereby providing accurate AOI inspection for each type of product.
- the calculation amount is small, the calculation method is simple and convenient, and the detection efficiency can be greatly improved.
- the invention increases the definition of image sharpness in the image processing process, sets the sharpness threshold, and provides the same product detection standard in the presence of the influence of the machine vibration, thereby improving the detection quality of the AOI device.
- FIG. 1 is a schematic diagram of RGB sub-pixel distribution of a color filter in the prior art
- FIG. 2 is a schematic diagram of the blur caused by the overlap in the scanned image caused by the vibration
- FIG. 3 is a schematic diagram of functional blocks of an image processing apparatus in an embodiment of the present invention.
- FIG. 4 is a schematic diagram of a rectangular detection area determined in an embodiment of the present invention.
- Figure 5 is a schematic diagram of functional modules of an automatic optical inspection machine in an embodiment of the present invention.
- FIG. 6 is a flow chart of an image processing method in an embodiment of the present invention.
- Figure 7b is a sample image with greater definition in the embodiment of the present invention.
- FIG. 8 is a schematic diagram showing a distribution of pixel gradations in a rectangular detection area of a sample image according to an embodiment of the present invention.
- FIG. 9 is a schematic diagram showing the distribution of pixel gradation in a rectangular area of product image in the embodiment of the present invention.
- FIG. 3 is a schematic diagram of functional blocks of an image processing apparatus according to an embodiment of the present invention.
- the image processing apparatus 300 includes:
- the detection area determining unit 301 is configured to determine a rectangular detection area in the detection image; preferably, but not limited to, one of the areas in the rectangular frames 401, 402, and 403 shown in FIG. 4 is selected as the rectangular detection area;
- a sharpness threshold determining unit 302 configured to calculate an image sharpness threshold according to a gray value of each pixel in the detection area of the plurality of sample images
- the sharpness threshold determining unit 302 further includes a sample scanning subunit 3021, configured to perform multiple scans on the same detected sample to obtain a plurality of sample images that match the standard slice; and a threshold calculating subunit 3022, configured to acquire multiple The gray level value of each pixel in the detection area in the sample image is used to calculate an image sharpness threshold.
- the threshold calculation sub-unit 3022 is configured to acquire the gray level of the pixel in the detection area in each sample image.
- the absolute value of the difference between the gray values of adjacent pixels in each detection area in each sample image is calculated, and the maximum value of the absolute value is selected as the gray level difference of the sample image line;
- the maximum value of the difference value is the sample resolution of the current sample image; the minimum value of the resolution of all samples is selected as the image sharpness threshold;
- a product image sharpness determining unit 303 configured to calculate a product image sharpness according to a gray value of each pixel point of the product image in the detection area
- the product image definition determining unit 303 further includes a product scanning sub-unit 3031 for scanning the product to obtain a product image; and an image sharpness calculating sub-unit 3032 for acquiring each pixel in the detection area in the product image. Gray value, calculate the image clarity of the product;
- the image sharpness calculation sub-unit 3032 is configured to obtain a gray value of a pixel of the product image in the detection area, and calculate a difference between the gray values of each adjacent pixel in the detection area in the product image.
- the absolute value of the absolute value is selected as the gray value difference of the product image line; the maximum value of the gray line difference of the product image line is selected as the image clarity of the product;
- the comparing unit 304 is configured to compare the product image sharpness with the image sharpness threshold; when the product image sharpness is greater than the image sharpness threshold, calling the product image selecting unit 305; when the product image sharpness is smaller than the image At the sharpness threshold, the product image sharpness determining unit 303 is called.
- the product image selecting unit 305 is configured to select the current product image as the image to be detected when the product image sharpness is greater than the image sharpness threshold.
- an embodiment of the present invention further provides an automatic optical detector 500, including an image processing apparatus 300, and a detecting apparatus 501 for performing optical detection based on the image to be detected.
- the image processing method of this embodiment is used in the automatic optical detection process, and the image processing method in the embodiment of the present invention will be described below with reference to FIG.
- step S601 the detection area determining unit 301 determines a rectangular detection area in the detected image.
- the detection area near the measurement item is selected, for example, one of the areas within the rectangular frames 401, 402, 403 shown in FIG. 4 or a combination thereof is set as the detection area;
- the detection area in the rectangular frames 401, 402 can be used to measure the overlap width, and the detection area in the rectangular frame 403 can be used to measure the line width CD.
- the area inside the rectangular frame 401 is taken as a rectangular detection area.
- Step S602 the sharpness threshold determining unit 302 performs multiple scans on the same detected sample to obtain a plurality of sample images matched with the standard slice, and respectively acquires gray values of each pixel in the detection region in the plurality of sample images. Calculate the image sharpness threshold.
- the sample scanning subunit 3021 performs multiple scans on the same detection sample to obtain a plurality of sample images matching the standard piece;
- Figures 7a and 7b show two samples of the same test sample that are scanned twice and matched with the standard slice.
- the sample image in Figure 7a is less sharp and the image is blurred, indicating that the AOI is in the process of scanning. Machine vibration Larger;
- Figure 7b shows a sharper image and clearer image, indicating that the AOI machine shakes less during the scanning process;
- the sample images shown in Figures 7a and 7b match the standard, indicating the vibration caused by the AOI device. Measurement errors are tolerable.
- the threshold calculation sub-unit 3022 acquires the gradation value of each pixel in the detection area among the plurality of sample images, respectively, and calculates an image sharpness threshold.
- FIG. 8 is a schematic diagram showing the distribution of pixel gradation in the detection area in the rectangular frame 401 in the sample image 7a.
- the detection area includes 4 rows and 6 columns of pixels, and each row is named A, B, C, and D rows.
- 10,
- 50,
- I A4-A5 I 20, I A5-A6
- 60, the maximum value of the absolute value in the A line is 60, and the gamma difference of the A line in Fig. 7a is 60.
- the same method is used to calculate the line gray difference values in lines B, C, and D, which are 59, 64, and 66, respectively.
- Selecting the sample image Figure 7a The maximum value of the gray level difference of the line 66 is the sample resolution of the sample image in Figure 7a.
- the minimum value of the resolution of all samples is selected as the image sharpness threshold.
- the sample resolution is 70 in Fig. 7b
- the image sharpness threshold is 66.
- the same product detection standard is provided in the presence of the influence of the machine vibration, thereby eliminating the influence of the machine vibration and improving the detection quality of the AOI device.
- Step S603 the product image sharpness determining unit 303 scans the product to obtain the product image, obtains the gray value of each pixel in the detection area in the product image, and calculates the product image sharpness.
- the product scanning subunit 3031 scans the product to obtain a product image; then, the image sharpness calculating subunit 3032 acquires the gray value of each pixel in the detection area in the product image, and calculates the product image sharpness.
- 9,
- 50,
- I A4-A5 I 15, I A5-A6
- 56, the maximum value of the absolute value in the A line is 56, the line of the product image The gamma difference is 56.
- Step S604 the comparing unit 304 determines whether the product image sharpness is greater than the image sharpness threshold; when the product image sharpness is greater than the image sharpness threshold, performing a product image selecting step S605; when the product image sharpness is less than the At the image sharpness threshold, the product image sharpness determining step S603 is executed, and the product image sharpness determining unit 303 is called to scan the product again.
- Step S605 the product image selecting unit 305 selects the current product image as the image to be detected when the product image sharpness is greater than the image sharpness threshold.
- the product image sharpness 67 is greater than the image sharpness threshold 66, and the current product image is selected as the image to be detected for subsequent optical detection.
- the sharpness threshold determination unit 302 in step S602 can calculate the sharpness threshold corresponding to the product by multiple scans, thereby providing accurate AOI inspection for each type of product.
- the embodiment of the present invention since the embodiment of the present invention only calculates the pixel gradation difference in the detection area near the measurement item, the calculation amount is small, and the calculation method is simple and convenient, so that the detection efficiency can be greatly improved.
- the embodiment of the invention increases the definition of image sharpness in the image processing process, sets the definition threshold, and provides the same product detection standard in the presence of the influence of the machine vibration, thereby improving the detection of the AOI device. quality.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
一种图像处理方法、装置和自动光学检测机。所述方法包括检测区域确定步骤,确定检测图像中矩形检测区域;清晰度阈值确定步骤,根据多个样本图像在所述检测区域内的每个像素点的灰度值,计算图像清晰度阈值;产品图像清晰度确定步骤,根据产品图像在所述检测区域内的每个像素点的灰度值,计算产品图像清晰度;比较步骤,比较产品图像清晰度与所述图像清晰度阈值;产品图像选定步骤,选定当前产品图像为待检测图像。
Description
图像处理方法、 装置和自动光学检测机
技术领域
本发明涉及自动光学测量技术领域, 具体地说, 涉及一种图像处理方法、装置和自动 光学检测机。
背景技术
薄膜晶体管液晶显示屏 TFT-LCD (Thin Film Transistor Liquid Crystal Display)正朝着 大尺寸、高分辨率的方向发展, 并且消费者对显示屏的视觉感受的要求也越来越高。通常 采用自动光学检测 AOI (Automated Optical Inspection) 设备监控 TFT-LCD产品品质。 但 是, 在 AOI设备运行过程中通常会出现震动, 导致相机拍摄图像的清晰度降低, 以至于 影响检测结果。
图 1 为现有技术中彩色滤光片的 RGB 子像素分布示意图。 黑色矩阵 BM (Black Matrix) 12排列设置在基板 11 上, 红色子像素 13a, 绿色子像素 13b, 蓝色子像素 13c 间隔设置在黑色矩阵 12之间。通常黑色矩阵的宽度 D作为线宽 CD (Critical Demention)。 在 TFT-LCD制程中, RGB子像素与黑色矩阵 12产生交叠, 形成重叠区域(overlay) 14。 在生产过程中采用 AOI设备测量 TFT-LCD产品的线宽 CD (Critical Demention) 和 重叠区域 (overlay) 宽度来检测产品质量。 通常 AOI设备在运行中会出现震动, 引起扫 描图像清晰度降低, 例如线宽和重叠区域模糊, 边缘不清晰, 导致测量结果不准确甚至误 判。 图 2所示为震动造成扫描图像中重叠区域模糊的示意图。如图 2所示, 测量所得重叠 区域宽度 L2比正常的重叠区域宽度 L1小。 目前, 降低 AOI设备震动的方法是在机台下方的地板做防震处理, 或者将机台基板 承载体的滚压装置改造为平台, 进一步, 还可在平台底座加入气浮功能, 从而减少环境震 动引起的基板震动。
基于上述情况, 亟需一种改进的 AOI图像处理方法解决设备震动引起的扫描图像清 晰度降低的技术问题。
发明内容
本发明针对现有技术中存在的上述问题,提供了一种图像处理方法,用于自动光学检 测过程中, 包括:
检测区域确定步骤, 确定检测图像中的矩形检测区域;
清晰度阈值确定步骤, 根据多个样本图像在所述检测区域内的每个像素点的灰度值, 计算图像清晰度阈值;
产品图像清晰度确定步骤, 根据产品图像在所述检测区域内的每个像素点的灰度值, 计算产品图像清晰度;
比较步骤, 比较产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度大于所述 图像清晰度阈值时,执行产品图像选定步骤; 当产品图像清晰度小于所述图像清晰度阈值 时, 执行所述产品图像清晰度确定步骤;
产品图像选定步骤, 选定当前产品图像为待检测图像。
根据本发明的一个实施例, 所述清晰度阈值确定步骤包括:
样本扫描步骤, 对同一检测样本进行多次扫描获得与标准片匹配的多个样本图像; 阈值计算步骤,分别获取多个样本图像在所述检测区域内的每个像素点的灰度值,计 算图像清晰度阈值。
根据本发明的一个实施例, 所述产品图像清晰度确定步骤包括:
产品扫描步骤, 扫描产品获得产品图像;
图像清晰度计算步骤,获取产品图像中所述检测区域内的每个像素点的灰度值,计算 产品图像清晰度。
根据本发明的一个实施例, 所述阈值计算步骤包括:
获取每一样本图像在所述检测区域内像素点的灰度值;
分别计算每一样本图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
根据本发明的一个实施例, 所述图像清晰度计算步骤包括:
获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。 根据本发明的另一个方面, 提供一种图像处理装置, 用于自动光学检测机中, 包括: 检测区域确定单元, 用于确定检测图像中矩形检测区域; 清晰度阈值确定单元,用于根据多个样本图像在所述检测区域内的每个像素点的灰度 值, 计算图像清晰度阈值; 产品图像清晰度确定单元,用于根据产品图像在所述检测区域内的每个像素点的灰度 值, 计算产品图像清晰度; 比较单元,用于比较产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度大于 所述图像清晰度阈值时,调用产品图像选定单元; 当产品图像清晰度小于所述图像清晰度 阈值时, 调用产品图像清晰度确定单元;
产品图像选定单元, 用于选定当前产品图像为待检测图像。 根据本发明的一个实施例, 所述清晰度阈值确定单元还包括:
样本扫描子单元,用于对同一检测样本进行多次扫描获得与标准片匹配的多个样本图 像;
阈值计算子单元, 用于获取每一样本图像中所述检测区域内像素点的灰度值; 分别计算每一样本图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
根据本发明的一个实施例, 所述产品图像清晰度确定单元还包括: 产品扫描子单元, 用于扫描产品获得产品图像;
图像清晰度计算子单元, 用于获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清
晰度。
根据本发明的另一方面, 提供一种自动光学检测机, 包括上述图像处理装置, 还包括 检测装置, 用于根据所述待检测图像进行光学检测。
本发明带来了以下有益效果:对于不同类型的产品,清晰度阈值确定单元可以通过多 次扫描计算得到相应于该产品的清晰度阈值, 从而针对每种类型的产品提供精准的 AOI 检验。在图像处理过程中, 由于仅仅计算测量项目附近的检测区域内的像素灰度差, 计算 量小, 计算方式简单便捷, 可大幅度提高检测效率。本发明在图像处理过程中增加了图像 清晰度的计算, 设定了清晰度阈值, 在存在机台震动影响的情况下, 提供了针对同一的产 品检测标准, 提高了 AOI设备的检测质量。
本发明的其它特征和优点将在随后的说明书中阐述,并且,部分地从说明书中变得显 而易见, 或者通过实施本发明而了解。本发明的目的和其他优点可通过在说明书、权利要 求书以及附图中所特别指出的结构来实现和获得。 附图说明
下面结合附图对本发明的具体实施方式作进一步详细的说明。
图 1是现有技术中彩色滤光片的 RGB子像素分布示意图;
图 2是震动造成扫描图像中重叠区域模糊的示意图;
图 3是本发明的实施例中图像处理装置的功能模块示意图;
图 4是本发明的实施例中确定的矩形检测区域示意图;
图 5是本发明的实施例中自动光学检测机的功能模块示意图;
图 6是本发明的实施例中图像处理方法的流程图;
图 7a是本发明实施例中清晰度较小的样本图像;
图 7b是本发明实施例中清晰度较大的样本图像;
图 8是本发明实施例中样本图像矩形检测区域内像素灰度的分布示意图;
图 9是本发明实施例中产品图像矩形检测区域内像素灰度的分布示意图。 具体实施方式
以下将结合附图及实施例来详细说明本发明的实施方式,借此对本发明如何应用技术 手段来解决技术问题,并达成技术效果的实现过程能充分理解并据以实施。需要说明的是, 只要不构成冲突,本发明中的各个实施例以及各实施例中的各个特征可以相互结合,所形 成的技术方案均在本发明的保护范围之内。
实施例一
图 3为本发明一个实施例的图像处理装置的功能模块示意图。 图像处理装置 300包 括:
检测区域确定单元 301, 用于确定检测图像中矩形检测区域; 优选的, 但不限于, 选 定图 4所示的矩形框 401、 402、 403内的区域其中之一作为矩形检测区域;
清晰度阈值确定单元 302,用于根据多个样本图像在所述检测区域内的每个像素点的 灰度值, 计算图像清晰度阈值;
优选的, 清晰度阈值确定单元 302进一步包括样本扫描子单元 3021, 用于对同一检 测样本进行多次扫描获得与标准片匹配的多个样本图像; 阈值计算子单元 3022, 用于分 别获取多个样本图像中所述检测区域内的每个像素点的灰度值, 计算图像清晰度阈值; 具体的, 阈值计算子单元 3022用于获取每一样本图像中所述检测区域内像素点的灰 度值; 分别计算每一样本图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对 值,选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当 前样本图像的样本清晰度; 选取全部样本清晰度的最小值为所述图像清晰度阈值;
产品图像清晰度确定单元 303,用于根据产品图像在所述检测区域内的每个像素点的 灰度值, 计算产品图像清晰度;
优选的, 产品图像清晰度确定单元 303进一步包括产品扫描子单元 3031, 用于扫描 产品获得产品图像; 图像清晰度计算子单元 3032, 用于获取产品图像中所述检测区域内 的每个像素点的灰度值, 计算产品图像清晰度;
具体的, 图像清晰度计算子单元 3032用于获取产品图像在所述检测区域内像素点的 灰度值, 分别计算产品图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图 像清晰度;
比较单元 304, 用于比较产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度 大于所述图像清晰度阈值时, 调用产品图像选定单元 305; 当产品图像清晰度小于所述图 像清晰度阈值时, 调用产品图像清晰度确定单元 303。
产品图像选定单元 305, 用于当产品图像清晰度大于所述图像清晰度阈值时, 选定当 前产品图像为待检测图像。
实施例二
如图 5 所示, 本发明的实施例还提供一种自动光学检测机 500, 包括图像处理装置 300, 还包括检测装置 501, 用于根据所述待检测图像进行光学检测。
其中图像处理装置 300的功能参见图 3的描述, 在此不再赘述。
实施例三
本实施例的图像处理方法用于自动光学检测过程中,以下结合图 6对本发明实施例中 的图像处理方法进行说明。
步骤 S601, 检测区域确定单元 301确定检测图像中的矩形检测区域。
优选的, 在 AOI软件 Recipe参数设置的步骤中, 选择测量项目附近的检测区域, 例 如, 设置图 4所示的矩形框 401、 402、 403内的区域其中之一或者其组合为检测区域; 其 中, 矩形框 401、 402内的检测区域可用于测量重叠区域 (overlay) 宽度, 矩形框 403内 的检测区域可用于测量线宽 CD。
在本实施例的计算中, 以矩形框 401内的区域做为矩形检测区域。
步骤 S602, 清晰度阈值确定单元 302对同一检测样本进行多次扫描获得与标准片匹 配的多个样本图像,分别获取多个样本图像中所述检测区域内的每个像素点的灰度值,计 算图像清晰度阈值。
具体的, 首先, 样本扫描子单元 3021对同一检测样本进行多次扫描获得与标准片匹 配的多个样本图像;
举例而言, 图 7a和图 7b为同一检测样本进行两次扫描的,并且与标准片匹配的两个 样本图像, 图 7a中的样本图像清晰度较小, 图像模糊, 说明在扫描过程中 AOI机台震动
较大; 图 7b中样本图像清晰度较大, 图像清晰, 说明在扫描过程中 AOI机台震动较小; 图 7a和图 7b所示的样本图像与标准片相匹配, 说明 AOI设备震动引起的测量误差是可 容忍的。
然后, 阈值计算子单元 3022分别获取多个样本图像中所述检测区域内的每个像素点 的灰度值, 计算图像清晰度阈值。
图 8所示为样本图像 7a中矩形框 401内检测区域内像素灰度的分布示意图, 检测区 域包括有 4行 6列像素, 每行命名为 A、 B、 C、 D行。 以图 8所示的像素灰度分布为例 说明图像清晰度阈值计算子单元 3022的计算过程: 获取每行像素点的灰度值;分别计算每一行相邻像素点灰度值的差值的绝对值,例如 A行中, 计算得到 I A1-A2 I =14, I A2-A3 | =10, | A3-A4 | =50,
I A4-A5 I =20, I A5-A6 | =60, 得到 A行中绝对值的最大值为 60, 样本图像图 7a 的 A行灰度差值为 60。
用同样的方法计算 B、 C、 D行中的行灰度差值, 分别为 59、 64、 66。 选择样本图像图 7a行灰度差值的最大值 66为样本图像图 7a的样本清晰度。 选取全部样本清晰度的最小值为所述图像清晰度阈值, 例如, 图 7b中样本清晰度为 70, 则图像清晰度阈值为 66。 本步骤中,通过样本图像清晰度阈值的计算,在存在机台震动影响的情况下提供针对 同一的产品检测标准, 从而消除机台震动的影响, 提高 AOI设备的检测质量。
步骤 S603, 产品图像清晰度确定单元 303扫描产品获得产品图像, 获取产品图像中 所述检测区域内的每个像素点的灰度值, 计算产品图像清晰度。
首先, 产品扫描子单元 3031扫描产品获得产品图像; 然后, 图像清晰度计算子单元 3032获取产品图像中所述检测区域内的每个像素点的灰度值, 计算产品图像清晰度。
图 9所示为产品图像中矩形框 401内检测区域内像素灰度的分布示意图,以图 9所示 的像素灰度分布为例说明产品图像清晰度计算子单元 3032的计算过程: 获取每行像素点的灰度值;分别计算每一行相邻像素点灰度值的差值的绝对值,例如 A行中, 计算得到 I A1-A2 I =21, I A2-A3 | =9, | A3-A4 | =50,
I A4-A5 I =15, I A5-A6 | =56, 得到 A行中绝对值的最大值为 56, 产品图像的行
灰度差值为 56。
用同样的方法计算 B、 C、 D行中的行灰度差值, 分别为 60、 64、 67。 选择产品图像行灰度差值的最大值 67为产品图像清晰度。
步骤 S604, 比较单元 304判断产品图像清晰度是否大于所述图像清晰度阈值; 当产 品图像清晰度大于所述图像清晰度阈值时, 执行产品图像选定步骤 S605; 当产品图像清 晰度小于所述图像清晰度阈值时, 执行产品图像清晰度确定步骤 S603, 调用所述产品图 像清晰度确定单元 303以再次扫描产品。
步骤 S605, 产品图像选定单元 305, 当产品图像清晰度大于所述图像清晰度阈值时, 选定当前产品图像为待检测图像。
本实施例中产品图像清晰度 67大于图像清晰度阈值 66,选定当前产品图像为待检测 图像, 进行后续的光学检测。
对于不同类型的产品, 步骤 S602中清晰度阈值确定单元 302可以通过多次扫描计算 得到相应于该产品的清晰度阈值, 从而针对每种类型的产品提供精准的 AOI检验。
在图像处理过程中,由于本发明的实施例仅仅计算测量项目附近的检测区域内的像素 灰度差, 计算量小, 计算方式简单便捷, 因此可大幅度提高检测效率。
本发明的实施例在图像处理过程中增加了图像清晰度的计算,设定了清晰度阈值,在 存在机台震动影响的情况下, 提供了针对同一的产品检测标准, 提高了 AOI设备的检测 质量。
虽然本发明所公开的实施方式如上,但所述的内容只是为了便于理解本发明而采用的 实施方式, 并非用以限定本发明。任何本发明所属技术领域内的技术人员, 在不脱离本发 明所揭露的精神和范围的前提下,可以在实施的形式上及细节上作任何的修改与变化,但 本发明的专利保护范围, 仍须以所附的权利要求书所界定的范围为准。
Claims
1、 一种图像处理方法, 用于自动光学检测过程中, 包括:
检测区域确定步骤, 确定检测图像中的矩形检测区域;
清晰度阈值确定步骤, 根据多个样本图像在所述检测区域内的每个像素点的灰度值, 计算图像清晰度阈值;
产品图像清晰度确定步骤, 根据产品图像在所述检测区域内的每个像素点的灰度值, 计算产品图像清晰度;
比较步骤, 比较所述产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度大于 所述图像清晰度阈值时,执行产品图像选定步骤; 当产品图像清晰度小于所述图像清晰度 阈值时, 执行产品图像清晰度确定步骤;
产品图像选定步骤, 选定当前产品图像为待检测图像。
2、 根据权利要求 1所述的图像处理方法, 其中, 所述清晰度阈值确定步骤包括: 样本扫描步骤, 对同一检测样本进行多次扫描获得与标准片匹配的多个样本图像; 阈值计算步骤,分别获取多个样本图像中所述检测区域内的每个像素点的灰度值,计 算图像清晰度阈值。
3、根据权利要求 2所述的图像处理方法, 其中, 所述产品图像清晰度确定步骤包括: 产品扫描步骤, 扫描产品获得产品图像;
图像清晰度计算步骤,获取产品图像中所述检测区域内的每个像素点的灰度值,计算 产品图像清晰度。
4、 根据权利要求 2所述的图像处理方法, 其中, 所述阈值计算步骤包括: 获取每一样本图像在所述检测区域内像素点的灰度值;
分别计算每一样本图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
5、 根据权利要求 3所述的图像处理方法, 其中, 所述阈值计算步骤包括: 获取每一样本图像在所述检测区域内像素点的灰度值;
分别计算每一样本图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
6、 根据权利要求 4所述的图像处理方法, 其中, 所述图像清晰度计算步骤包括: 获取产品图像在所述检测区域内像素点的灰度值;
分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
7、 根据权利要求 5所述的图像处理方法, 其中, 所述图像清晰度计算步骤包括: 获取产品图像在所述检测区域内像素点的灰度值;
分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
8、 一种图像处理装置, 用于自动光学检测机中, 包括:
检测区域确定单元, 用于确定检测图像中矩形检测区域;
清晰度阈值确定单元,用于根据多个样本图像在所述检测区域内的每个像素点的灰度 值, 计算图像清晰度阈值;
产品图像清晰度确定单元,用于根据产品图像在所述检测区域内的每个像素点的灰度 值, 计算产品图像清晰度;
比较单元,用于比较所述产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度 大于所述图像清晰度阈值时,调用产品图像选定单元; 当产品图像清晰度小于所述图像清 晰度阈值时, 调用产品图像清晰度确定单元;
产品图像选定单元, 用于选定当前产品图像为待检测图像。
9、 根据权利要求 8所述的图像处理装置, 其中, 所述清晰度阈值确定单元还包括: 样本扫描子单元,用于对同一检测样本进行多次扫描获得与标准片匹配的多个样本图 像;
阈值计算子单元, 用于获取每一样本图像中所述检测区域内像素点的灰度值; 分别计算每一样本图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
10、根据权利要求 8所述的图像处理装置, 其中, 所述产品图像清晰度确定单元还包 括:
产品扫描子单元, 用于扫描产品获得产品图像;
图像清晰度计算子单元, 用于获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
11、根据权利要求 9所述的图像处理装置, 其中, 所述产品图像清晰度确定单元还包 括:
产品扫描子单元, 用于扫描产品获得产品图像;
图像清晰度计算子单元, 用于获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
12、 一种自动光学检测机, 包括图像处理装置, 用于自动光学检测机中; 所述图像处 理装置包括:
检测区域确定单元, 用于确定检测图像中矩形检测区域;
清晰度阈值确定单元,用于根据多个样本图像在所述检测区域内的每个像素点的灰度 值, 计算图像清晰度阈值;
产品图像清晰度确定单元,用于根据产品图像在所述检测区域内的每个像素点的灰度 值, 计算产品图像清晰度;
比较单元,用于比较所述产品图像清晰度与所述图像清晰度阈值; 当产品图像清晰度 大于所述图像清晰度阈值时,调用产品图像选定单元; 当产品图像清晰度小于所述图像清 晰度阈值时, 调用产品图像清晰度确定单元;
产品图像选定单元, 用于选定当前产品图像为待检测图像;
还包括检测装置, 用于根据所述待检测图像进行后续检测。
13、 根据权利要求 12所述的自动光学检测机, 其中, 所述清晰度阈值确定单元还包 括:
样本扫描子单元,用于对同一检测样本进行多次扫描获得与标准片匹配的多个样本图 像;
阈值计算子单元, 用于获取每一样本图像中所述检测区域内像素点的灰度值; 分别计算每一样本图像中所述检测区域内每一行相邻像素点灰度值的差值的绝对值, 选取绝对值的最大值为样本图像行灰度差值;选取样本图像行灰度差值的最大值为当前样 本图像的样本清晰度;
选取全部样本清晰度的最小值为所述图像清晰度阈值。
14、 根据权利要求 12所述的自动光学检测机, 其中, 所述产品图像清晰度确定单元 还包括:
产品扫描子单元, 用于扫描产品获得产品图像;
图像清晰度计算子单元, 用于获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
15、 根据权利要求 13所述的自动光学检测机, 其中, 所述产品图像清晰度确定单元 还包括:
产品扫描子单元, 用于扫描产品获得产品图像;
图像清晰度计算子单元, 用于获取产品图像在所述检测区域内像素点的灰度值; 分别计算产品图像在所述检测区域内每一行相邻像素点灰度值的差值的绝对值,选取 绝对值的最大值为产品图像行灰度差值;选取产品图像行灰度差值的最大值为产品图像清 晰度。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US14/375,714 US9646224B2 (en) | 2014-04-14 | 2014-05-15 | Image processing method, image processing device and automated optical inspection machine |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410147746.5 | 2014-04-14 | ||
| CN201410147746.5A CN103927749A (zh) | 2014-04-14 | 2014-04-14 | 图像处理方法、装置和自动光学检测机 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2015158024A1 true WO2015158024A1 (zh) | 2015-10-22 |
Family
ID=51145960
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2014/077579 Ceased WO2015158024A1 (zh) | 2014-04-14 | 2014-05-15 | 图像处理方法、装置和自动光学检测机 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103927749A (zh) |
| WO (1) | WO2015158024A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116152101A (zh) * | 2023-02-15 | 2023-05-23 | 河南科技大学 | 图片清晰度绝对值归一化评价方法 |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106934804B (zh) * | 2017-03-13 | 2019-12-13 | 重庆贝奥新视野医疗设备有限公司 | 图像清晰度检测方法及装置 |
| CN108961201A (zh) * | 2017-05-19 | 2018-12-07 | 广州康昕瑞基因健康科技有限公司 | 图像清晰度识别方法和自动聚焦方法 |
| CN107424128B (zh) * | 2017-06-20 | 2020-12-29 | 南京泰立瑞信息科技有限公司 | 一种从所录制的眼底视频中提取清晰图像帧的方法及系统 |
| CN107770520A (zh) * | 2017-11-13 | 2018-03-06 | 凌云光技术集团有限责任公司 | 一种线阵相机辅助对焦方法及装置 |
| CN110838099A (zh) * | 2019-10-10 | 2020-02-25 | 深圳市燕麦科技股份有限公司 | 一种异物检测方法、装置、系统及终端设备 |
| CN111368812B (zh) * | 2020-05-27 | 2020-12-08 | 北京伟杰东博信息科技有限公司 | 一种列车前方异物侵限检测方法及系统 |
| CN111626207B (zh) * | 2020-05-27 | 2021-07-06 | 深圳智锐通科技有限公司 | 一种基于图像处理的列车前方异物入侵检测方法及系统 |
| CN111626203B (zh) * | 2020-05-27 | 2021-11-16 | 北京伟杰东博信息科技有限公司 | 一种基于机器学习的铁路异物识别方法及系统 |
| CN111626204B (zh) * | 2020-05-27 | 2022-01-11 | 汪海洋 | 一种铁路异物入侵监测方法及系统 |
| CN115719360B (zh) * | 2022-11-25 | 2025-08-22 | 中国航空无线电电子研究所 | 一种面向复杂背景的红外弱小目标检测跟踪方法 |
| CN116091613A (zh) * | 2023-02-06 | 2023-05-09 | 盛视科技股份有限公司 | 一种摄像头清晰度检测方法 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090041349A1 (en) * | 2007-08-07 | 2009-02-12 | Masaru Suzuki | Image determining device, image determining method, and program |
| CN101527040A (zh) * | 2008-03-05 | 2009-09-09 | 深圳华为通信技术有限公司 | 图像处理方法及系统 |
| CN201440128U (zh) * | 2009-07-13 | 2010-04-21 | 北京航星科技有限公司 | 一种面向pcb缺陷检测的自动光学检测系统 |
| CN102609939A (zh) * | 2012-01-16 | 2012-07-25 | 北京航空航天大学 | 一种tfds图像质量评价方法及系统 |
| CN103686148A (zh) * | 2013-12-05 | 2014-03-26 | 北京华戎京盾科技有限公司 | 一种基于图像处理的自动检测视频图像清晰度的方法 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006140594A (ja) * | 2004-11-10 | 2006-06-01 | Pentax Corp | デジタルカメラ |
| CN102404602A (zh) * | 2011-09-23 | 2012-04-04 | 浙江工业大学 | 一种基于清晰度测试卡的摄像机清晰度检测方法 |
| CN102800082B (zh) * | 2012-06-19 | 2015-04-15 | 浙江捷尚视觉科技股份有限公司 | 一种无参考的图像清晰度检测方法 |
| CN103051836B (zh) * | 2012-12-14 | 2016-03-30 | 中兴通讯股份有限公司 | 移动终端防抖动拍照方法和装置 |
-
2014
- 2014-04-14 CN CN201410147746.5A patent/CN103927749A/zh active Pending
- 2014-05-15 WO PCT/CN2014/077579 patent/WO2015158024A1/zh not_active Ceased
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090041349A1 (en) * | 2007-08-07 | 2009-02-12 | Masaru Suzuki | Image determining device, image determining method, and program |
| CN101527040A (zh) * | 2008-03-05 | 2009-09-09 | 深圳华为通信技术有限公司 | 图像处理方法及系统 |
| CN201440128U (zh) * | 2009-07-13 | 2010-04-21 | 北京航星科技有限公司 | 一种面向pcb缺陷检测的自动光学检测系统 |
| CN102609939A (zh) * | 2012-01-16 | 2012-07-25 | 北京航空航天大学 | 一种tfds图像质量评价方法及系统 |
| CN103686148A (zh) * | 2013-12-05 | 2014-03-26 | 北京华戎京盾科技有限公司 | 一种基于图像处理的自动检测视频图像清晰度的方法 |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116152101A (zh) * | 2023-02-15 | 2023-05-23 | 河南科技大学 | 图片清晰度绝对值归一化评价方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103927749A (zh) | 2014-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2015158024A1 (zh) | 图像处理方法、装置和自动光学检测机 | |
| CN107507571B (zh) | 一种对amoled进行外部光学补偿的方法及装置 | |
| CN110225336B (zh) | 评估图像采集精度的方法及装置、电子设备、可读介质 | |
| CN109345528B (zh) | 一种基于人眼视觉特性的显示屏缺陷检测方法及装置 | |
| CN103792705B (zh) | 检测基板缺陷的检测方法及检测装置 | |
| CN103885217B (zh) | 一种检测液晶面板柱状隔垫物缺陷的方法及装置 | |
| CN103809309B (zh) | 基板检测设备及方法 | |
| US9646224B2 (en) | Image processing method, image processing device and automated optical inspection machine | |
| WO2015035744A1 (zh) | 一种图像处理方法 | |
| US20170048518A1 (en) | Method and apparatus for adjusting installation flatness of lens in real time | |
| TW201546781A (zh) | 影像分析和影像顯示的裝置和方法 | |
| CN104166250A (zh) | 一种平面显示器面板均匀度检测方法及其系统 | |
| KR101946581B1 (ko) | 패널 검사 방법 | |
| CN106713907B (zh) | 一种显示器的hdr图像显示性能评测方法及装置 | |
| KR20200081541A (ko) | 촬상 장치 및 이의 구동 방법 | |
| CN105427315B (zh) | 数字仪表图像位置测试方法及装置 | |
| CN112465780A (zh) | 绝缘层膜厚异常监控方法及装置 | |
| JP5846100B2 (ja) | 表示装置の欠陥検査方法 | |
| JPH08327497A (ja) | カラー液晶表示パネルの検査方法 | |
| CN105509643A (zh) | 一种亚像素单元线宽的测量方法及装置 | |
| TW201518695A (zh) | 光源穩定性測試系統及方法 | |
| KR100842460B1 (ko) | 평판 디스플레이 패널의 도트 결함 검출 방법 | |
| CN110672631A (zh) | 面板缺陷拍照方法和面板缺陷拍照装置 | |
| JP5603964B2 (ja) | 平面表示パネルの検査方法、検査装置及び検査プログラム | |
| KR100765294B1 (ko) | 액정 디스플레이 패널의 얼룩 결함 검출 방법 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| WWE | Wipo information: entry into national phase |
Ref document number: 14375714 Country of ref document: US |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14889711 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: 14889711 Country of ref document: EP Kind code of ref document: A1 |