WO2022062812A1 - 屏幕缺陷检测方法、装置和电子设备 - Google Patents

屏幕缺陷检测方法、装置和电子设备 Download PDF

Info

Publication number
WO2022062812A1
WO2022062812A1 PCT/CN2021/114453 CN2021114453W WO2022062812A1 WO 2022062812 A1 WO2022062812 A1 WO 2022062812A1 CN 2021114453 W CN2021114453 W CN 2021114453W WO 2022062812 A1 WO2022062812 A1 WO 2022062812A1
Authority
WO
WIPO (PCT)
Prior art keywords
area
defect
screen
suspected
image
Prior art date
Application number
PCT/CN2021/114453
Other languages
English (en)
French (fr)
Inventor
宋秀峰
张一凡
刘杰
张文超
Original Assignee
歌尔股份有限公司
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 歌尔股份有限公司 filed Critical 歌尔股份有限公司
Priority to US18/246,806 priority Critical patent/US20230368361A1/en
Publication of WO2022062812A1 publication Critical patent/WO2022062812A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0004Industrial image inspection
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/11Region-based segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/60Analysis of geometric attributes
    • G06T7/62Analysis of geometric attributes of area, perimeter, diameter or volume
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30108Industrial image inspection
    • G06T2207/30164Workpiece; Machine component

Definitions

  • the present application relates to the field of workpiece detection, and in particular, to a screen defect detection method, device and electronic device.
  • the quality of the screen has a great impact on the user experience, especially for VR (Virtual Reality, virtual reality) devices, AR (Augmented Reality, augmented reality) devices and other devices that mainly rely on the screen to interact with users.
  • VR Virtual Reality, virtual reality
  • AR Augmented Reality, augmented reality
  • Many screen manufacturers will perform defect detection on the screen before it leaves the factory, such as detecting whether the screen has dead pixels, scratches, etc., and many times rely on manual inspection, which is not efficient.
  • defect detection on the screen before it leaves the factory, such as detecting whether the screen has dead pixels, scratches, etc., and many times rely on manual inspection, which is not efficient.
  • transparent defects caused by contamination of foreign objects on the screen it is difficult to find manually, and the rate of missed detection and false detection is high.
  • the screen defect detection method, device and electronic device of the present application are proposed to overcome the above problems.
  • a screen defect detection method comprising: identifying a number of suspected defective pixel points from a detection image of a target screen; determining a suspected defect corresponding to the suspected defective pixel point in the detection image area; divide the suspected defect area to obtain a general area and a core area; judge the target according to the gray mean value of the suspected defect area, the gray mean value of the general area, and the gray value minimum value of the core area Whether the screen is defective.
  • a screen defect detection device comprising:
  • an identification unit for identifying a number of suspected defective pixels from the detection image of the target screen
  • the suspected defect area determination unit is used to determine the suspected defect area corresponding to the suspected defect pixel point in the inspection image
  • the division unit is used to divide the suspected defect area to obtain the general area and the core area;
  • the judgment unit is used for judging whether the target screen has a transparent defect according to the average gray value of the suspected defect area, the average gray value of the general area and the minimum value of the gray value of the core area.
  • an electronic device comprising: a processor; and a memory arranged to store computer-executable instructions that, when executed, cause the processor to perform any of the above The described screen defect detection method.
  • a computer-readable storage medium storing one or more programs that, when executed by an electronic device including a plurality of application programs, cause the electronic device to Perform the screen defect detection method as described in any of the above.
  • an automated screen defect detection solution replaces manual detection, improves efficiency and saves labor costs; uses image filtering algorithms and other methods to first detect suspected defective pixels, and then The combination of area division and grayscale comparison finally determines whether the target screen has transparent defects, which improves the accuracy of defect detection.
  • FIG. 1 shows a schematic flowchart of a method for detecting screen defects according to an embodiment of the present application
  • Figure 2 shows a grayscale image containing a screen
  • Fig. 3 is the binarized image of Fig. 2;
  • Fig. 4 is the schematic diagram that identifies the outline in Fig. 2;
  • Fig. 5 shows the detection image obtained according to Fig. 2;
  • FIG. 6 shows a schematic diagram of a suspected defective area according to an embodiment of the present application.
  • Fig. 7 is an image obtained by binarizing a suspected defect area
  • FIG. 8 shows a schematic structural diagram of a screen defect detection apparatus according to an embodiment of the present application.
  • FIG. 9 shows a schematic structural diagram of an electronic device according to an embodiment of the present application.
  • FIG. 10 shows a block diagram of a computer-readable storage medium according to an embodiment of the present application.
  • the transparent defects of the screen mainly refer to the defects formed due to the adhesion of fluff, paper scraps, etc. on the surface of the screen.
  • "Transparency” means that the screen can emit light through these foreign objects, resulting in the screen glowing in this part, but the brightness is darker. condition. This kind of defect generally does not need to be returned to the factory for treatment, and can be improved by wiping, but if it is assembled or packaged for sale without treatment, the user will still get a poor experience.
  • the technical concept of the present application is: considering the manifestation of transparent defects, the gray value is used to determine the transparent defects, and the defect area can also be used for review, thereby avoiding the problems of time-consuming and laborious manual detection and low accuracy.
  • the solution of the present application can be applied to various types of screens, such as screens of VR glasses and so on.
  • FIG. 1 shows a schematic flowchart of a screen defect detection method according to an embodiment of the present application.
  • the screen defect detection method includes:
  • step S110 several suspected defective pixels are identified from the detected image of the target screen.
  • the detection image of the target screen is the basis for automatic defect detection, which can generally be obtained by photographing the target screen.
  • the detection image may be a gray image in particular.
  • the above-mentioned screen defect detection method further includes: acquiring an original image that can contain the target screen; converting the original image into a grayscale image; detecting a valid image corresponding to the target screen from the grayscale image
  • the grayscale image is cropped according to the effective area to obtain the detection image of the target screen.
  • the camera can also be selected to directly capture a grayscale image including the target screen, omitting the step of converting the original image into a grayscale image.
  • Screen detection usually only requires the image content of this part of the screen, and when shooting the screen, it is inevitable to capture the screen border, etc., so the grayscale image can be cropped. Specifically, it can be achieved in the following ways:
  • the screen contour is searched based on the binarized image of the grayscale image, and the screen contour is obtained; the effective area is determined according to the circumscribed rectangle of the screen contour.
  • Figure 2 shows a grayscale image containing a screen. It can be seen that there is a light gray screen border around the screen, which will affect subsequent processing.
  • the binarization process can be performed on FIG. 2 to obtain a binarized image, as shown in FIG. 3 , and then the screen contour search is performed on the binarized image. This step can be realized by relying on the prior art.
  • map the searched screen contour line to the grayscale image, as shown in Fig. 4 it can be seen that the white line around the screen is the screen contour line.
  • the circumscribed rectangle is determined according to the outline of the screen, and the detection image as shown in Figure 5 is obtained by cropping.
  • the identification of suspected defective pixels can be realized by a filter based on an image filtering algorithm.
  • the filter here generally refers to filtering software, such as machine vision software halcon.
  • the filter can detect various types of pixel points.
  • corresponding parameters can be set based on the characteristics of transparent defects, for example, pixels with darker brightness can be screened out as suspected defective pixels.
  • step S120 a suspected defective area corresponding to the suspected defective pixel point is determined in the inspection image.
  • Suspected defective pixels are generally scattered points, and there may be false detection or missed detection. Therefore, it can be realized by determining the suspected defective area corresponding to the suspected defective pixel, and performing subsequent detection based on the suspected defective area, instead of relying only on Suspected defective pixels themselves.
  • the suspected defective area corresponding to each suspected defective pixel point may be determined separately, that is, a one-to-one correspondence is achieved.
  • Step S130 dividing the suspected defective area to obtain a general area and a core area.
  • the general area mainly corresponds to normal pixels
  • the core area mainly corresponds to suspected defective pixels, so as to facilitate the calculation and comparison of gray values below.
  • Step S140 according to the average gray value of the suspected defective area, the average gray value of the general area, and the minimum value of the gray value of the core area, determine whether the target screen has a transparent defect. In this way, whether there is a transparent defect on the target screen is judged by judging whether there is an abnormal local gray value in the suspected defect area, and the accuracy is very high.
  • the method shown in Figure 1 replaces manual detection through an automated screen defect detection scheme, which improves efficiency and saves labor costs; using image filtering algorithms and other methods to first detect suspected defective pixels, and then through regional division, The combination of grayscale comparison method finally determines whether the target screen has transparent defects, which improves the accuracy of defect detection.
  • determining the suspected defect area corresponding to the suspected defective pixel point in the detection image includes: Set the value as the side length to extract the square area as the suspected defective area corresponding to the suspected defective pixel point.
  • the first preset value may be determined to be 50 (pixels).
  • dividing the suspected defective area to obtain the general area and the core area includes: in the suspected defective area including the suspected defective pixel point, taking the suspected defective pixel point as the center of the circle , using the second preset value to divide a circular area as the core area corresponding to the suspected defective pixel point; all areas except the core area in the suspected defective area are general areas corresponding to the suspected defective pixel point.
  • FIG. 6 shows a schematic diagram of a suspected defect area according to an embodiment of the present application.
  • the circular area in the figure is a core area, and the area outside the circle is a general area. The combination of the two forms a suspected defect area.
  • the circle as the core area can cover the pixels representing transparent defects in the suspected defect area as fully as possible, and minimize the coverage of normal pixels.
  • the square suspected defect area is convenient for statistical calculation and extraction.
  • the second preset value may be selected to be 15 (pixels).
  • judging whether the target screen has a transparent defect according to the average gray value of the suspected defective area, the average gray value of the general area, and the minimum value of the gray value of the core area includes: calculating the core The ratio of the gray minimum value of the area to the gray average value of the general area is used as the local contrast; the ratio of the gray minimum value of the core area to the gray average value of the suspected defect area is calculated as the global contrast; the local contrast and the global contrast are used to judge the target screen. Whether there is a transparency defect.
  • the boundary between the core area and the general area may be marked with contour lines. Then, the pixels belonging to the core area and the general area can be determined according to the contour lines, and the gray average value of the suspected defect area, the gray average value of the general area, and The grayscale minimum value of the core region.
  • the pointPolygonTest(vec,cv::Point(x,y),false) function can be used to determine whether the pixel is inside the contour line or outside the contour line.
  • gray_min is the gray minimum value of the core area.
  • the determining whether the target screen has a transparency defect by using the local contrast and the global contrast includes: if the absolute value of the difference between the local contrast and the global contrast is in the first Between a threshold value and a second threshold value, the target screen has a transparency defect.
  • the judgment can also be realized by setting only a minimum threshold and other methods.
  • screen defects such as screen dead pixels may also cause similar appearances.
  • these defects are generally accidental.
  • there are one or two dead pixels on the screen which may also cause low brightness (no light) of individual points on the screen. Therefore, transparent defects can be distinguished from other defects by calculating the defect area. . That is, in an embodiment of the present application, in the above-mentioned screen defect detection method, when it is determined that the target screen has a transparent defect, the defect area in the suspected defect area is calculated; if the defect area is smaller than the area threshold, it is determined that the transparent defect is an error. other deficiencies found.
  • 6 suspected defective pixels are identified in the inspection image, then 6 suspected defective areas are obtained, which are determined according to the average gray value of the suspected defective area, the average gray value of the general area, and the minimum value of the gray value of the core area. If four suspected defect areas are identified, which respectively represent a transparent defect, in the further misjudgment detection, the defect areas in these four suspected defect areas are calculated respectively, and each transparent defect is judged whether it is a misjudged other defect. .
  • calculating the defect area includes: performing adaptive binarization processing on the suspected defect area, so as to obtain a first color to indicate that the screen is normal, and a second color to indicate that the screen is normal
  • the binarized image of the defect is taken as the defect area.
  • FIG. 7 is an image obtained by binarizing a suspected defect area, in which black indicates that the screen is normal, and white indicates screen defects. It is easy to understand that the expressions of colors can be reversed, and FIG. 7 is an exemplary image that can highlight the defective part. It is more complicated to directly obtain the area of the white area, and according to the cause of the transparent defect, foreign objects are often aggregated, so the defect area can be obtained by finding the circumscribed rectangle. Of course, in other embodiments, it can also be obtained by finding the circumscribed circle, etc. As the defect area, the circumscribed rectangle has the advantage of low computational complexity.
  • the transparent defect detection can be carried out only through the aforementioned statistical comparison of gray values, and then manually reviewed; and if a higher detection accuracy rate is desired, the The transparent defects detected by the statistical comparison of gray values are regarded as suspected transparent defects, and then the existence of transparent defects is finally determined by calculating the defect area.
  • FIG. 8 shows a schematic structural diagram of a screen defect detection apparatus according to an embodiment of the present application.
  • the screen defect detection apparatus 800 includes:
  • the identifying unit 810 is configured to identify several suspected defective pixel points from the detected image of the target screen.
  • the detection image of the target screen is the basis for automatic defect detection, which can generally be obtained by photographing the target screen.
  • the detection image may be a gray image in particular.
  • the suspected defective area determination unit 820 determines the suspected defective area corresponding to the suspected defective pixel point in the inspection image.
  • Suspected defective pixels are generally scattered points, and there may be false detection or missed detection. Therefore, it can be realized by determining the suspected defective area corresponding to the suspected defective pixel, and performing subsequent detection based on the suspected defective area, instead of relying only on Suspected defective pixels themselves.
  • the suspected defective area corresponding to each suspected defective pixel point may be determined separately, that is, a one-to-one correspondence is achieved.
  • the dividing unit 830 is configured to divide the suspected defect area to obtain a general area and a core area.
  • the general area mainly corresponds to normal pixels
  • the core area mainly corresponds to suspected defective pixels, so as to facilitate the calculation and comparison of gray values below.
  • the judging unit 840 is configured to judge whether the target screen has a transparent defect according to the average gray value of the suspected defective area, the average gray value of the general area and the minimum value of the gray value of the core area.
  • the device shown in Figure 8 replaces manual detection through an automated screen defect detection scheme, improves efficiency and saves labor costs; using image filtering algorithms and other methods to first detect suspected defective pixels, and then through regional division, The combination of grayscale comparison method finally determines whether the target screen has transparent defects, which improves the accuracy of defect detection.
  • the determination unit 840 is configured to calculate the ratio of the minimum gray value of the core area to the average gray value of the general area as the local contrast; The ratio of the minimum gray value to the average gray value of the suspected defect area is used as the global contrast; whether the target screen has a transparent defect is determined by using the local contrast and the global contrast.
  • the determination unit 840 is specifically configured to, when the absolute value of the difference between the local contrast and the global contrast is between the first threshold and the second threshold, It is determined that the target screen has a transparency defect.
  • the judgment unit 840 is further configured to calculate the defect area in the suspected defect area when it is determined that the target screen has a transparent defect; if the defect area is smaller than the area threshold , then it is determined that the transparent defect is another defect that is misjudged.
  • the judging unit 840 is configured to perform adaptive binarization processing on the suspected defect area, and obtain a second color that indicates that the screen is normal and the second color indicates that the screen is defective. Binarizing the image; taking the area of the circumscribed rectangle of the pixel points of the second color in the binarizing image as the defect area.
  • the above-mentioned device further includes: a detection image acquisition unit, configured to acquire an original image that can include the target screen; convert the original image into a grayscale image; detect from the grayscale image For an effective area corresponding to the target screen, the grayscale image is trimmed according to the effective area to obtain a detection image of the target screen.
  • a detection image acquisition unit configured to acquire an original image that can include the target screen; convert the original image into a grayscale image; detect from the grayscale image For an effective area corresponding to the target screen, the grayscale image is trimmed according to the effective area to obtain a detection image of the target screen.
  • the suspected defective area determination unit 820 is configured to, in the detection image, take the suspected defective pixel as the center, and use the first preset value to extract a square area as the side length. The suspected defective area corresponding to the suspected defective pixel point.
  • the area dividing unit 830 is configured to divide the suspected defective pixel point as the center of the circle in the suspected defective area including the suspected defective pixel point, and use the second preset value to divide the radius
  • a circular area is drawn as the core area corresponding to the suspected defective pixel point; all areas in the suspected defective area except the core area are general areas corresponding to the suspected defective pixel point.
  • the technical solution of the present application is aimed at the problem that transparent defects are difficult to detect, and the inspection orientation needs to be changed at all times during manual inspection, which will increase the labor intensity of operators.
  • the method of comparison and defect area review is used to determine whether the target screen has transparent defects. It is proved by experiments that the solution proposed in this application can be well applied to the automatic detection of transparent defects of OLED and other types of screens in the production line, which not only reduces the cost, but also reduces the The operation damages the eyes of the workers, improves the work efficiency and reduces the missed detection rate.
  • modules in the device in the embodiment can be adaptively changed and arranged in one or more devices different from the embodiment.
  • the modules or units or components in the embodiments may be combined into one module or unit or component, and further they may be divided into multiple sub-modules or sub-units or sub-assemblies. All features disclosed in this specification (including accompanying claims, abstract and drawings) and any method so disclosed may be employed in any combination, unless at least some of such features and/or procedures or elements are mutually exclusive. All processes or units of equipment are combined.
  • Each feature disclosed in this specification may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
  • Various component embodiments of the present application may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof.
  • a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all functions of some or all of the components in the apparatus 800 for detecting screen defects according to embodiments of the present application.
  • DSP digital signal processor
  • the present application can also be implemented as an apparatus or apparatus program (eg, computer programs and computer program products) for performing part or all of the methods described herein.
  • Such a program implementing the present application may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from Internet sites, or provided on carrier signals, or in any other form.
  • FIG. 9 shows a schematic structural diagram of an electronic device according to an embodiment of the present application.
  • the electronic device 900 includes a processor 910 and a memory 920 arranged to store computer-executable instructions (computer-readable program code).
  • the memory 920 may be electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, hard disk, or ROM.
  • the memory 920 has storage space 930 for storing computer readable program code 931 for performing any of the method steps in the above-described methods.
  • the storage space 930 for storing computer-readable program code may include various computer-readable program codes 931 for implementing various steps in the above methods, respectively.
  • Computer readable program code 931 can be read from or written to one or more computer program products.
  • FIG. 10 shows a schematic structural diagram of a computer-readable storage medium according to an embodiment of the present application.
  • the computer-readable storage medium 1000 stores computer-readable program code 931 for performing the method steps according to the present application, which can be read by the processor 910 of the electronic device 900 when the computer-readable program code 931 is executed by the electronic device 900 , causing the electronic device 900 to execute each step in the above-described method.
  • the computer-readable program code 931 stored in the computer-readable storage medium can execute the method shown in any of the above-described embodiments.
  • the computer readable program code 931 may be compressed in a suitable form.

Abstract

一种屏幕缺陷检测方法、装置和电子设备。所述方法包括:从目标屏幕的检测图像中识别出若干个疑似缺陷像素点(S110);在所述检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域(S120);对所述疑似缺陷区域进行划分得到一般区域和核心区域(S130);根据所述疑似缺陷区域的灰度均值、所述一般区域的灰度均值以及所述核心区域的灰度最小值判断所述目标屏幕是否存在透明缺陷(S140)。通过自动化的屏幕缺陷检测方案,替代了人工检测,提升了效率并节约了人工成本;利用图像滤波算法等方式先检测出疑似缺陷像素点,再通过区域划分、灰度比较方式的结合最终确定目标屏幕是否存在透明缺陷,提升了缺陷检测的准确性。

Description

屏幕缺陷检测方法、装置和电子设备 技术领域
本申请涉及工件检测领域,特别涉及屏幕缺陷检测方法、装置和电子设备。
发明背景
屏幕的质量对用户体验有着很大影响,尤其是对VR(Virtual Reality,虚拟现实)设备、AR(Augmented Reality,增强现实)设备等主要依赖屏幕与用户进行互动的设备更是如此。许多屏幕制造商都会在屏幕出厂前对其进行缺陷检测,例如检测屏幕是否存在坏点、划痕等,很多时候要依赖人工检测,效率不高。尤其是对于因为屏幕上沾染异物等原因而产生的透明缺陷,人工很难发现,漏检、误检率较高。
发明内容
鉴于现有技术无法有效检测屏幕中存在缺陷的问题,提出了本申请的屏幕缺陷检测方法、装置和电子设备,以便克服上述问题。
为了实现上述目的,本申请采用了如下技术方案:
依据本申请的一个方面,提供了一种屏幕缺陷检测方法,包括:从目标屏幕的检测图像中识别出若干个疑似缺陷像素点;在所述检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域;对所述疑似缺陷区域进行划分得到一般区域和核心区域;根据所述疑似缺陷区域的灰度均值、所述一般区域的灰度均值以及所述核心区域的灰度最小值判断所述目标屏幕是否存在缺陷。
依据本申请的另一方面,提供了一种屏幕缺陷检测装置,包括:
识别单元,用于从目标屏幕的检测图像中识别出若干个疑似缺陷像素点;
疑似缺陷区域确定单元,用于在检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域;
划分单元,用于对疑似缺陷区域进行划分得到一般区域和核心区域;
判断单元,用于根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值判断目标屏幕是否存在透明缺陷。
依据本申请的又一方面,提供了一种电子设备,包括:处理器;以及被安排成存储计算机可执行指令的存储器,所述可执行指令在被执行时使所述处理器执行如上任一所述的屏幕缺陷检测方法。
依据本申请的再一方面,提供了一种计算机可读存储介质,存储一个或多个程序,所述一个或多个程序当被包括多个应用程序的电子设备执行时,使得所述电子设备执行如上任一所述的屏幕缺陷检测方法。
综上所述,本申请的有益效果是:通过自动化的屏幕缺陷检测方案,替代了人工检测,提升了效率并节约了人工成本;利用图像滤波算法等方式先检测出疑似缺陷像素点,再通过区域划分、灰度比较方式的结合最终确定目标屏幕是否存在透明缺陷,提升了缺陷检测的准确性。
附图简要说明
图1示出了根据本申请一个实施例的一种屏幕缺陷检测方法的流程示意图;
图2示出了包含屏幕的灰度图像;
图3为图2的二值化图像;
图4是在图2中标识出轮廓线的示意图;
图5示出了根据图2得到的检测图像;
图6示出了根据本申请一个实施例的一个疑似缺陷区域的示意图;
图7是对一个疑似缺陷区域进行二值化处理得到的图像;
图8示出了根据本申请一个实施例的一种屏幕缺陷检测装置的结构示意图;
图9示出了根据本申请一个实施例的一种电子设备的结构示意图;
图10示出了根据本申请一个实施例的一种计算机可读存储介质的框图。
具体实施方式
下面将参照附图更详细地描述本申请的示例性实施例。虽然附图中显示了本申请的示例性实施例,然而应当理解,可以以各种形式实现本申请而不应被这里阐述的实施例所限制。相反,提供这些实施例是为了能够更透彻地理解本申请,并且能够将本申请的范围完整的传达给本领域的技术人员。
屏幕的透明缺陷,主要是指因为屏幕表面因为粘上绒毛、纸屑等原因形成的缺陷,“透明”是指屏幕发光可以透过这些异物,产生屏幕在该部位虽然发光,但亮度较暗的情况。这种缺陷一般不需要返厂处理,通过擦拭即可改善,但如不进行处理就组装或包装发售等,仍会使用户得到较差的体验。
本申请的技术构思是:考虑到透明缺陷的表现形式,利用灰度值确定透明缺陷,还可以利用缺陷面积进行复核,从而避免了人工检测耗时费力,准确度 不高的问题。本申请的方案可以适用于各类屏幕,例如VR眼镜的屏幕等等。
下面结合具体实施例进行本申请技术方案的示例性说明。
图1示出了根据本申请一个实施例的一种屏幕缺陷检测方法的流程示意图。如图1所示,屏幕缺陷检测方法包括:
步骤S110,从目标屏幕的检测图像中识别出若干个疑似缺陷像素点。
得到目标屏幕的检测图像是实现自动化缺陷检测的基础,一般可以通过对目标屏幕进行拍摄得到。而为了便于后续灰度值的统计和计算,检测图像可以具体是灰度图像。
例如,在本申请的一个实施例中,上述屏幕缺陷检测方法还包括:获取能够包含目标屏幕的原始图像;将原始图像转换为灰度图像;从灰度图像中检测出与目标屏幕对应的有效区域,根据有效区域对灰度图像进行剪裁,得到目标屏幕的检测图像。
当然,在其他实施例中,也可以选择使用相机直接拍摄包含目标屏幕的灰度图像,省去将原始图像转换为灰度图像的步骤。
进行屏幕检测通常只需要屏幕这部分的图像内容,而在对屏幕进行拍摄时,难免会将屏幕边框等拍摄进来,因此可以对灰度图像进行剪裁。具体地,可以通过如下方式实现:
基于灰度图像的二值化图像进行屏幕轮廓搜索,得到屏幕轮廓线;根据屏幕轮廓线的外接矩形确定有效区域。
例如,图2示出了包含屏幕的灰度图像。可以看出,在屏幕周围有一圈呈淡灰色的屏幕边框,会影响后续的处理。为解决这一问题,可以对图2进行二值化处理,得到二值化图像,如图3所示,然后对二值化图像进行屏幕轮廓搜索,这一步骤可以依赖于现有技术实现。再根据图2与图3的对应关系,将搜索到的屏幕轮廓线映射到灰度图像中,如图4所示,可以看到屏幕周围的白线即是屏幕轮廓线。接下来根据屏幕轮廓线确定外接矩形,裁剪得到如图5所示的检测图像。
疑似缺陷像素点的识别可以利用基于图像滤波算法的滤波器实现,这里的滤波器一般是指滤波软件,如机器视觉软件halcon。滤波器可以检测出多种类型的像素点,在本申请的实施例中,可以基于透明缺陷的特点设置相应的参数,例如筛选出亮度较暗的像素点作为疑似缺陷像素点。
步骤S120,在检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域。
疑似缺陷像素点一般是散点,而且可能存在误检或漏检,因此可以通过确定与疑似缺陷像素点对应的疑似缺陷区域,基于疑似缺陷区域进行后续检测的方式来实现,而不仅仅依赖于疑似缺陷像素点本身。
需要说明的是,检测出的疑似缺陷像素点可能有多个,那么可以分别确定与每个疑似缺陷像素点对应的疑似缺陷区域,即实现一一对应。
步骤S130,对疑似缺陷区域进行划分得到一般区域和核心区域。这里一般区域主要是对应正常的像素点,核心区域主要对应疑似缺陷像素点,以方便下面进行灰度值的计算和比较。
步骤S140,根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值判断目标屏幕是否存在透明缺陷。这样就通过判断疑似缺陷区域中是否存在局部灰度值异常来判断目标屏幕是否存在透明缺陷,准确度很高。
可见,图1所示的方法,通过自动化的屏幕缺陷检测方案,替代了人工检测,提升了效率并节约了人工成本;利用图像滤波算法等方式先检测出疑似缺陷像素点,再通过区域划分、灰度比较方式的结合最终确定目标屏幕是否存在透明缺陷,提升了缺陷检测的准确性。
在本申请的一个实施例中,上述屏幕缺陷检测方法中,在检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域包括:在检测图像中,以疑似缺陷像素点为中心,以第一预设值为边长提取正方形区域作为与该疑似缺陷像素点对应的疑似缺陷区域。
根据对实际场景中透明缺陷的分析,一般造成透明缺陷的异物不会很大。因此在一些实施例中,可以将第一预设值确定为50(像素)。
在本申请的一个实施例中,上述屏幕缺陷检测方法中,对疑似缺陷区域进行划分得到一般区域和核心区域包括:在包含疑似缺陷像素点的疑似缺陷区域中,以该疑似缺陷像素点为圆心,以第二预设值为半径划分出圆形区域作为与该疑似缺陷像素点对应的核心区域;疑似缺陷区域中除核心区域以外的全部区域为与该疑似缺陷像素点对应的一般区域。
例如,图6示出了根据本申请一个实施例的一个疑似缺陷区域的示意图,图中圆形区域为核心区域,圆形外的区域为一般区域,二者组合形成疑似缺陷区域。
以圆形作为核心区域,能够实现尽可能全面覆盖疑似缺陷区域中表征透明 缺陷的像素点,并尽量减少覆盖正常像素点。而正方形的疑似缺陷区域可以方便统计计算和提取。在一些实施例中,第二预设值可以选择为15(像素)。
在本申请的一个实施例中,上述屏幕缺陷检测方法中,根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值判断目标屏幕是否存在透明缺陷包括:计算核心区域的灰度最小值与一般区域的灰度均值的比值作为局部对比度;计算核心区域的灰度最小值与疑似缺陷区域的灰度均值的比值作为全局对比度;利用局部对比度和全局对比度判断目标屏幕是否存在透明缺陷。
具体来说,在划分核心区域与一般区域时,可以将核心区域与一般区域的边界以轮廓线进行标记。则可以根据轮廓线确定分别属于核心区域与一般区域的各像素点,根据各像素点的灰度值和各像素点所在的区域,确定疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值。
例如,可以通过pointPolygonTest(vec,cv::Point(x,y),false)函数判定像素点在轮廓线内还是在轮廓线外。
然后,统计轮廓线外(即一般区域)的像素点灰度值灰度和为gray_out_sum、像素点总个数为sum_out,从而可计算出一般区域的灰度均值gray_out_mean=gray_out_sum/sum_out。
统计疑似缺陷区域的像素点灰度值灰度和为gray_sum,以第一预设值取50,即疑似缺陷区域为50*50像素区域为例,则疑似缺陷区域的灰度均值gray_global_mean=gray_sum/(50*50)。
那么,局部对比度gray_mean_contrast=gray_min/gray_out_mean,全局对比度gray_global_mean_contrast=gray_min/gray_global_mean。其中,gray_min就是核心区域的灰度最小值。
然后,可以根据gray_mean_contrast与gray_global_mean_contrast来判断目标屏幕是否存在透明缺陷。在本申请的一个实施例中,所述利用所述局部对比度和所述全局对比度判断所述目标屏幕是否存在透明缺陷包括:若所述局部对比度和所述全局对比度的差值的绝对值在第一阈值与第二阈值之间,则所述目标屏幕存在透明缺陷。
即如果gray_mean_contrast与gray_global_mean_contrast差值的绝对值大于第一阈值thres_contrast_min,小于第二阈值thres_contrast_max,则判定屏幕存在透明缺陷。
当然,除了上面给出的以局部对比度和全局对比度之间的差值绝对值和两个预设阈值的比较方式外,也可以通过仅设置一个最小阈值等其他方式实现判断。
除了透明缺陷外,屏幕坏点等屏幕自身缺陷也可能造成相类似的表象。但这些缺陷一般是偶发性的,例如屏幕存在一到两个坏点,也可能造成屏幕的个别点亮度低(不发光),因此,可以通过计算缺陷面积的方式将透明缺陷与其他缺陷区分开。即在本申请的一个实施例中,上述屏幕缺陷检测方法中,在确定目标屏幕存在透明缺陷的情况下,计算疑似缺陷区域中的缺陷面积;若缺陷面积小于面积阈值,则确定透明缺陷为误判的其他缺陷。
举个例子,检测图像中识别到了6个疑似缺陷像素点,则得到6个疑似缺陷区域,分别根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值,确定了4个疑似缺陷区域,分别表征一处透明缺陷,则在进一步误判的检测时,分别计算这4个疑似缺陷区域中的缺陷面积,分别判断每一处透明缺陷是否为误判的其他缺陷。
具体地,在本申请的一个实施例中,上述屏幕缺陷检测方法中,计算缺陷面积包括:对疑似缺陷区域进行自适应二值化处理,得到以第一颜色表征屏幕正常,第二颜色表征屏幕缺陷的二值化图像;将二值化图像中呈现第二颜色的像素点的外接矩形的面积作为缺陷面积。
例如,图7是对一个疑似缺陷区域进行二值化处理得到的图像,其中以黑色表征屏幕正常,白色表征屏幕缺陷。容易理解颜色的表达可以进行对调,图7是能够凸显缺陷部分的示例性图像。直接求得白色区域的面积会比较复杂,而根据透明缺陷的成因,异物往往是聚集存在的,因此可以通过求外接矩形的方式得到缺陷面积,当然在其他实施例中也可以通过求外接圆等方式来作为缺陷面积,而外接矩形具有计算量低的优点。
可以看出,通过这种方式可以避免误检。在具体实施中,如果追求前期的检测效率,可以仅通过前述进行灰度值统计比较的方式进行透明缺陷的检测,再通过人工进行复核;而如果希望有较高的检测准确率,则可以将灰度值统计比较检测出的透明缺陷认为是疑似透明缺陷,再通过求缺陷面积的方式最终确定是否存在透明缺陷。
图8示出了根据本申请一个实施例的一种屏幕缺陷检测装置的结构示意图。如图8所示,屏幕缺陷检测装置800包括:
识别单元810,用于从目标屏幕的检测图像中识别出若干个疑似缺陷像素点。
得到目标屏幕的检测图像是实现自动化缺陷检测的基础,一般可以通过对目标屏幕进行拍摄得到。而为了便于后续灰度值的统计和计算,检测图像可以具体是灰度图像。
疑似缺陷区域确定单元820,在检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域。
疑似缺陷像素点一般是散点,而且可能存在误检或漏检,因此可以通过确定与疑似缺陷像素点对应的疑似缺陷区域,基于疑似缺陷区域进行后续检测的方式来实现,而不仅仅依赖于疑似缺陷像素点本身。
需要说明的是,检测出的疑似缺陷像素点可能有多个,那么可以分别确定与每个疑似缺陷像素点对应的疑似缺陷区域,即实现一一对应。
划分单元830,用于对疑似缺陷区域进行划分得到一般区域和核心区域。这里一般区域主要是对应正常的像素点,核心区域主要对应疑似缺陷像素点,以方便下面进行灰度值的计算和比较。
判断单元840,用于根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值判断目标屏幕是否存在透明缺陷。
这样就通过判断疑似缺陷区域中是否存在局部灰度值异常来判断目标屏幕是否存在透明缺陷,准确度很高。
可见,图8所示的装置,通过自动化的屏幕缺陷检测方案,替代了人工检测,提升了效率并节约了人工成本;利用图像滤波算法等方式先检测出疑似缺陷像素点,再通过区域划分、灰度比较方式的结合最终确定目标屏幕是否存在透明缺陷,提升了缺陷检测的准确性。
在本申请的一个实施例中,上述装置中,判断单元840,用于计算所述核心区域的灰度最小值与所述一般区域的灰度均值的比值作为局部对比度;计算所述核心区域的灰度最小值与所述疑似缺陷区域的灰度均值的比值作为全局对比度;利用所述局部对比度和所述全局对比度判断所述目标屏幕是否存在透明缺陷。
在本申请的一个实施例中,上述装置中,判断单元840,具体用于在所述局部对比度和所述全局对比度的差值的绝对值在第一阈值与第二阈值之间的情况下,判定所述目标屏幕存在透明缺陷。
在本申请的一个实施例中,上述装置中,判断单元840,还用于在确定所述目标屏幕存在透明缺陷的情况下,计算疑似缺陷区域中的缺陷面积;若所述缺陷面积小于面积阈值,则确定所述透明缺陷为误判的其他缺陷。
在本申请的一个实施例中,上述装置中,判断单元840,用于对所述疑似缺陷区域进行自适应二值化处理,得到以第一颜色表征屏幕正常,第二颜色表征屏幕缺陷的二值化图像;将所述二值化图像中呈现第二颜色的像素点的外接矩形的面积作为缺陷面积。
在本申请的一个实施例中,上述装置还包括:检测图像获取单元,用于获取能够包含目标屏幕的原始图像;将所述原始图像转换为灰度图像;从所述灰度图像中检测出与所述目标屏幕对应的有效区域,根据所述有效区域对所述灰度图像进行剪裁,得到所述目标屏幕的检测图像。
在本申请的一个实施例中,上述装置中,疑似缺陷区域确定单元820,用于在所述检测图像中,以疑似缺陷像素点为中心,以第一预设值为边长提取正方形区域作为与该疑似缺陷像素点对应的疑似缺陷区域。
在本申请的一个实施例中,上述装置中,区域划分单元830,用于在包含疑似缺陷像素点的疑似缺陷区域中,以该疑似缺陷像素点为圆心,以第二预设值为半径划分出圆形区域作为与该疑似缺陷像素点对应的核心区域;所述疑似缺陷区域中除所述核心区域以外的全部区域为与该疑似缺陷像素点对应的一般区域。
需要说明的是,上述各装置实施例的具体实施方式可以参照前述对应方法实施例的具体实施方式进行,在此不再赘述。
综上所述,本申请的技术方案,针对透明缺陷较难察觉,且人工进行检测时需要时刻转变检测方位,会增加作业员劳动强度的问题,可以通过疑似缺陷像素点确定、灰度值统计比较、缺陷面积复核的方式,确定目标屏幕是否存在透明缺陷,通过实验证明本申请提出的方案可以很好的应用到产线的OLED等类型屏幕的透明缺陷自动检测,不仅降低了成本,而且减少该操作对工人眼睛的伤害,提高了工作效率,减小了漏检率。
需要说明的是:
在此提供的算法和显示不与任何特定计算机、虚拟装置或者其它设备固有相关。各种通用装置也可以与基于在此的示教一起使用。根据上面的描述,构造这类装置所要求的结构是显而易见的。此外,本申请也不针对任何特定编程 语言。应当明白,可以利用各种编程语言实现在此描述的本申请的内容,并且上面对特定语言所做的描述是为了披露本申请的最佳实施方式。
本领域那些技术人员可以理解,可以对实施例中的设备中的模块进行自适应性地改变并且把它们设置在与该实施例不同的一个或多个设备中。可以把实施例中的模块或单元或组件组合成一个模块或单元或组件,以及此外可以把它们分成多个子模块或子单元或子组件。除了这样的特征和/或过程或者单元中的至少一些是相互排斥之外,可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明确陈述,本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由提供相同、等同或相似目的的替代特征来代替。
此外,本领域的技术人员能够理解,尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征,但是不同实施例的特征的组合意味着处于本申请的范围之内并且形成不同的实施例。例如,在下面的权利要求书中,所要求保护的实施例的任意之一都可以以任意的组合方式来使用。
本申请的各个部件实施例可以以硬件实现,或者以在一个或者多个处理器上运行的软件模块实现,或者以它们的组合实现。本领域的技术人员应当理解,可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本申请实施例的屏幕缺陷检测装置800中的一些或者全部部件的一些或者全部功能。本申请还可以实现为用于执行这里所描述的方法的一部分或者全部的设备或者装置程序(例如,计算机程序和计算机程序产品)。这样的实现本申请的程序可以存储在计算机可读介质上,或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到,或者在载体信号上提供,或者以任何其他形式提供。
例如,图9示出了根据本申请一个实施例的电子设备的结构示意图。该电子设备900包括处理器910和被安排成存储计算机可执行指令(计算机可读程序代码)的存储器920。存储器920可以是诸如闪存、EEPROM(电可擦除可编程只读存储器)、EPROM、硬盘或者ROM之类的电子存储器。存储器920具有存储用于执行上述方法中的任何方法步骤的计算机可读程序代码931的存储空间930。例如,用于存储计算机可读程序代码的存储空间930可以包括分别用于实现上面的方法中的各种步骤的各个计算机可读程序代码931。计算机可读程序 代码931可以从一个或者多个计算机程序产品中读出或者写入到这一个或者多个计算机程序产品中。这些计算机程序产品包括诸如硬盘,紧致盘(CD)、存储卡或者软盘之类的程序代码载体。这样的计算机程序产品通常为例如图10所述的计算机可读存储介质。图10示出了根据本申请一个实施例的一种计算机可读存储介质的结构示意图。该计算机可读存储介质1000存储有用于执行根据本申请的方法步骤的计算机可读程序代码931,可以被电子设备900的处理器910读取,当计算机可读程序代码931由电子设备900运行时,导致该电子设备900执行上面所描述的方法中的各个步骤,具体来说,该计算机可读存储介质存储的计算机可读程序代码931可以执行上述任一实施例中示出的方法。计算机可读程序代码931可以以适当形式进行压缩。
应该注意的是上述实施例对本申请进行说明而不是对本申请进行限制,并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中,不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的元件或步骤。位于元件之前的单词“一”或“一个”不排除存在多个这样的元件。本申请可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中,这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。

Claims (11)

  1. 一种屏幕缺陷检测方法,其中,所述屏幕缺陷检测方法包括:
    从目标屏幕的检测图像中识别出若干个疑似缺陷像素点;
    在所述检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域;
    对所述疑似缺陷区域进行划分得到一般区域和核心区域;
    根据所述疑似缺陷区域的灰度均值、所述一般区域的灰度均值以及所述核心区域的灰度最小值判断所述目标屏幕是否存在透明缺陷。
  2. 根据权利要求1所述的屏幕缺陷检测方法,其中,所述根据所述疑似缺陷区域的灰度均值、所述一般区域的灰度均值以及所述核心区域的灰度最小值判断所述目标屏幕是否存在透明缺陷包括:
    计算所述核心区域的灰度最小值与所述一般区域的灰度均值的比值作为局部对比度;
    计算所述核心区域的灰度最小值与所述疑似缺陷区域的灰度均值的比值作为全局对比度;
    利用所述局部对比度和所述全局对比度判断所述目标屏幕是否存在透明缺陷。
  3. 根据权利要求2所述的屏幕缺陷检测方法,其中,所述利用所述局部对比度和所述全局对比度判断所述目标屏幕是否存在透明缺陷包括:
    若所述局部对比度和所述全局对比度的差值的绝对值在第一阈值与第二阈值之间,则所述目标屏幕存在透明缺陷。
  4. 根据权利要求1或2或3所述的屏幕缺陷检测方法,其中,所述屏幕缺陷检测方法还包括:
    在确定所述目标屏幕存在透明缺陷的情况下,计算疑似缺陷区域中的缺陷面积;
    若所述缺陷面积小于面积阈值,则确定所述透明缺陷为误判的其他缺陷。
  5. 根据权利要求4所述的屏幕缺陷检测方法,其中,所述计算缺陷面积包括:
    对所述疑似缺陷区域进行自适应二值化处理,得到以第一颜色表征屏幕正常,第二颜色表征屏幕缺陷的二值化图像;
    将所述二值化图像中呈现第二颜色的像素点的外接矩形的面积作为缺陷面积。
  6. 根据权利要求1所述的屏幕缺陷检测方法,其中,所述屏幕缺陷检测方法还包括:
    获取能够包含目标屏幕的原始图像;
    将所述原始图像转换为灰度图像;
    从所述灰度图像中检测出与所述目标屏幕对应的有效区域,根据所述有效区域对所述灰度图像进行剪裁,得到所述目标屏幕的检测图像。
  7. 根据权利要求1所述的屏幕缺陷检测方法,其中,所述在所述检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域包括:
    在所述检测图像中,以疑似缺陷像素点为中心,以第一预设值为边长提取正方形区域作为与该疑似缺陷像素点对应的疑似缺陷区域。
  8. 根据权利要求1所述的屏幕缺陷检测方法,其中,所述对所述疑似缺陷区域进行划分得到一般区域和核心区域包括:
    在包含疑似缺陷像素点的疑似缺陷区域中,以该疑似缺陷像素点为圆心,以第二预设值为半径划分出圆形区域作为与该疑似缺陷像素点对应的核心区域;
    所述疑似缺陷区域中除所述核心区域以外的全部区域为与该疑似缺陷像素点对应的一般区域。
  9. 一种屏幕缺陷检测装置,其中,所述屏幕缺陷检测装置包括:
    识别单元810,用于从目标屏幕的检测图像中识别出若干个疑似缺陷像素点;
    疑似缺陷区域确定单元820,用于在检测图像中确定与疑似缺陷像素点对应的疑似缺陷区域;
    划分单元830,用于对疑似缺陷区域进行划分得到一般区域和核心区域;
    判断单元840,用于根据疑似缺陷区域的灰度均值、一般区域的灰度均值以及核心区域的灰度最小值判断目标屏幕是否存在透明缺陷。
  10. 根据权利要求9所述的装置,其中,所述判断单元840,用于计算所述核心区域的灰度最小值与所述一般区域的灰度均值的比值作为局部对比度;计算所述核心区域的灰度最小值与所述疑似缺陷区域的灰度均值的比值作为全局对比度;利用所述局部对比度和所述全局对比度判断所述目标屏幕是否存在透明缺陷。
  11. 一种电子设备,其中,所述电子设备包括:
    处理器;以及
    被安排成存储计算机可执行指令的存储器,所述可执行指令在被执行时使所述处理器执行根据权利要求1-8中任一项所述的屏幕缺陷检测方法。
PCT/CN2021/114453 2020-09-28 2021-08-25 屏幕缺陷检测方法、装置和电子设备 WO2022062812A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/246,806 US20230368361A1 (en) 2020-09-28 2021-08-25 Screen defect detection method, apparatus, and electronic device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011039225.X 2020-09-28
CN202011039225.XA CN112348773B (zh) 2020-09-28 2020-09-28 屏幕缺陷检测方法、装置和电子设备

Publications (1)

Publication Number Publication Date
WO2022062812A1 true WO2022062812A1 (zh) 2022-03-31

Family

ID=74361203

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/114453 WO2022062812A1 (zh) 2020-09-28 2021-08-25 屏幕缺陷检测方法、装置和电子设备

Country Status (3)

Country Link
US (1) US20230368361A1 (zh)
CN (1) CN112348773B (zh)
WO (1) WO2022062812A1 (zh)

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114820617A (zh) * 2022-06-29 2022-07-29 苏州大学 基于四焦距相位相干机器视觉的晶体缺陷检测方法和系统
CN115049604A (zh) * 2022-06-09 2022-09-13 佛山科学技术学院 一种大幅面板材超高分辨率图像的微小缺陷快速检测方法
CN115272350A (zh) * 2022-09-30 2022-11-01 江苏宝翼通讯科技有限公司 一种计算机pcb主板生产质量检测方法
CN115311283A (zh) * 2022-10-12 2022-11-08 山东鲁玻玻璃科技有限公司 一种玻璃拉管缺陷检测方法及系统
CN115439476A (zh) * 2022-11-07 2022-12-06 成都博视广达科技有限责任公司 基于图像分析的丝印缺陷检测方法及装置
CN115457035A (zh) * 2022-11-10 2022-12-09 山东鲁旺机械设备有限公司 基于机器视觉的建筑吊篮焊接质量检测方法
CN115457037A (zh) * 2022-11-10 2022-12-09 澳润(山东)药业有限公司 一种用于阿胶制品生产的原材料质检方法
CN115511907A (zh) * 2022-11-24 2022-12-23 深圳市晶台股份有限公司 一种用于led屏幕的划痕检测方法
CN115601368A (zh) * 2022-12-16 2023-01-13 山东天意高科技有限公司(Cn) 一种建材装备钣金件缺陷检测方法
CN115841488A (zh) * 2023-02-21 2023-03-24 聊城市飓风工业设计有限公司 一种基于计算机视觉的pcb板的检孔方法
CN115862006A (zh) * 2023-03-01 2023-03-28 山东长有面粉有限公司 一种面粉磨粉过程中麸星的检测方法
CN116049863A (zh) * 2023-03-28 2023-05-02 东莞锐视光电科技有限公司 采用lcd显示屏产生条纹光的系统、方法及应用
CN116071355A (zh) * 2023-03-06 2023-05-05 山东第一医科大学第二附属医院 一种用于外周血管影像的辅助分割系统及方法
CN116224632A (zh) * 2022-12-30 2023-06-06 中山市巨正电子有限公司 一种背光源装配精度检测方法及检测系统
CN116416246A (zh) * 2023-06-08 2023-07-11 临沂中科芯华新材料科技有限公司 基于机器视觉的全降解塑料制品淋膜效果评价方法
CN116433669A (zh) * 2023-06-14 2023-07-14 山东兴华钢结构有限公司 基于机器视觉的抗震结构钢架焊缝质量检测方法
CN116612306A (zh) * 2023-07-17 2023-08-18 山东顺发重工有限公司 基于计算机视觉的法兰盘智能对位方法及系统
CN116630332A (zh) * 2023-07-26 2023-08-22 山东华航高分子材料有限公司 一种基于图像处理的pvc塑料管口缺陷检测方法
CN116758057A (zh) * 2023-08-10 2023-09-15 山东贺铭电气有限公司 基于人工智能的通信设备缺陷检测方法
CN116778520A (zh) * 2023-08-11 2023-09-19 山东省大数据中心 一种海量证照数据质量检验方法
CN116843678A (zh) * 2023-08-28 2023-10-03 青岛冠宝林活性炭有限公司 一种硬碳电极生产质量检测方法
CN116977333A (zh) * 2023-09-22 2023-10-31 山东恒海钢结构有限公司 基于图像数据的装配式钢结构构件质量检测方法
CN117115146A (zh) * 2023-10-19 2023-11-24 深圳宝铭微电子有限公司 一种碳化硅材质半导体表面镀膜质量检测方法
CN115829907B (zh) * 2022-04-18 2024-02-23 宁德时代新能源科技股份有限公司 对复合料带的隔膜进行缺陷检测的方法、装置、系统

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112348773B (zh) * 2020-09-28 2022-09-23 歌尔股份有限公司 屏幕缺陷检测方法、装置和电子设备
CN112951737B (zh) * 2021-02-18 2021-12-17 长江存储科技有限责任公司 沟道孔缺陷的改进方法、检测方法及检测系统
CN113808136B (zh) * 2021-11-19 2022-02-22 中导光电设备股份有限公司 一种基于最近邻算法的液晶屏幕缺陷检测方法和系统
CN113888657B (zh) * 2021-12-08 2022-03-29 珠海视熙科技有限公司 一种屏幕颜色采集方法、装置、摄像设备以及存储介质
CN114486732B (zh) * 2021-12-30 2024-04-09 武汉光谷卓越科技股份有限公司 一种基于线扫描三维的瓷砖缺陷在线检测方法
CN114693676B (zh) * 2022-05-31 2022-08-19 南通市通州翔鹏纺织品有限公司 一种新材料纺织品漂白缺陷光学检测方法及装置
CN115183990B (zh) * 2022-09-09 2022-12-23 杭州光粒科技有限公司 一种显示屏幕检测方法及装置
CN116664551B (zh) * 2023-07-21 2023-10-31 深圳市长荣科机电设备有限公司 基于机器视觉的显示屏检测方法、装置、设备及存储介质
CN117314894B (zh) * 2023-11-27 2024-03-29 深圳市金三维实业有限公司 一种手表底盖缺陷快速检测方法
CN117522876A (zh) * 2024-01-08 2024-02-06 成都数智创新精益科技有限公司 一种基于颜色和面积的缺陷检测方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180246044A1 (en) * 2015-10-26 2018-08-30 Huawei Technologies Co., Ltd. Display defect detection method, apparatus, and device for display screen
CN110111711A (zh) * 2019-04-30 2019-08-09 京东方科技集团股份有限公司 屏幕的检测方法及装置、计算机可读存储介质
CN111563889A (zh) * 2020-05-06 2020-08-21 深圳市斑马视觉科技有限公司 基于计算机视觉的液晶屏幕Mura缺陷检测方法
CN111612781A (zh) * 2020-05-27 2020-09-01 歌尔股份有限公司 一种屏幕缺陷检测方法、装置及头戴显示设备
CN111627009A (zh) * 2020-05-27 2020-09-04 歌尔股份有限公司 一种屏幕检测方法、装置及头戴显示设备
CN112348773A (zh) * 2020-09-28 2021-02-09 歌尔股份有限公司 屏幕缺陷检测方法、装置和电子设备

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9230316B2 (en) * 2013-12-31 2016-01-05 Shenzhen China Star Optoelectronics Technology Co., Ltd Defect inspection device for display panel and method for the same
TWI603074B (zh) * 2015-07-03 2017-10-21 元智大學 光學薄膜缺陷辨識方法及其系統
CN108683907B (zh) * 2018-05-31 2020-10-09 歌尔股份有限公司 光学模组像素缺陷检测方法、装置及设备
CN110766736B (zh) * 2019-10-29 2022-10-14 京东方科技集团股份有限公司 缺陷检测方法、装置、电子设备以及存储介质
CN111062910B (zh) * 2019-11-13 2023-06-20 易思维(杭州)科技有限公司 局部阈值分割方法及缺陷检测方法
CN111462113B (zh) * 2020-04-24 2021-12-28 上海精测半导体技术有限公司 无图形晶圆的复检方法
CN111612757B (zh) * 2020-05-18 2023-07-18 苏州精濑光电有限公司 一种屏幕裂纹的检测方法、装置、设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180246044A1 (en) * 2015-10-26 2018-08-30 Huawei Technologies Co., Ltd. Display defect detection method, apparatus, and device for display screen
CN110111711A (zh) * 2019-04-30 2019-08-09 京东方科技集团股份有限公司 屏幕的检测方法及装置、计算机可读存储介质
CN111563889A (zh) * 2020-05-06 2020-08-21 深圳市斑马视觉科技有限公司 基于计算机视觉的液晶屏幕Mura缺陷检测方法
CN111612781A (zh) * 2020-05-27 2020-09-01 歌尔股份有限公司 一种屏幕缺陷检测方法、装置及头戴显示设备
CN111627009A (zh) * 2020-05-27 2020-09-04 歌尔股份有限公司 一种屏幕检测方法、装置及头戴显示设备
CN112348773A (zh) * 2020-09-28 2021-02-09 歌尔股份有限公司 屏幕缺陷检测方法、装置和电子设备

Cited By (37)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115829907B (zh) * 2022-04-18 2024-02-23 宁德时代新能源科技股份有限公司 对复合料带的隔膜进行缺陷检测的方法、装置、系统
CN115049604A (zh) * 2022-06-09 2022-09-13 佛山科学技术学院 一种大幅面板材超高分辨率图像的微小缺陷快速检测方法
CN115049604B (zh) * 2022-06-09 2023-04-07 佛山科学技术学院 一种大幅面板材超高分辨率图像的微小缺陷快速检测方法
CN114820617B (zh) * 2022-06-29 2022-09-20 苏州大学 基于四焦距相位相干机器视觉的晶体缺陷检测方法和系统
CN114820617A (zh) * 2022-06-29 2022-07-29 苏州大学 基于四焦距相位相干机器视觉的晶体缺陷检测方法和系统
CN115272350A (zh) * 2022-09-30 2022-11-01 江苏宝翼通讯科技有限公司 一种计算机pcb主板生产质量检测方法
CN115311283A (zh) * 2022-10-12 2022-11-08 山东鲁玻玻璃科技有限公司 一种玻璃拉管缺陷检测方法及系统
CN115439476B (zh) * 2022-11-07 2023-03-14 成都博视广达科技有限责任公司 基于图像分析的丝印缺陷检测方法及装置
CN115439476A (zh) * 2022-11-07 2022-12-06 成都博视广达科技有限责任公司 基于图像分析的丝印缺陷检测方法及装置
CN115457035A (zh) * 2022-11-10 2022-12-09 山东鲁旺机械设备有限公司 基于机器视觉的建筑吊篮焊接质量检测方法
CN115457037A (zh) * 2022-11-10 2022-12-09 澳润(山东)药业有限公司 一种用于阿胶制品生产的原材料质检方法
CN115511907A (zh) * 2022-11-24 2022-12-23 深圳市晶台股份有限公司 一种用于led屏幕的划痕检测方法
CN115601368A (zh) * 2022-12-16 2023-01-13 山东天意高科技有限公司(Cn) 一种建材装备钣金件缺陷检测方法
CN116224632A (zh) * 2022-12-30 2023-06-06 中山市巨正电子有限公司 一种背光源装配精度检测方法及检测系统
CN115841488B (zh) * 2023-02-21 2023-06-09 聊城市飓风工业设计有限公司 一种基于计算机视觉的pcb板的检孔方法
CN115841488A (zh) * 2023-02-21 2023-03-24 聊城市飓风工业设计有限公司 一种基于计算机视觉的pcb板的检孔方法
CN115862006B (zh) * 2023-03-01 2023-05-16 山东长有面粉有限公司 一种面粉磨粉过程中麸星的检测方法
CN115862006A (zh) * 2023-03-01 2023-03-28 山东长有面粉有限公司 一种面粉磨粉过程中麸星的检测方法
CN116071355A (zh) * 2023-03-06 2023-05-05 山东第一医科大学第二附属医院 一种用于外周血管影像的辅助分割系统及方法
CN116049863A (zh) * 2023-03-28 2023-05-02 东莞锐视光电科技有限公司 采用lcd显示屏产生条纹光的系统、方法及应用
CN116416246B (zh) * 2023-06-08 2023-08-11 临沂中科芯华新材料科技有限公司 基于机器视觉的全降解塑料制品淋膜效果评价方法
CN116416246A (zh) * 2023-06-08 2023-07-11 临沂中科芯华新材料科技有限公司 基于机器视觉的全降解塑料制品淋膜效果评价方法
CN116433669B (zh) * 2023-06-14 2023-08-18 山东兴华钢结构有限公司 基于机器视觉的抗震结构钢架焊缝质量检测方法
CN116433669A (zh) * 2023-06-14 2023-07-14 山东兴华钢结构有限公司 基于机器视觉的抗震结构钢架焊缝质量检测方法
CN116612306B (zh) * 2023-07-17 2023-09-26 山东顺发重工有限公司 基于计算机视觉的法兰盘智能对位方法及系统
CN116612306A (zh) * 2023-07-17 2023-08-18 山东顺发重工有限公司 基于计算机视觉的法兰盘智能对位方法及系统
CN116630332A (zh) * 2023-07-26 2023-08-22 山东华航高分子材料有限公司 一种基于图像处理的pvc塑料管口缺陷检测方法
CN116630332B (zh) * 2023-07-26 2023-09-26 山东华航高分子材料有限公司 一种基于图像处理的pvc塑料管口缺陷检测方法
CN116758057A (zh) * 2023-08-10 2023-09-15 山东贺铭电气有限公司 基于人工智能的通信设备缺陷检测方法
CN116778520B (zh) * 2023-08-11 2023-11-10 山东省大数据中心 一种海量证照数据质量检验方法
CN116778520A (zh) * 2023-08-11 2023-09-19 山东省大数据中心 一种海量证照数据质量检验方法
CN116843678A (zh) * 2023-08-28 2023-10-03 青岛冠宝林活性炭有限公司 一种硬碳电极生产质量检测方法
CN116843678B (zh) * 2023-08-28 2023-11-21 青岛冠宝林活性炭有限公司 一种硬碳电极生产质量检测方法
CN116977333A (zh) * 2023-09-22 2023-10-31 山东恒海钢结构有限公司 基于图像数据的装配式钢结构构件质量检测方法
CN116977333B (zh) * 2023-09-22 2023-12-22 山东恒海钢结构有限公司 基于图像数据的装配式钢结构构件质量检测方法
CN117115146A (zh) * 2023-10-19 2023-11-24 深圳宝铭微电子有限公司 一种碳化硅材质半导体表面镀膜质量检测方法
CN117115146B (zh) * 2023-10-19 2024-03-19 深圳宝铭微电子有限公司 一种碳化硅材质半导体表面镀膜质量检测方法

Also Published As

Publication number Publication date
CN112348773A (zh) 2021-02-09
US20230368361A1 (en) 2023-11-16
CN112348773B (zh) 2022-09-23

Similar Documents

Publication Publication Date Title
WO2022062812A1 (zh) 屏幕缺陷检测方法、装置和电子设备
CN111627009B (zh) 一种屏幕检测方法、装置及头戴显示设备
CN111612781B (zh) 一种屏幕缺陷检测方法、装置及头戴显示设备
CN110070524A (zh) 一种智能终端面板视觉故障检测系统
CN110858395A (zh) 一种卷装长丝的污丝缺陷检测方法
CN112330598A (zh) 一种化纤表面僵丝缺陷检测的方法、装置及存储介质
CN111122590A (zh) 一种陶瓷表面缺陷检测装置及检测方法
WO2021189259A1 (zh) Oled屏幕点缺陷判定方法、装置、存储介质及电子设备
CN115880248A (zh) 一种表面划痕缺陷识别方法及视觉检测设备
CN108107611B (zh) 一种自适应缺陷检测方法、装置及电子设备
KR20070005968A (ko) 용접부 결함 검출 방법 및 장치
CN114897881A (zh) 一种基于边缘特征的晶粒缺陷检测方法
CN109086643B (zh) 一种基于机器视觉的彩盒标签检测方法及系统
CN113793322A (zh) 一种对磁性材料自动检测的方法、电子设备和存储介质
CN111008960B (zh) 基于机器视觉的铝电解电容底部外观检测方法及装置
CN111833350A (zh) 机器视觉检测方法与系统
CN114813783B (zh) 芯片表面瑕疵检测方法
CN114125435A (zh) 一种用于网络摄像机的图像缺陷智能检测方法
CN112200805A (zh) 一种工业品图像目标提取及缺陷判断方法
CN112987356A (zh) 液晶面板底部异物误检滤除方法及装置
CN108171691B (zh) 一种塑料容器的检测方法
JP4036048B2 (ja) 欠陥識別方法
JP2004286708A (ja) 欠陥検出装置、方法及びプログラム
TWI816549B (zh) 自動化瑕疵檢測方法
JP6043178B2 (ja) フラットパネルディスプレイの自動ムラ検出装置および自動ムラ検出方法

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: 21871185

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: 21871185

Country of ref document: EP

Kind code of ref document: A1