WO2021134996A1 - Led支架缺陷判定方法及系统 - Google Patents

Led支架缺陷判定方法及系统 Download PDF

Info

Publication number
WO2021134996A1
WO2021134996A1 PCT/CN2020/086824 CN2020086824W WO2021134996A1 WO 2021134996 A1 WO2021134996 A1 WO 2021134996A1 CN 2020086824 W CN2020086824 W CN 2020086824W WO 2021134996 A1 WO2021134996 A1 WO 2021134996A1
Authority
WO
WIPO (PCT)
Prior art keywords
tested
image
target
sub
led bracket
Prior art date
Application number
PCT/CN2020/086824
Other languages
English (en)
French (fr)
Inventor
陈润康
林淼
张春平
刘志永
Original Assignee
研祥智能科技股份有限公司
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 研祥智能科技股份有限公司 filed Critical 研祥智能科技股份有限公司
Publication of WO2021134996A1 publication Critical patent/WO2021134996A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/001—Industrial image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
    • H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
    • H01L22/10—Measuring as part of the manufacturing process
    • H01L22/12—Measuring as part of the manufacturing process for structural parameters, e.g. thickness, line width, refractive index, temperature, warp, bond strength, defects, optical inspection, electrical measurement of structural dimensions, metallurgic measurement of diffusions
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection

Definitions

  • the present invention relates to the technical field of image processing, and in particular to a method and system for determining defects of an LED bracket.
  • LED Light Emitting Diode, light-emitting diode
  • the LED bracket is the carrier of the LED lamp beads, that is, the LED bracket is the bottom base of the LED lamp beads before packaging.
  • the chip is fixed in, the positive and negative electrodes are welded, and the LED is sealed with encapsulating glue.
  • the lamp beads are packaged to fix the LED lamp beads on the mounting holes on the LED bracket. Therefore, if the mounting holes on the LED bracket are defective, the LED lamp beads will not function properly.
  • the existing LED bracket generally needs to install a lot of LED lamp beads.
  • most of the manual inspection methods are used to verify each mounting hole on the LED bracket to improve the product qualification rate.
  • disadvantages of manual detection such as high intensity, which is harmful to the body. High-intensity work causes serious vision loss and cannot persist for a long time; low work efficiency, limited detection intensity per unit time, large fluctuations in accuracy for a long time, and high labor costs.
  • the method and system for determining the defect of the LED bracket provided by the present invention can accurately and quickly determine the target to be tested on the LED bracket, and can save labor.
  • the present invention provides a method for judging LED bracket defects, including:
  • the image to be tested is segmented to obtain at least one first sub-image to be tested, wherein each first sub-image to be tested contains at least A target to be tested;
  • the image to be tested corresponding to the LED bracket is segmented according to the gray value of the image to be tested corresponding to the LED bracket, and at least one second to be tested is obtained Sub-images, where each second sub-image to be tested contains at least one target to be tested;
  • the judging whether the target to be tested in the LED bracket is basically normal according to the matching coefficient includes:
  • the matching coefficient is less than the preset first threshold, it is determined that the target to be tested in the LED bracket has a defect
  • the matching coefficient is greater than or equal to the preset first threshold, it is determined that the target to be tested in the LED bracket is basically normal.
  • the size of the template image is smaller than the size of the first sub-object to be measured
  • the matching each first sub-image to be tested by using the preset template image and obtaining the corresponding matching coefficient includes:
  • the preset step size use the preset template image to traverse each first sub-image to be tested, and obtain at least one preliminary matching coefficient on each first sub-image to be tested, and use the preliminary matching coefficient with the largest value as the corresponding The matching coefficient.
  • the method further includes:
  • the determining whether the LED bracket has a defect according to the attribute information of the target to be tested includes:
  • the contrast coefficient is less than the preset second threshold, it is determined that the target to be tested in the LED bracket has a defect
  • the contrast coefficient is greater than or equal to the preset second threshold, it is determined that the target to be tested in the LED bracket is normal.
  • the present invention provides a system for judging LED bracket defects, including:
  • the first acquisition module is configured to acquire an image to be tested of the LED bracket, and the image to be tested contains at least one target to be tested;
  • the first segmentation module is configured to segment the image to be tested according to the position information of the LED bracket of the target to be tested in the image to be tested, and obtain at least one first sub-image to be tested, wherein each of the first sub-images is A sub-image to be tested contains at least one target to be tested;
  • the matching module is configured to use a preset template image to match each first sub-image to be tested, and to obtain a corresponding matching coefficient, wherein the matching coefficient is used to indicate that the corresponding first sub-image to be tested and the corresponding first sub-image to be tested are Describe the degree of similarity of the panel images;
  • a judging module configured to judge whether the target to be tested in the LED bracket is basically normal according to the matching coefficient
  • the second segmentation module is configured to segment the image to be tested corresponding to the LED stent according to the gray value of the image to be tested corresponding to the LED stent when it is determined that the object to be tested in the LED stent is basically normal , And obtain at least one second sub-image to be tested, wherein each second sub-image to be tested contains at least one target to be tested;
  • the first analysis module is configured to perform connected domain analysis on each second sub-image to be tested, and obtain corresponding attribute information of the target to be tested;
  • the determining module is configured to determine whether the LED bracket has a defect according to the attribute information of the target to be tested.
  • the judgment module includes:
  • the first determination sub-module is configured to determine that the target to be tested in the LED bracket has a defect if the matching coefficient is less than a preset first threshold
  • the second determination sub-module is configured to determine that the target to be tested in the LED bracket is basically normal if the matching coefficient is greater than or equal to a preset first threshold.
  • the size of the template image is smaller than the size of the first sub-object to be measured
  • the matching module is further configured to use a preset template image to traverse each first sub-image to be tested according to a preset step size, and to obtain at least one preliminary matching coefficient on each first sub-image to be tested, The preliminary matching coefficient with the largest value is used as the corresponding matching coefficient.
  • system further includes:
  • the second acquisition module is configured to acquire the template image according to the normal LED bracket, and the template image contains at least one normal sample target;
  • the second analysis module is configured to perform connected domain analysis on the template image, and obtain attribute information of the corresponding sample target;
  • the storage module is configured to store the attribute information of the sample target.
  • the determination module includes:
  • the comparison sub-module is configured to compare the attribute information of the target to be tested with the attribute information of the sample target to obtain a contrast coefficient, where the contrast coefficient is used to represent the attribute information of the target to be tested and the attribute information of the sample target;
  • the third determination sub-module is configured to determine that the target to be tested in the LED bracket has a defect if the contrast coefficient is less than a preset second threshold;
  • the fourth determination sub-module is configured to determine that the target to be tested in the LED bracket is normal if the contrast coefficient is greater than or equal to a second preset threshold.
  • the LED bracket defect judgment method and system can segment the target to be tested in the image to be tested according to the position information of the LED bracket in the image to be tested, and use a preset template image Match each first sub-image to be tested, determine whether the target to be tested is basically normal according to the matching result, and then segment the target to be tested in the image to be tested according to the gray value of the image to be tested, And through the method of connected domain analysis, each second sub-image to be tested is analyzed, and according to the analysis result, it is determined whether the basically normal target to be tested has defects, so first according to the LED bracket of the target to be tested in the image to be tested.
  • the determination of location information can avoid the defect of low light sensitivity to the connected domain analysis method, thereby improving the accuracy of the determination and improving the robustness of the system.
  • FIG. 1 is a schematic flowchart of a method for determining defects of an LED bracket according to an embodiment of the application
  • Figure 2 is a schematic structural diagram of the positional relationship between the LED bracket and the target to be tested in the image to be tested according to an embodiment of the application;
  • FIG. 3 is a schematic principle diagram of the principle of the NCC algorithm according to an embodiment of the application.
  • FIG. 4 is a schematic structural diagram of a second image to be tested according to an embodiment of the application.
  • FIG. 5 is an effect diagram of Blob analysis according to an embodiment of the application.
  • FIG. 6 is a schematic flowchart of a method for determining defects of an LED bracket according to an embodiment of the application
  • FIG. 7 is a schematic structural diagram of an LED bracket defect determination system according to an embodiment of the application.
  • the present invention provides a method for determining a defect of an LED bracket.
  • FIG. 1 shows a schematic flowchart of a method for determining a defect of an LED bracket according to an embodiment of the present application.
  • the method includes steps S101 to S107, as follows:
  • Step S101 Obtain an image to be tested of the LED bracket, where the image to be tested includes at least one target to be tested.
  • each LED bracket that is, the image to be tested contains 64 targets to be tested, and the targets to be tested are the ones used to fix the LED lamp beads on the LED bracket. Encapsulation glue.
  • Step S102 According to the position information of the LED bracket of the target to be tested in the image to be tested, the image to be tested is segmented, and at least one first sub-image to be tested is obtained, wherein each first sub-image to be tested Contains at least one target to be tested.
  • FIG. 2 shows a schematic structural diagram of the positional relationship between the LED bracket and the target to be measured in the image to be measured according to an embodiment of the present application, and the target to be measured on each LED bracket They are all arranged on the corresponding LED brackets according to preset rules; according to the preset position information of the LED brackets of the target to be tested in the image to be tested, the image to be tested is divided, so that each first to be tested The sub-image contains a target to be measured.
  • the black part in FIG. 2 is the LED bracket, each square is a first sub-image to be measured, and the white part in each square is the target to be measured.
  • Step S103 Use a preset template image to match each first sub-image to be tested, and obtain a corresponding matching coefficient, where the matching coefficient is used to indicate the corresponding first sub-image to be tested and the panel image The degree of similarity.
  • the NCC (normalized cross correlation) algorithm is used to normalize the find_ncc_model operator in the cross-correlation matching algorithm to obtain the matching coefficient.
  • the size of the template image is smaller than the size of the first sub-object to be measured.
  • Said matching each first sub-image to be tested with a preset template image and obtaining the corresponding matching coefficient includes: traversing each first sub-image to be tested by using the preset template image according to the preset step length Image, and at least one preliminary matching coefficient is obtained on each first sub-image to be tested, and the preliminary matching coefficient with the largest value is used as the corresponding matching coefficient.
  • FIG. 3 shows a schematic diagram of the principle of the NCC algorithm according to an embodiment of the present application.
  • J and K which are the image to be tested and the template image, respectively.
  • the size of K n ⁇ m should be smaller than the size of J N ⁇ M (n ⁇ N, m ⁇ M).
  • the matching method is to move J in the horizontal and vertical directions on K, traverse the entire J, and calculate the NCC coefficient at each position, that is, the preliminary matching coefficient, and the maximum value of the NCC coefficient is the matching coefficient.
  • the movement of K on J can obtain the sub-image Q covered by K, that is, the two overlapping sub-images Q.
  • (X, Y) in FIG. 3 represents the coordinates of the upper left point of the sub-image Q in J.
  • the range of K traversal is 1 ⁇ x ⁇ M-m, 1 ⁇ y ⁇ N-n.
  • Step S104 Determine whether the target to be tested in the LED bracket is basically normal according to the matching coefficient.
  • the judging whether the object to be tested in the LED bracket is basically normal according to the matching coefficient includes: if the matching coefficient is less than a preset first threshold, determining the LED bracket If the matching coefficient is greater than or equal to the preset first threshold, it is determined that the target to be tested in the LED bracket is basically normal.
  • the first threshold is 0.8; if the matching coefficient is less than 0.8, it is determined that the target to be tested in the LED bracket is defective; if the matching coefficient is greater than or equal to 0.8, it is determined that the The target to be tested in the LED bracket is basically normal.
  • Step S105 If it is determined that the object to be tested in the LED bracket is basically normal, segment the image to be tested corresponding to the LED bracket according to the gray value of the image to be tested corresponding to the LED bracket, and obtain at least one first Two sub-images to be tested, wherein each second sub-image to be tested contains at least one target to be tested.
  • FIG. 4 shows a schematic structural diagram of the second to-be-tested image according to an embodiment of the present application.
  • the black part in 4 is the background area, and the other parts are the sample areas.
  • Each second sub-image to be tested contains a target to be tested.
  • Step S106 Perform connected domain analysis on each second sub-image to be tested, and obtain corresponding attribute information of the target to be tested.
  • FIG. 5 shows an effect diagram of Blob analysis according to an embodiment of the present application, where the black part in the white strip in FIG. 5 is the sample area, and the white strip The outer black part is the background area.
  • Step S107 Determine whether the LED bracket has a defect according to the attribute information of the target to be tested.
  • the attribute information of the target to be tested includes geometric features such as the area, length, width, and rectangularity of the target to be tested.
  • the determining whether the LED bracket has a defect according to the attribute information of the target to be tested includes:
  • the attribute information of the target to be tested is compared with the attribute information of the sample target to obtain a contrast coefficient, which is used to represent the attribute information of the target to be tested and the attribute information of the sample target. If the contrast coefficient is less than the preset second threshold, it is determined that the target to be tested in the LED bracket has a defect. If the contrast coefficient is greater than or equal to the preset second threshold, it is determined that the target to be tested in the LED bracket is normal.
  • the attribute information of the sample target includes geometric features such as the area, length, width, and rectangularity of the sample target. Since there are many geometric features of the target to be tested, this application does not limit the second threshold, and can be set according to actual working conditions.
  • the method further includes:
  • the template image is obtained, and the template image contains at least one normal sample target.
  • the connected domain analysis is performed on the template image, and the attribute information of the corresponding sample target is obtained. Store the attribute information of the sample target.
  • the NCC algorithm is also used to obtain the template image
  • the Blob analysis method is used to obtain the attribute information of the sample target. Storing the attribute information of the sample target can facilitate the system to execute step S107 quickly and efficiently.
  • the method for judging the defect of the LED bracket can segment the target to be tested in the image to be tested according to the position information of the LED bracket in the image to be tested, and use a preset template image for each first to be tested.
  • the sub-images are matched, and according to the matching result, it is determined whether the target under test is basically normal, and then according to the gray value of the test image, the target under test in the test image is segmented, and the connected domain analysis method is adopted Analyze each second sub-image to be tested, and determine whether there is a defect in the basically normal target to be tested according to the result of the analysis, so that the first determination based on the position information of the LED bracket of the target to be tested in the image to be tested can avoid connection
  • the domain analysis method is sensitive to the defects of low light sensitivity, which can improve the accuracy of judgment and improve the robustness of the system.
  • the present invention provides a method for judging a defect of an LED bracket. See FIG. 6, which shows a schematic flowchart of a method for judging a defect of an LED bracket according to an embodiment of the present application.
  • the method includes steps S601 to S608, as follows:
  • Step S601 Obtain an image to be tested of the LED bracket, where the image to be tested includes at least one target to be tested.
  • Step S602 According to the position information of the LED bracket of the target to be tested in the image to be tested, the image to be tested is segmented, and at least one first sub-image to be tested is obtained, wherein each first sub-image to be tested Contains a target to be tested.
  • Step S603 According to the preset step size, the NCC algorithm uses the preset template image to traverse each first sub-image to be tested, and at least one preliminary matching coefficient is obtained on each first sub-image to be tested, and the value is the largest The preliminary matching coefficient of is used as the corresponding matching coefficient.
  • Step S604 Determine whether the target to be tested in the LED bracket is basically normal according to the matching coefficient.
  • Step S605 If it is determined that the object to be tested in the LED bracket is basically normal, segment the image to be tested corresponding to the LED bracket according to the gray value of the image to be tested corresponding to the LED bracket, and obtain at least one first Two sub-images to be tested, wherein each second sub-image to be tested includes a target to be tested.
  • Step S606 Use the Blob analysis method to perform connected domain analysis on each second sub-image to be tested, and obtain corresponding attribute information of the target to be tested.
  • Step S607 Compare the attribute information of the target to be tested with the attribute information of the sample target to obtain a contrast coefficient, where the contrast coefficient is used to represent the attribute information of the target to be tested and the attribute information of the sample target.
  • Step S608 Compare the second threshold with the contrast coefficient. If the contrast coefficient is less than the preset second threshold, it is determined that the target to be tested in the LED bracket is defective, and if the contrast coefficient is greater than or equal to the preset The second threshold determines that the target to be tested in the LED bracket is normal.
  • this application adds a normalized cross-correlation matching algorithm to determine the defect of the LED bracket. Compared with the existing recognition method, the recognition accuracy of this application can reach 99.9%. At the same time, the missed detection rate can be Reduced to 0%.
  • the method for judging the defect of the LED bracket can segment the target to be tested in the image to be tested according to the position information of the LED bracket in the image to be tested, and use a preset template image pair through the NCC algorithm.
  • Each first sub-image to be tested is matched, and according to the matching result, it is determined whether the target to be tested is basically normal, and then according to the gray value of the image to be tested, the target to be tested in the image to be tested is segmented, and Analyze each second sub-image to be tested by the method of Blob analysis, and determine whether there is a defect in the basically normal target to be tested according to the result of the analysis, so first according to the position information of the target to be tested in the LED bracket in the image to be tested The determination can avoid the defect of low light sensitivity to the connected domain analysis method, thereby improving the accuracy of the determination, improving the robustness of the system, and reducing the missed detection rate.
  • the present invention provides an LED bracket defect determination system 700.
  • FIG. 7 shows a schematic structural diagram of an LED bracket defect determination system according to an embodiment of the present application.
  • the system includes:
  • the first acquisition module 701 is configured to acquire an image to be tested of the LED bracket, and the image to be tested contains at least one target to be tested.
  • the first segmentation module 702 is configured to segment the image to be tested according to the position information of the LED bracket in the image to be tested, and to obtain at least one first sub-image to be tested, wherein each The first sub-image to be tested contains at least one target to be tested.
  • the matching module 703 is configured to use a preset template image to match each first sub-image to be tested, and obtain a corresponding matching coefficient, where the matching coefficient is used to indicate that the corresponding first sub-image to be tested and The degree of similarity of the panel images.
  • the judging module 704 is configured to judge whether the target under test in the LED bracket is basically normal according to the matching coefficient.
  • the second segmentation module 705 is configured to, when determining that the target to be tested in the LED bracket is basically normal, perform the measurement on the image to be tested corresponding to the LED bracket according to the gray value of the image to be tested corresponding to the LED bracket Divide, and obtain at least one second sub-image to be tested, wherein each second sub-image to be tested contains at least one target to be tested.
  • the first analysis module 706 is configured to perform a connected domain analysis on each second sub-image to be tested, and obtain corresponding attribute information of the target to be tested.
  • the determining module 707 is configured to determine whether the LED bracket has a defect according to the attribute information of the target to be tested.
  • the judgment module 704 includes:
  • the first determination sub-module is configured to determine that the target to be tested in the LED bracket has a defect if the matching coefficient is less than a preset first threshold.
  • the second determination sub-module is configured to determine that the target to be tested in the LED bracket is basically normal if the matching coefficient is greater than or equal to a preset first threshold.
  • the size of the template image is smaller than the size of the first sub-object to be measured.
  • the matching module 703 is further configured to use a preset template image to traverse each first sub-image to be tested according to a preset step size, and obtain at least one preliminary matching coefficient on each first sub-image to be tested , And use the preliminary matching coefficient with the largest value as the corresponding matching coefficient.
  • system further includes:
  • the second acquisition module is configured to acquire the template image according to the normal LED bracket, and the template image contains at least one normal sample target.
  • the second analysis module is configured to perform connected domain analysis on the template image and obtain attribute information of the corresponding sample target.
  • the storage module is configured to store the attribute information of the sample target.
  • the determining module 707 includes:
  • the comparison sub-module is configured to compare the attribute information of the target to be tested with the attribute information of the sample target to obtain a contrast coefficient, and the contrast coefficient is used to represent the attribute information of the target to be tested and the attribute information of the sample target.
  • the third determination sub-module is configured to determine that the target to be tested in the LED bracket has a defect if the contrast coefficient is less than a preset second threshold.
  • the fourth determination sub-module is configured to determine that the target to be tested in the LED bracket is normal if the contrast coefficient is greater than or equal to a second preset threshold.
  • the LED bracket defect determination system can segment the target to be tested in the image to be tested according to the position information of the LED bracket of the target to be tested in the image to be tested, and use a preset template image for each first to be tested
  • the sub-images are matched, and according to the matching result, it is determined whether the target under test is basically normal, and then according to the gray value of the test image, the target under test in the test image is segmented, and the connected domain analysis method is adopted Analyze each second sub-image to be tested, and determine whether there is a defect in the basically normal target to be tested according to the result of the analysis, so that the first determination based on the position information of the LED bracket of the target to be tested in the image to be tested can avoid connection
  • the domain analysis method is sensitive to the defects of low light sensitivity, which can improve the accuracy of judgment and improve the robustness of the system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Signal Processing (AREA)
  • Chemical & Material Sciences (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Power Engineering (AREA)
  • Quality & Reliability (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
  • Image Analysis (AREA)

Abstract

一种LED支架缺陷判定方法及系统，方法包括：获取LED支架的待测图像（S101）；根据待测目标在所述待测图像中LED支架的位置信息，对待测图像进行分割，并得到至少一个第一待测子图像（S102）；采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数（S103）；根据匹配系数判断所述LED支架中的待测目标是否基本正常（S104）；若判定LED支架基本正常，根据LED支架对应的待测图像的灰度值，对LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像（S105）；对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息（S106）；根据待测目标的属性信息判定所述LED支架是否存在缺陷（S107）。该方法和相应系统能够精准快速的对LED支架上的待测目标进行判定，并节省劳动力。

Description

LED支架缺陷判定方法及系统 技术领域
本发明涉及图像处理技术领域，尤其涉及一种LED支架缺陷判定方法及系统。
背景技术
LED(Light Emitting Diode，发光二极管)在现代生活中必不可少，电视机屏幕、手机屏幕、汽车灯等等，都有LED的身影。而LED支架是LED灯珠的载具，即LED支架是LED灯珠在封装之前的底基座，在LED支架的基础上，将芯片固定进去，焊上正负电极，再用封装胶对LED灯珠进行封装，以将LED灯珠固定在LED支架上的安装孔上。因此LED支架上的安装孔出现不良，会使得LED灯珠也不能正常发挥作用。
而现有的LED支架一般需要安装很多个LED灯珠，为提高LED灯珠的安装效率，需要保证安装孔注入的封装胶的形状统一。目前大多是采用人工检测的方法，对LED支架上的每个安装孔进行校验，以提高产品合格率。而人工检测缺点较多，如强度大有害身体，高强度工作使得视力下降严重，无法长时间坚持；工作效能低，单位时间检测强度有限，且长时间准确率波动大，以及人工成本高，人工检验有上限，产能提升需要大量人员等。
发明内容
为解决上述问题，本发明提供的LED支架缺陷判定方法及系统，能够精准快速的对LED支架上的待测目标进行判定，并能够节省劳动力。
第一方面，本发明提供一种LED支架缺陷判定方法，包括：
获取LED支架的待测图像，所述待测图像中包含至少一个待测目标；
根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标；
采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子图像与所述面板图像的相似程度；
根据所述匹配系数判断所述LED支架中的待测目标是否基本正常；
若判定所述LED支架中的待测目标基本正常，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标；
对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息；
根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
可选地，所述根据所述匹配系数判断所述LED支架中的待测目标是否基本正常包括：
若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷；
若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
可选地，所述模板图像的尺寸小于所述第一待测子对象尺寸；
所述采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，包括：
根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每 个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
可选地，所述方法还包括：
根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标；
对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息；
存储所述样本目标的属性信息。
可选地，所述根据所述待测目标的属性信息判定所述LED支架是否存在缺陷，包括：
将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息；
若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷；
若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
第二方面，本发明提供一种LED支架缺陷判定系统，包括：
第一获取模块，被配置为获取LED支架的待测图像，所述待测图像中包含至少一个待测目标；
第一分割模块，被配置为根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标；
匹配模块，被配置为采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子 图像与所述面板图像的相似程度；
判断模块，被配置为根据所述匹配系数判断所述LED支架中的待测目标是否基本正常；
第二分割模块，被配置为在判定所述LED支架中的待测目标基本正常时，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标；
第一分析模块，被配置为对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息；
判定模块，被配置为根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
可选地，所述判断模块包括：
第一判定子模块，被配置为若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷；
第二判定子模块，被配置为若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
可选地，所述模板图像的尺寸小于所述第一待测子对象尺寸；
所述匹配模块，进一步被配置为根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
可选地，所述系统还包括：
第二获取模块，被配置为根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标；
第二分析模块，被配置为对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息；
存储模块，被配置为存储所述样本目标的属性信息。
可选地，所述判定模块包括：
对比子模块，被配置为将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息；
第三判定子模块，被配置为若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷；
第四判定子模块，被配置为若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
本发明实施例提供的LED支架缺陷判定方法及系统，能够根据待测目标在所述待测图像中LED支架的位置信息，对待测图像中的待测目标进行分割，并采用预设的模板图像对每个第一待测子图像进行匹配，根据匹配的结果判定所述待测目标是否基本正常，并在此之后根据待测图像的灰度值，对待测图像中的待测目标进行分割，并通过连通域分析的方法对每个第二待测子图像进行分析，根据分析的结果判定基本正常的待测目标是否存在缺陷，如此先根据待测目标在所述待测图像中LED支架的位置信息进行判定能够避免连通域分析的方法对光敏感程度低的缺陷，进而能够提高判定的准确率，提高系统的鲁棒性。
附图说明
图1为本申请实施例的LED支架缺陷判定方法的示意性流程图；
图2为本申请实施例的在待测图像中LED支架与待测目标位置关系的示 意性结构图；
图3为本申请实施例的NCC算法原理的示意性原理图；
图4为本申请实施例的第二待测图像的示意性结构图；
图5为本申请实施例的Blob分析的效果图；
图6为本申请实施例的LED支架缺陷判定方法的示意性流程图；
图7为本申请实施例的LED支架缺陷判定系统的示意性结构图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
第一方面，本发明提供一种LED支架缺陷判定方法，参见图1，图1示出了根据本申请一实施例的LED支架缺陷判定方法的示意性流程图，所述方法包括步骤S101至步骤S107，如下：
步骤S101：获取LED支架的待测图像，所述待测图像中包含至少一个待测目标。
在本实施例中，每个LED支架上排列有64个待测目标，即所述待测图像中包含64个待测目标，且所述待测目标为LED支架上用于固定LED灯珠的封装胶。
步骤S102：根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标。
在本实施例中，参见图2，图2示出了根据本申请一实施例的在待测图像中LED支架与待测目标位置关系的示意性结构图，每个LED支架上的待测目标都是根据预设的规则排布在相应的LED支架上的；根据预设的待测目标在所述待测图像中LED支架的位置信息，对待测图像进行分割，使得每个第一待测子图像中包含一个待测目标。其中，图2中的黑色部分为LED支架，每个方格即为一个第一待测子图像，且每个方格中的白色部分即为待测目标。
步骤S103：采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子图像与所述面板图像的相似程度。
在本实施例中，通过NCC(normalized cross correlation)算法，归一化互相关匹配算法中的find_ncc_model算子以得到匹配系数。
在一种可选的实施例中，所述模板图像的尺寸小于所述第一待测子对象尺寸。所述采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，包括：根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
具体的，参见图3，图3示出了根据本申请一实施例的NCC算法原理的示意性原理图，例如，存在两幅图像J和K，分别为待测图像和模板图像，K的尺寸n×m应小于J的尺寸N×M(n≤N,m≤M)。匹配方法是，将J在K上进行水平和垂直方向移动，遍历整个J，计算每个位置上的NCC系数，即预备匹配系数，NCC系数最大值即为匹配系数。其中，K在J上移动可以得到被K覆盖、即二者重叠的子图像Q，图3中的(X,Y)表示子图像Q左上点在J中的坐标。K遍历的范围1≤x≤M-m，1≤y≤N-n。
步骤S104：根据所述匹配系数判断所述LED支架中的待测目标是否基本正常。
在一种可选的实施例中，所述根据所述匹配系数判断所述LED支架中的待测目标是否基本正常包括：若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷；若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
在本实施例中，所述第一阈值为0.8；若所述匹配系数小于0.8，则判定所述LED支架中的待测目标存在缺陷；若所述匹配系数大于或等于0.8，则判定所述LED支架中的待测目标基本正常。
步骤S105：若判定所述LED支架中的待测目标基本正常，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标。
在本实施例中，在该步骤中共得到64个第二待测子图像，其中，参见图4，图4示出了根据本申请一实施例的第二待测图像的示意性结构图，图4中的黑色部分为背景区域，其他部分为样本区域。每个第二待测子图像中包含一个待测目标。
步骤S106：对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息。
在本实施例中，参见图5，图5示出了根据本申请一实施例的Blob分析的效果图，其中，图5中的白色长条形内的黑色部分为样本区域，白色长条形外的黑色部分为背景区域。通过Blob分析的方法对每个第二待测子图像进行连通域分析，具体的，通过Blob分析方法中的快速阈值全局分割算子将待测 图像中灰度值g满足MinGray<＝g<＝MaxGray的区域聚合为一个样本区域，并分离样品区域与背景区域，其中，MinGray为128，MaxGray为255。之后采用Connection算子进行连通域的分析，将待测图像中不连通的噪点与样本区域分离，以得到纯净的样本区域，该样本区域即为待测目标在第二待测子图像中的图像。
步骤S107：根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
在本实施例中，所述待测目标的属性信息包括：待测目标的面积、长度、宽度、矩形度等几何特征。
在一种可选的实施例中，所述根据所述待测目标的属性信息判定所述LED支架是否存在缺陷，包括：
将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息。若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷。若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
在本实施例中，所述样本目标的属性信息包括：样本目标的面积、长度、宽度、矩形度等几何特征。由于待测目标的几何特征较多，本申请不对第二阈值进行限定，可根据实际工况进行设置。
在一种可选的实施例中，所述方法还包括：
根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标。对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息。存储所述样本目标的属性信息。
在本实施例中，同样采用NCC算法获得模板图像，并通过Blob分析的方法获得样本目标的属性信息。存储所述样本目标的属性信息能够便于系统快捷高效的执行步骤S107。
所述LED支架缺陷判定方法能够根据待测目标在所述待测图像中LED支架的位置信息，对待测图像中的待测目标进行分割，并采用预设的模板图像对每个第一待测子图像进行匹配，根据匹配的结果判定所述待测目标是否基本正常，并在此之后根据待测图像的灰度值，对待测图像中的待测目标进行分割，并通过连通域分析的方法对每个第二待测子图像进行分析，根据分析的结果判定基本正常的待测目标是否存在缺陷，如此先根据待测目标在所述待测图像中LED支架的位置信息进行判定能够避免连通域分析的方法对光敏感程度低的缺陷，进而能够提高判定的准确率，提高系统的鲁棒性。
第二方面，本发明提供一种LED支架缺陷判定方法，参见图6，图6示出了根据本申请一实施例的LED支架缺陷判定方法的示意性流程图，所述方法包括步骤S601至步骤S608，如下：
步骤S601：获取LED支架的待测图像，所述待测图像中包含至少一个待测目标。
步骤S602：根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含一个待测目标。
步骤S603：根据预设的步长，通过NCC算法使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
步骤S604：根据所述匹配系数判断所述LED支架中的待测目标是否基本 正常。
步骤S605：若判定所述LED支架中的待测目标基本正常，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含一个待测目标。
步骤S606：采用Blob分析的方法对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息。
步骤S607：将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息。
步骤S608：对比第二阈值与所述对比系数，若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷，若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
本申请通过在Blob分析的基础上，增加归一化互相关匹配算法对LED支架缺陷的判定，相对于现有的识别方法，本申请的识别准确率能够达到99.9％，同时，漏检率能够减少到了0％。具体的，所述LED支架缺陷判定方法能够根据待测目标在所述待测图像中LED支架的位置信息，对待测图像中的待测目标进行分割，并通过NCC算法采用预设的模板图像对每个第一待测子图像进行匹配，根据匹配的结果判定所述待测目标是否基本正常，并在此之后根据待测图像的灰度值，对待测图像中的待测目标进行分割，并通过Blob分析的方法对每个第二待测子图像进行分析，根据分析的结果判定基本正常的待测目标是否存在缺陷，如此先根据待测目标在所述待测图像中LED支架的位置信息进行判定能够避免连通域分析的方法对光敏感程度低的缺陷，进而能够提高判 定的准确率，提高系统的鲁棒性，减少漏检率。
第三方面，本发明提供一种LED支架缺陷判定系统700，参见图7，图7示出了根据本申请一实施例的LED支架缺陷判定系统的示意性结构图，所述系统包括：
第一获取模块701，被配置为获取LED支架的待测图像，所述待测图像中包含至少一个待测目标。
第一分割模块702，被配置为根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标。
匹配模块703，被配置为采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子图像与所述面板图像的相似程度。
判断模块704，被配置为根据所述匹配系数判断所述LED支架中的待测目标是否基本正常。
第二分割模块705，被配置为在判定所述LED支架中的待测目标基本正常时，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标。
第一分析模块706，被配置为对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息。
判定模块707，被配置为根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
在一种可选的实施例中，所述判断模块704包括：
第一判定子模块，被配置为若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷。
第二判定子模块，被配置为若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
在一种可选的实施例中，所述模板图像的尺寸小于所述第一待测子对象尺寸。
所述匹配模块703，进一步被配置为根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
在一种可选的实施例中，所述系统还包括：
第二获取模块，被配置为根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标。
第二分析模块，被配置为对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息。
存储模块，被配置为存储所述样本目标的属性信息。
在一种可选的实施例中，所述判定模块707包括：
对比子模块，被配置为将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息。
第三判定子模块，被配置为若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷。
第四判定子模块，被配置为若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
所述LED支架缺陷判定系统能够根据待测目标在所述待测图像中LED支架的位置信息，对待测图像中的待测目标进行分割，并采用预设的模板图像对每个第一待测子图像进行匹配，根据匹配的结果判定所述待测目标是否基本正常，并在此之后根据待测图像的灰度值，对待测图像中的待测目标进行分割，并通过连通域分析的方法对每个第二待测子图像进行分析，根据分析的结果判定基本正常的待测目标是否存在缺陷，如此先根据待测目标在所述待测图像中LED支架的位置信息进行判定能够避免连通域分析的方法对光敏感程度低的缺陷，进而能够提高判定的准确率，提高系统的鲁棒性。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到的变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应该以权利要求的保护范围为准。

Claims (10)

  1. 一种LED支架缺陷判定方法，其特征在于，包括：
    获取LED支架的待测图像，所述待测图像中包含至少一个待测目标；
    根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标；
    采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子图像与所述面板图像的相似程度；
    根据所述匹配系数判断所述LED支架中的待测目标是否基本正常；
    若判定所述LED支架中的待测目标基本正常，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标；
    对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息；
    根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述匹配系数判断所述LED支架中的待测目标是否基本正常包括：
    若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷；
    若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
  3. 根据权利要求1或2所述的方法，其特征在于，所述模板图像的尺寸小于所述第一待测子对象尺寸；
    所述采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，包括：
    根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
  4. 根据权利要求3所述的方法，其特征在于，所述方法还包括：
    根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标；
    对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息；
    存储所述样本目标的属性信息。
  5. 根据权利要求4所述的方法，其特征在于，所述根据所述待测目标的属性信息判定所述LED支架是否存在缺陷，包括：
    将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息；
    若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷；
    若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
  6. 一种LED支架缺陷判定系统，其特征在于，包括：
    第一获取模块，被配置为获取LED支架的待测图像，所述待测图像中包含至少一个待测目标；
    第一分割模块，被配置为根据待测目标在所述待测图像中LED支架的位置信息，对所述待测图像进行分割，并得到至少一个第一待测子图像，其中，每个第一待测子图像中包含至少一个待测目标；
    匹配模块，被配置为采用预设的模板图像对每个第一待测子图像进行匹配，并得到相应的匹配系数，其中，所述匹配系数用于表示相应的第一待测子图像与所述面板图像的相似程度；
    判断模块，被配置为根据所述匹配系数判断所述LED支架中的待测目标是否基本正常；
    第二分割模块，被配置为在判定所述LED支架中的待测目标基本正常时，根据所述LED支架对应的待测图像的灰度值，对所述LED支架对应的待测图像进行分割，并得到至少一个第二待测子图像，其中，每个第二待测子图像中包含至少一个待测目标；
    第一分析模块，被配置为对每个第二待测子图像进行连通域分析，并得到相应的待测目标的属性信息；
    判定模块，被配置为根据所述待测目标的属性信息判定所述LED支架是否存在缺陷。
  7. 根据权利要求6所述的系统，其特征在于，所述判断模块包括：
    第一判定子模块，被配置为若所述匹配系数小于预设的第一阈值，判定所述LED支架中的待测目标存在缺陷；
    第二判定子模块，被配置为若所述匹配系数大于或等于预设的第一阈值，判定所述LED支架中的待测目标基本正常。
  8. 根据权利要求6或7所述的系统，其特征在于，所述模板图像的尺寸小于所述第一待测子对象尺寸；
    所述匹配模块，进一步被配置为根据预设的步长，使用预设的模板图像遍历每个第一待测子图像，并在每个第一待测子图像上得到至少一预备匹配系数，将数值最大的预备匹配系数作为相应的匹配系数。
  9. 根据权利要求8所述的系统，其特征在于，所述系统还包括：
    第二获取模块，被配置为根据正常的LED支架，获取所述模板图像，所述模板图像中包含至少一正常的样本目标；
    第二分析模块，被配置为对所述模板图像进行连通域分析，并得到相应的样本目标的属性信息；
    存储模块，被配置为存储所述样本目标的属性信息。
  10. 根据权利要求9所述的系统，其特征在于，所述判定模块包括：
    对比子模块，被配置为将待测目标的属性信息与样本目标的属性信息进行对比，得到对比系数，所述对比系数用于表示待测目标的属性信息与样本目标的属性信息；
    第三判定子模块，被配置为若所述对比系数小于预设的第二阈值，判定所述LED支架中的待测目标存在缺陷；
    第四判定子模块，被配置为若所述对比系数大于或等于预设的第二阈值，判定所述LED支架中的待测目标正常。
PCT/CN2020/086824 2019-12-30 2020-04-24 Led支架缺陷判定方法及系统 WO2021134996A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911395306.0A CN111060519A (zh) 2019-12-30 2019-12-30 Led支架缺陷判定方法及系统
CN201911395306.0 2019-12-30

Publications (1)

Publication Number Publication Date
WO2021134996A1 true WO2021134996A1 (zh) 2021-07-08

Family

ID=70304673

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/086824 WO2021134996A1 (zh) 2019-12-30 2020-04-24 Led支架缺陷判定方法及系统

Country Status (2)

Country Link
CN (1) CN111060519A (zh)
WO (1) WO2021134996A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI788002B (zh) * 2021-09-09 2022-12-21 英業達股份有限公司 用於檢測電路板上發光二極體之檢測系統

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030179294A1 (en) * 2002-03-22 2003-09-25 Martins Fernando C.M. Method for simultaneous visual tracking of multiple bodies in a closed structured environment
CN101995223A (zh) * 2009-08-25 2011-03-30 比亚迪股份有限公司 一种芯片外观检测方法及系统
CN103076345A (zh) * 2013-01-16 2013-05-01 东莞市三姆森光电科技有限公司 一种ito导电玻璃的检测方法及其全自动光学检测系统
EP1427212B1 (en) * 2002-11-27 2014-07-16 Bosch Security Systems, Inc. Video tracking system and method
CN104700085A (zh) * 2015-03-10 2015-06-10 华中科技大学 一种基于模板匹配的芯片定位方法
JP2015177510A (ja) * 2014-03-18 2015-10-05 株式会社リコー カメラシステム、画像処理方法及びプログラム
CN107543828A (zh) * 2017-08-25 2018-01-05 广东工业大学 一种工件表面缺陷检测方法及系统
CN109829914A (zh) * 2019-02-26 2019-05-31 视睿（杭州）信息科技有限公司 检测产品缺陷的方法和装置
CN109839385A (zh) * 2019-03-04 2019-06-04 佛山市南海区广工大数控装备协同创新研究院 一种自适应的pcb板缺陷视觉定位检测及分类系统
CN109884070A (zh) * 2019-03-08 2019-06-14 武汉大学 一种基于机器视觉的铝铝泡罩包装药片缺陷检测方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9494567B2 (en) * 2012-12-31 2016-11-15 Omni Medsci, Inc. Near-infrared lasers for non-invasive monitoring of glucose, ketones, HBA1C, and other blood constituents
CN106056597B (zh) * 2016-05-26 2019-07-09 广州视源电子科技股份有限公司 物体视觉检测方法及装置
CN110555829B (zh) * 2019-08-12 2021-07-16 华南理工大学 一种用于芯片点胶情况的检测方法

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030179294A1 (en) * 2002-03-22 2003-09-25 Martins Fernando C.M. Method for simultaneous visual tracking of multiple bodies in a closed structured environment
EP1427212B1 (en) * 2002-11-27 2014-07-16 Bosch Security Systems, Inc. Video tracking system and method
CN101995223A (zh) * 2009-08-25 2011-03-30 比亚迪股份有限公司 一种芯片外观检测方法及系统
CN103076345A (zh) * 2013-01-16 2013-05-01 东莞市三姆森光电科技有限公司 一种ito导电玻璃的检测方法及其全自动光学检测系统
JP2015177510A (ja) * 2014-03-18 2015-10-05 株式会社リコー カメラシステム、画像処理方法及びプログラム
CN104700085A (zh) * 2015-03-10 2015-06-10 华中科技大学 一种基于模板匹配的芯片定位方法
CN107543828A (zh) * 2017-08-25 2018-01-05 广东工业大学 一种工件表面缺陷检测方法及系统
CN109829914A (zh) * 2019-02-26 2019-05-31 视睿（杭州）信息科技有限公司 检测产品缺陷的方法和装置
CN109839385A (zh) * 2019-03-04 2019-06-04 佛山市南海区广工大数控装备协同创新研究院 一种自适应的pcb板缺陷视觉定位检测及分类系统
CN109884070A (zh) * 2019-03-08 2019-06-14 武汉大学 一种基于机器视觉的铝铝泡罩包装药片缺陷检测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WEN ZHAOLIN, CHEN XIN, ZHENG DE-TAO: "METHOD OF FAST NORMALIZATION AND CROSS CORREL ATION IN DEFECTION DETECTION", JOURNAL OF GUANGZHOU MARITIME COLLEGE, vol. 14, no. 2, 1 December 2006 (2006-12-01), pages 29 - 40, XP055826754, ISSN: 1009-8526 *

Also Published As

Publication number Publication date
CN111060519A (zh) 2020-04-24

Similar Documents

Publication Publication Date Title
CN101995223B (zh) 一种芯片外观检测方法及系统
CN103500459B (zh) 一种基于规则化参数的印刷电路板缺陷检测方法
CN107945184B (zh) 一种基于彩色图像分割和梯度投影定位的贴装元件检测方法
CN113221889B (zh) 一种芯片字符抗干扰识别方法及装置
WO2021238300A1 (zh) 一种用于测量鞋子刷胶边缘胶线的装置及其测量方法
WO2022082904A1 (zh) 镜头脏污检测方法、装置和设备
WO2007062563A1 (fr) Procede d&#39;examen automatique en ligne pour des detections de defauts de surface d&#39;acier pendant le pretraitement de toles d&#39;acier
CN102053093A (zh) 一种晶圆表面切割芯片的表面缺陷检测方法
KR102168724B1 (ko) 이미지 검사를 이용한 이상 판별 방법 및 장치
CN106501272B (zh) 机器视觉焊锡定位检测系统
WO2017071406A1 (zh) 金针类元件的引脚检测方法和系统
CN109788281A (zh) 一种摄像模组的漏光检测方法及装置
WO2021134996A1 (zh) Led支架缺陷判定方法及系统
CN106442560A (zh) 一种显示屏的定位测量与缺陷检测方法
WO2024002187A1 (zh) 缺陷检测方法、缺陷检测设备及存储介质
CN115601359A (zh) 一种焊缝检测方法及装置
CN113504239A (zh) 一种品质管控数据分析方法
CN110428764B (zh) 显示面板检测方法
CN106226270A (zh) 检测图像传感器表面脏污缺陷的方法
CN102890089A (zh) 晶圆缺陷扫描方法及晶圆缺陷扫描机台
WO2022222467A1 (zh) 开口圆环工件外观缺陷检测方法、系统及计算机存储介质
CN109387527A (zh) 一种膜上膜内缺陷的检测方法
TWI558999B (zh) 瑕疵檢測方法及其裝置
CN109387525A (zh) 一种膜上膜内缺陷的判定方法
WO2005008223A3 (en) Methods for defect detection and process monitoring based on sem images

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: 20909773

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: 20909773

Country of ref document: EP

Kind code of ref document: A1