CN114858805A - Glass coated surface defect online detection device and defect classification identification method - Google Patents
Glass coated surface defect online detection device and defect classification identification methodDownload PDFInfo
- Publication number
- CN114858805A CN114858805ACN202210306997.8ACN202210306997ACN114858805ACN 114858805 ACN114858805 ACN 114858805ACN 202210306997 ACN202210306997 ACN 202210306997ACN 114858805 ACN114858805 ACN 114858805A
- Authority
- CN
- China
- Prior art keywords
- glass
- module
- line
- light source
- coating surface
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000011521glassSubstances0.000titleclaimsabstractdescription153
- 230000007547defectEffects0.000titleclaimsabstractdescription65
- 238000001514detection methodMethods0.000titleclaimsabstractdescription58
- 238000000034methodMethods0.000titleclaimsabstractdescription19
- 238000003384imaging methodMethods0.000claimsabstractdescription64
- 239000011248coating agentSubstances0.000claimsabstractdescription62
- 238000000576coating methodMethods0.000claimsabstractdescription62
- 238000007781pre-processingMethods0.000claimsdescription40
- 230000033001locomotionEffects0.000claimsdescription21
- 230000002950deficientEffects0.000claimsdescription7
- 230000017525heat dissipationEffects0.000claimsdescription6
- 238000003672processing methodMethods0.000claimsdescription4
- 230000009286beneficial effectEffects0.000abstractdescription2
- 238000010586diagramMethods0.000description6
- 230000003287optical effectEffects0.000description5
- 238000004049embossingMethods0.000description4
- 238000004519manufacturing processMethods0.000description4
- 230000000694effectsEffects0.000description3
- 230000005540biological transmissionEffects0.000description2
- 238000013135deep learningMethods0.000description2
- 238000001914filtrationMethods0.000description2
- 239000000463materialSubstances0.000description2
- 238000002310reflectometryMethods0.000description2
- 230000032258transportEffects0.000description2
- 230000000007visual effectEffects0.000description2
- 238000003705background correctionMethods0.000description1
- 238000004422calculation algorithmMethods0.000description1
- 238000003709image segmentationMethods0.000description1
- 238000010801machine learningMethods0.000description1
- 238000007637random forest analysisMethods0.000description1
- 239000002699waste materialSubstances0.000description1
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
- G01N21/8806—Specially adapted optical and illumination features
- 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
- 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
- G01N2021/8854—Grading and classifying of flaws
Landscapes
- Chemical & Material Sciences (AREA)
- Pathology (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及一种玻璃检测技术结构，尤其涉及一种玻璃镀膜面缺陷在线检测装置，还涉及一种基于所述玻璃镀膜面缺陷在线检测装置的玻璃镀膜面缺陷在线检测分类识别方法。The invention relates to a glass detection technology structure, in particular to an on-line detection device for glass coating surface defects, and a method for online detection, classification and identification of glass coating surface defects based on the online detection device for glass coating surface defects.
背景技术Background technique
随着玻璃的应用越来越广，玻璃的性能要求越来越高，为了满足特殊玻璃的性能要求，镀膜玻璃得到越来越广泛的运用。镀膜玻璃性能指标不仅仅取决于玻璃本身材质的影响，还取决于镀膜质量。玻璃有两个表面，我们把需要镀膜的玻璃表面称为镀膜面，另外一面不需要镀膜的面称为非镀膜面，而镀膜面是否平整没有缺陷，不仅是玻璃本身材质缺陷，又是决定镀膜质量的重要因素，所以镀膜面是否有缺陷成为影响镀膜玻璃质量关键因素，同时检测玻璃镀膜面是否平整有没有缺陷成为关键技术。With the increasing application of glass, the performance requirements of glass are getting higher and higher. In order to meet the performance requirements of special glass, coated glass is more and more widely used. The performance index of coated glass depends not only on the influence of the material of the glass itself, but also on the quality of the coating. Glass has two surfaces. We call the glass surface that needs to be coated as the coated surface, and the other side that does not need coating is called the non-coated surface. Whether the coated surface is flat or not is not only a defect in the material of the glass itself, but also determines the coating. Therefore, whether the coating surface is defective or not has become a key factor affecting the quality of coated glass, and at the same time detecting whether the glass coating surface is flat or not has become a key technology.
以往镀膜面缺陷主要是在镀膜前，通过大量人力观察鉴别缺陷，这样不仅效率低，并且伴有严重的误判。镀膜玻璃镀膜面要求平整，但是非镀膜面却没有这个要求，甚至很多玻璃非镀膜面是有花纹的比如光伏压延玻璃，这样又给镀膜面检测带来了巨大的困难。如果镀膜面缺陷却无法分类识别出来，而大部分的非镀膜面缺陷不会影响镀膜玻璃的质量，如果都归为镀膜缺陷，镀膜玻璃的生产将会非常困难，不仅产能受限，也浪费大量资源。In the past, the defects of the coating surface were mainly observed and identified by a lot of manpower before coating, which is not only inefficient, but also accompanied by serious misjudgments. The coated surface of coated glass is required to be flat, but the non-coated surface does not have this requirement. Even many non-coated surfaces of glass have patterns, such as photovoltaic rolled glass, which brings huge difficulties to the detection of the coated surface. If the coated surface defects cannot be classified and identified, and most of the non-coated surface defects will not affect the quality of the coated glass, if they are all classified as coating defects, the production of the coated glass will be very difficult, not only the production capacity will be limited, but also a lot of waste resource.
现有检测方法检测玻璃缺陷，会检测包括镀膜面的缺陷在内的玻璃所有的缺陷，但是却没有针对镀膜面的缺陷进行分类识别。The existing detection method detects glass defects, and can detect all defects of glass including the defects of the coating surface, but does not classify and identify the defects of the coating surface.
此外，现有的检测方法主要是以软件算法为主，通过普通的光学系统对光伏进行成像，形成有压花的图像，然后对压花图像进行分析，找到压花的一些特性，经过复杂的算法处理，然后从图像中过滤压花，实现对缺陷的检测。In addition, the existing detection methods are mainly based on software algorithms. The photovoltaic is imaged through a common optical system to form an embossed image, and then the embossed image is analyzed to find some characteristics of the embossing. Algorithmic processing then filters the embossing from the image to detect defects.
发明内容SUMMARY OF THE INVENTION
为解决现有技术中的问题，本发明提供一种玻璃镀膜面缺陷在线检测装置，还提供了一种基于所述玻璃镀膜面缺陷在线检测装置的玻璃镀膜面缺陷在线检测分类识别方法，不需要复杂的算法，就能够进行镀膜面的缺陷检测。In order to solve the problems in the prior art, the present invention provides an online detection device for glass coating surface defects, and also provides an online detection, classification and identification method for glass coating surface defects based on the online detection device for glass coating surface defects, which does not require The complex algorithm can detect the defects of the coating surface.
本发明玻璃镀膜面缺陷在线检测装置包括线光源模块、成像模块和与成像模块输出端相连的终端数据处理模块，其中，所述线光源模块发出的光线宽度不大于所述玻璃的厚度，所述线光源模块发出的光线经玻璃反射后形成反射光，所述成像模块能够接收经玻璃上表面镀膜面反射的反射光，下表面的反射光设置在所述成像模块接收范围之外，所述成像模块输出端根据成像模块是否能够接收镀膜面的反射光判断所述玻璃的镀膜面是否存在缺陷。The online detection device for glass coating surface defects of the present invention includes a line light source module, an imaging module and a terminal data processing module connected to the output end of the imaging module, wherein the width of the light emitted by the line light source module is not greater than the thickness of the glass, and the The light emitted by the line light source module is reflected by the glass to form reflected light, the imaging module can receive the reflected light reflected by the coating surface on the upper surface of the glass, and the reflected light on the lower surface is set outside the receiving range of the imaging module. The module output end judges whether the coating surface of the glass has defects according to whether the imaging module can receive the reflected light of the coating surface.
本发明作进一步改进，还包括数据预处理模块，所述数据预处理模块设置在所述成像模块和终端数据处理模块之间，对所述成像模块输出的数据进行预处理。The present invention is further improved, and further includes a data preprocessing module, which is arranged between the imaging module and the terminal data processing module, and preprocesses the data output by the imaging module.
本发明作进一步改进，还包括散热模块，所述散热模块为所述玻璃镀膜面缺陷在线检测装置散热。The present invention is further improved, and further includes a heat dissipation module, and the heat dissipation module dissipates heat for the on-line detection device for defects on the glass coating surface.
本发明作进一步改进，还包括分别与线光源模块和成像模块相连的驱动模块，所述驱动模块用于调整所述线光源模块和成像模块的位置和角度，还包括控制所述驱动模块的运动控制模块。The present invention is further improved, and further includes a driving module respectively connected with the linear light source module and the imaging module, the driving module is used to adjust the position and angle of the linear light source module and the imaging module, and also includes controlling the movement of the driving module. control module.
本发明作进一步改进，所述驱动模块为伺服电机，所述运动控制模块为电机控制器。The present invention is further improved, the drive module is a servo motor, and the motion control module is a motor controller.
本发明作进一步改进，所述线光源为线激光光源，线激光在玻璃上投影发散宽度不超过玻璃厚度的1/3，线激光扩张角度小于60°，线激光与玻璃的入射角大于60°。The present invention is further improved, the line light source is a line laser light source, the projected divergence width of the line laser on the glass does not exceed 1/3 of the thickness of the glass, the expansion angle of the line laser is less than 60°, and the incident angle between the line laser and the glass is greater than 60° .
本发明作进另一种改进，所述线光源为普通光源及设置在普通光源的出光端的准直镜，所述准直镜将所述普通光源发出的发散光转换为直线光束，所述直线光束在玻璃上投影发散宽度不超过玻璃厚度的1/3，线激光扩张角度小于60°，线激光与玻璃的入射角大于60°。The present invention makes another improvement. The line light source is a common light source and a collimating mirror arranged at the light-emitting end of the common light source. The collimating mirror converts the divergent light emitted by the common light source into a straight light beam, and the straight line The projected divergence width of the beam on the glass does not exceed 1/3 of the thickness of the glass, the expansion angle of the line laser is less than 60°, and the incident angle between the line laser and the glass is greater than 60°.
本发明作进一步改进，所述成像模块包括线阵相机和与所述线阵相机相连的长焦镜头，提供高清晰度图像。The present invention is further improved, and the imaging module includes a line scan camera and a telephoto lens connected to the line scan camera to provide high-definition images.
本发明还提供了一种基于所述玻璃镀膜面缺陷在线检测装置的玻璃镀膜面缺陷在线检测分类识别方法，包括如下步骤：The invention also provides an online detection, classification and identification method for glass coating surface defects based on the glass coating surface defect online detection device, comprising the following steps:
S1：根据待检测的玻璃参数信息和厚度信息，调整线光源模块与成像模块，使成像模块能够接收到线光源经玻璃上表面反射的反射光；S1: Adjust the line light source module and the imaging module according to the glass parameter information and thickness information to be detected, so that the imaging module can receive the reflected light from the line light source reflected by the upper surface of the glass;
S2：输送玻璃进入检测区域，开始检测；S2: Convey the glass into the detection area and start the detection;
S3：在玻璃输送过程中，线光源模块发出线光源，成像模块将成像图像输出给终端数据处理模块；S3: During the glass conveying process, the line light source module emits a line light source, and the imaging module outputs the imaging image to the terminal data processing module;
S4：终端数据处理模块对成像模块输出的图像进行缺陷分类并识别。S4: The terminal data processing module performs defect classification and identification on the image output by the imaging module.
本发明作进一步改进，步骤S1中，还包括通过数据预处理模块进行预处理子步骤，所述预处理子步骤的处理方法如下：The present invention is further improved. In step S1, it also includes a sub-step of pre-processing through a data pre-processing module, and the processing method of the sub-step of pre-processing is as follows:
A1：当待检测玻璃进入检测之前，终端数据处理模块发送包含玻璃厚度信息的玻璃参数给数据预处理模块，A1: Before the glass to be tested enters the test, the terminal data processing module sends the glass parameters including the glass thickness information to the data preprocessing module.
A2：数据预处理模块会根据终端数据处理模块与运动控制模块的数据进行对比确认，判断玻璃厚度与当前位置是否匹配，如果是，执行步骤A3，如果否，执行步骤A4；A2: The data preprocessing module will compare and confirm the data of the terminal data processing module and the motion control module to determine whether the glass thickness matches the current position. If so, go to step A3, if not, go to step A4;
A3：如果玻璃厚度与当前位置匹配，数据预处理模块给终端数据模块OK信号，然后执行步骤S2，开始检测玻璃缺陷；A3: If the glass thickness matches the current position, the data preprocessing module sends an OK signal to the terminal data module, and then executes step S2 to start detecting glass defects;
A4：如果玻璃厚度与当前位置不匹配，数据预处理模块根据玻璃厚度计算出运动模块需要移动到位置，然后发送给运动控制模块，控制伺服电机，驱动成像模块、线光源模块移动到指定位置；A4: If the glass thickness does not match the current position, the data preprocessing module calculates the position that the motion module needs to move to according to the glass thickness, and then sends it to the motion control module to control the servo motor to drive the imaging module and the line light source module to move to the specified position;
移动到指定位置后，数据预处理模块开始采集玻璃图像，然后通过玻璃图像信息分析此时成像模块、线光源模块是否在正确位置上，如果在正确位置上，则给终端数据处理模块一个OK信号，然后执行步骤S2，开始检测玻璃；After moving to the designated position, the data preprocessing module starts to collect the glass image, and then analyzes whether the imaging module and the line light source module are in the correct position through the glass image information. If they are in the correct position, an OK signal is given to the terminal data processing module. , and then perform step S2 to start detecting glass;
如果发现此时位置不对，数据预处理模块会根据图像能量与清晰度判断还需要移动的距离，然后发送给运动控制模块进行运动控制，最终达到指定位置。If it is found that the position is wrong at this time, the data preprocessing module will judge the distance that needs to be moved according to the image energy and clarity, and then send it to the motion control module for motion control, and finally reach the specified position.
与现有技术相比，本发明的有益效果是：本发明使用光方向集中的线光源配合成像模块，使得在光学上解决了镀膜面与非镀膜面的区分检测，以及镀膜面玻璃缺陷的检测识别与分类，尤其是在玻璃单面有压花的领域有显著效果，大大缩小了软件工作量，又解决了软件算法容易误检的问题，在效率上有很大提升；同时由于没有下表面反射光的影响，使得能够更加精准测量出玻璃反射率，从而能够更容易的检测镀膜本身缺陷，例如镀膜过厚，镀膜不均匀等。Compared with the prior art, the beneficial effects of the present invention are: the present invention uses a line light source with a concentrated light direction to cooperate with an imaging module, so that the detection of the difference between the coated surface and the non-coated surface and the detection of glass defects on the coated surface are optically solved. Identification and classification, especially in the field with embossing on one side of the glass, have a significant effect, greatly reducing the software workload, and solving the problem that the software algorithm is prone to false detection, greatly improving the efficiency; at the same time, because there is no lower surface The influence of reflected light makes it possible to measure the reflectivity of the glass more accurately, so that it is easier to detect the defects of the coating itself, such as the coating is too thick, the coating is uneven, etc.
附图说明Description of drawings
图1为本发明玻璃镀膜面缺陷在线检测装置一实施例结构示意图；1 is a schematic structural diagram of an embodiment of an on-line detection device for glass coating surface defects of the present invention;
图2为本发明在线检测装置第二实施例结构示意图；2 is a schematic structural diagram of a second embodiment of the online detection device of the present invention;
图3为本发明在线检测装置第三实施例结构示意图；3 is a schematic structural diagram of a third embodiment of the online detection device of the present invention;
图4为本发明玻璃镀膜面无缺陷时光路示意图；4 is a schematic diagram of the optical path of the defect-free glass coating surface of the present invention;
图5为本发明玻璃镀膜面有缺陷时光路示意图；5 is a schematic diagram of the optical path when the glass coating surface of the present invention is defective;
图6和图7为本发明检测底部有缺陷或有花纹的玻璃镀膜面光路示意图；FIG. 6 and FIG. 7 are schematic diagrams of the optical path of the glass coating surface with defects or patterns at the bottom of the detection of the present invention;
图8为本发明处理方法流程图；8 is a flowchart of the processing method of the present invention;
图9为数据预处理模块处理方法流程图。FIG. 9 is a flowchart of a processing method of the data preprocessing module.
具体实施方式Detailed ways
下面结合附图和实施例对本发明做进一步详细说明。The present invention will be described in further detail below with reference to the accompanying drawings and embodiments.
本发明玻璃镀膜面缺陷在线检测装置为专门用来检测镀膜玻璃镀膜面玻璃缺陷，并且基于该装置的镀膜缺陷的检测识别方法，更通用更有效，使得检测范围涵盖不同类型的玻璃，并且检测速度快和检测精度高，误检率非常低。The on-line detection device for glass coating surface defects of the present invention is specially used to detect glass coating surface defects of coated glass, and the detection and identification method of coating defects based on the device is more versatile and effective, so that the detection range covers different types of glass, and the detection speed is Fast and high detection accuracy, the false detection rate is very low.
如图1所示，作为本发明的一个实施例，本发明玻璃镀膜面缺陷在线检测装置包括线光源模块2、成像模块3和与成像模块3输出端相连的终端数据处理模块1，其中，所述线光源模块2发出的光线宽度不大于所述玻璃的厚度，所述线光源模块3发出的光线经玻璃6反射后形成反射光，所述成像模块2能够接收经玻璃上表面镀膜面反射的反射光，下表面的反射光设置在所述成像模块2接收范围之外，所述成像模块2输出端根据成像模块是否能够接收镀膜面的反射光判断所述玻璃6的镀膜面是否存在缺陷。As shown in FIG. 1 , as an embodiment of the present invention, the on-line detection device for glass coating surface defects of the present invention includes a line
本发明通过线光源模块3发出光线经过被检测玻璃反射被成像模块接收形成图像，然后经过终端数据处理模块1分析得出结果。In the present invention, the light emitted by the line light source module 3 is reflected by the detected glass and received by the imaging module to form an image, and then analyzed by the terminal data processing module 1 to obtain a result.
本例线光源模块3中光源为高亮度线激光5，线激光5在玻璃上投影发散宽度不超过玻璃6的厚度，如果玻璃有花纹等，优选的，所述投影发散宽度不超过玻璃厚度的1/3，线激光扩张角度小于60°。线激光与玻璃的入射角大于60°。In this example, the light source in the line light source module 3 is a high-
所述成像模块2包含高速高精度线阵相机4和高清晰度长焦镜头，第一伺服电机2驱动高速线阵相机4上下移动。The
如图2所示，作为本发明的另一个实施例，本例的光源可以使用普通光源8，然后通过准直镜7将所述普通光源8发出的发散光转换为直线光束。普通光源8通过准直镜可以收束光源角度，同时在通过调节准直镜7出光口大小来达到与线激光同样的效果。As shown in FIG. 2 , as another embodiment of the present invention, a common light source 8 can be used as the light source in this example, and then a collimating mirror 7 converts the divergent light emitted by the common light source 8 into a straight beam. The ordinary light source 8 can converge the angle of the light source through the collimating mirror, and at the same time, the same effect as the line laser can be achieved by adjusting the size of the light outlet of the collimating mirror 7.
如图3所示，优选的，本例还包括对成像模块2的位置进行调整的伺服电机12，伺服电机12可以驱动成像模块2上下运动，同样，本例还可以对线光源模块3的位置进行调整的伺服电机12，第二伺服电机12可以驱动线光源模块3上下运动，其中，所述伺服电机12可以由额外的电机控制器或终端数据处理模块1控制，本例的成像模块2的镜头还可以上下摆动，从而能够更好的实现对镀膜面反射光的接收。As shown in FIG. 3 , preferably, this example also includes a
优选地，本例还包括预处理模块10，所述数据预处理模块10设置在所述成像模块2和终端数据处理模块1之间，对所述成像模块2输出的数据进行预处理。所述预处理模块10还分别通过数据传输线缆9与运动控制模块11相连，通过运动控制模块11控制伺服电机12。本例数据预处理模块10根据玻璃厚度控制线光源模块3与成像模块2中的伺服电机12上下移动，保证成像模块2与线光源模块3始终在需要的光学位置上。终端数据处理模块1通过数据传输电缆9与数据预处理模块10进行信息交换与控制。Preferably, this example further includes a
本例还包括散热模块13，本例散热模块13为水冷空调，水冷空调为线光源模块3、成像模块2、数据预处理模块10提供可靠的散热。This example also includes a
本例的数据预处理模块10针对数据的预处理方法包括如下步骤：The data preprocessing method of the
先将原始图像进行高斯滤波，然后将图像进行对比度增强，再对增强后的图像进行平场校正，接着采用基于卡尔曼滤波的图像分割方法，将缺陷图像与背景图像进行分离，提取缺陷图像块，对缺陷块图像进行分析和特征描述，形成缺陷图像数据，然后将其发送给终端数据处理模块1。The original image is first subjected to Gaussian filtering, then the contrast is enhanced, and then the enhanced image is subjected to flat-field correction. Then, the image segmentation method based on Kalman filtering is used to separate the defective image from the background image, and extract the defective image block. , analyze and characterize the defective block image, form defective image data, and then send it to the terminal data processing module 1.
本例终端数据处理模块1会将数据预处理模块10的缺陷数据通过采用传统机器学习方法、深度学习以及人工规则结合的方式对缺陷进行智能分类，并且提供可视化的接口，方便用户设置不同的参数以适应产线的要求。In this example, the terminal data processing module 1 will intelligently classify the defects from the defect data of the
如图4所示，光源5发出的光线14经过玻璃6上表面反射为光线15能被线阵相机4接收，经过下表面反射为光线16，其到达线阵相机4以外的地方，不能被线阵相机4接收到。As shown in FIG. 4 , the light 14 emitted by the
如图5所示，当玻璃6上表面即镀膜面上有缺陷时，光线14经过上表面反射正常为光线19，由于缺陷的存在，发生偏折变为光线17，光线17无法被线阵相机4接收到，此外，经过下表面反射为光线18，其到达线阵相机4以外的地方，同样也不能被线阵相机4接收到。As shown in FIG. 5 , when there is a defect on the upper surface of the glass 6, that is, the coating surface, the light 14 is normally reflected by the upper surface as the light 19. Due to the existence of the defect, the light 17 is deflected and becomes the light 17, and the light 17 cannot be reflected by the line scan camera. 4 received, in addition, reflected by the lower surface as
如图6和图7所示，当玻璃下表面上有缺陷，或者下表面本身就自带花纹的情况下，光线14经过上表面反射为光线15还是能够被线阵相机4接收，经过下表面反射形成的光线16是无法被相机接收的，这与没有带花纹的玻璃检测结果一致，因此，本发明的装置可以用来进行所有类型的玻璃的镀膜面的检测。As shown in FIGS. 6 and 7 , when there is a defect on the lower surface of the glass, or the lower surface itself has a pattern, the light 14 can still be received by the
以往技术无法分辨缺陷是在玻璃镀膜面还是非镀膜面，而本发明使用光方向集中的线激光配合线阵相机，使得在光学上解决了镀膜面与非镀膜面的区分检测，以及镀膜面玻璃缺陷的检测识别与分类，尤其是在玻璃单面有压花的领域有显著效果，大大缩小了软件工作量，又解决了软件算法容易误检的问题，在效率上有很大提升；同时由于没有下表面反射光的影响，使得能够更加精准测量出玻璃反射率，从而能够更容易的检测镀膜本身缺陷，例如镀膜过厚，镀膜不均匀等。The prior art cannot distinguish whether the defect is on the coated surface or the non-coated surface of the glass, but the present invention uses a line laser with a concentrated light direction and a line scan camera, which optically solves the problem of distinguishing and detecting the coated surface and the non-coated surface, as well as the detection of the coated surface glass. The detection, identification and classification of defects, especially in the field with embossing on one side of the glass, has a significant effect, greatly reducing the software workload, and solving the problem that the software algorithm is prone to false detection, greatly improving the efficiency; at the same time, due to the Without the influence of the reflected light from the lower surface, the reflectivity of the glass can be measured more accurately, so that the defects of the coating itself can be more easily detected, such as the coating is too thick, the coating is uneven, etc.
如图8所示，本发明基于上述装置的在先检测分类识别方法的处理过程为：As shown in FIG. 8 , the processing procedure of the prior detection, classification and identification method based on the above-mentioned device of the present invention is:
S1：根据待检测的玻璃参数信息和厚度信息，调整线光源模块与成像模块，使成像模块能够接收到线光源经玻璃上表面反射的反射光。具体地，S1: Adjust the line light source module and the imaging module according to the glass parameter information and thickness information to be detected, so that the imaging module can receive the reflected light reflected by the line light source on the upper surface of the glass. specifically,
当待检测玻璃6快进入检测时，终端数据处理模块1会发送包含玻璃厚度信息的玻璃参数给数据预处理模块10，When the glass 6 to be detected is about to enter the detection, the terminal data processing module 1 will send the glass parameters including the glass thickness information to the
数据预处理模块10会根据终端数据处理模块1与运动控制模块11的数据进行对比确认，如果玻璃厚度与当前位置是匹配的，数据预处理模块10会给终端数据模块1给定OK信号，开始检测玻璃缺陷。The
如果玻璃厚度与当前位置不匹配，数据预处理模块10会根据玻璃厚度计算出运动模块需要移动到位置，然后发送给运动控制模块11，驱动伺服电机12移动成像模块2、线光源模块3到指定位置。If the glass thickness does not match the current position, the
移动到指定位置后，数据预处理模块10会开始采集玻璃图像，然后通过玻璃图像信息分析此时成像模块2、线光源模块3是否在正确位置上。如果在正确位置上，则会给终端数据处理模块1一个OK信号开始检测玻璃。After moving to the designated position, the
如果发现此时位置不对，数据预处理模块10会根据图像能量与清晰度判断还需要移动的距离，在发送给运动控制模块11进行运动控制，最终达到指定位置可以进行缺陷检测。If it is found that the position is wrong at this time, the
S2：输送玻璃进入检测区域，开始检测。S2: Convey the glass into the detection area and start the detection.
本例的玻璃通过现有的玻璃输送装置，将待测试的玻璃输送至检测区域。The glass of this example transports the glass to be tested to the detection area through the existing glass transport device.
S3：在玻璃输送过程中，线光源模块发出线光源，成像模块将成像图像输出给数据预处理模块，数据预处理模块对数据进行处理后，输出给终端数据处理模块；S3: During the glass conveying process, the line light source module emits a line light source, the imaging module outputs the imaging image to the data preprocessing module, and the data preprocessing module processes the data and outputs it to the terminal data processing module;
S4：终端数据处理模块对数据进行分类并识别。S4: The terminal data processing module classifies and identifies the data.
所述终端数据处理模块通过现有的深度学习、随机森林或人工规则等方法进行缺陷的智能化分类，如图9所示。本例还能够将识别结果显示在显示屏上，供用户查看测试结果，本例也提供可视化的接口，方便用户设置不同的参数以适应产线的要求。The terminal data processing module performs intelligent classification of defects through existing methods such as deep learning, random forest or artificial rules, as shown in FIG. 9 . This example can also display the recognition results on the display screen for users to view the test results. This example also provides a visual interface, which is convenient for users to set different parameters to meet the requirements of the production line.
S5：如果测试完一种玻璃，需要测试其他厚度玻璃时，则重新执行步骤S1-S4，对更换后的玻璃进行检测。S5: If one glass is tested and other thicknesses of glass need to be tested, steps S1-S4 are performed again to test the replaced glass.
以上所述之具体实施方式为本发明的较佳实施方式，并非以此限定本发明的具体实施范围，本发明的范围包括并不限于本具体实施方式，凡依照本发明所作的等效变化均在本发明的保护范围内。The specific embodiments described above are the preferred embodiments of the present invention, and are not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to the specific embodiments. All equivalent changes made in accordance with the present invention are within the protection scope of the present invention.
Claims (10)
Translated fromChinesePriority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210306997.8ACN114858805A (en) | 2022-03-25 | 2022-03-25 | Glass coated surface defect online detection device and defect classification identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210306997.8ACN114858805A (en) | 2022-03-25 | 2022-03-25 | Glass coated surface defect online detection device and defect classification identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114858805Atrue CN114858805A (en) | 2022-08-05 |
Family
ID=82628733
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210306997.8APendingCN114858805A (en) | 2022-03-25 | 2022-03-25 | Glass coated surface defect online detection device and defect classification identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114858805A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115342729A (en)* | 2022-08-26 | 2022-11-15 | 湖南科洛德科技有限公司 | Automotive glass size detection device |
| CN117388314A (en)* | 2023-12-13 | 2024-01-12 | 徐州丰诚新材料科技有限公司 | Intelligent detection method and system for molding temperature of optical glass |
| CN118817609A (en)* | 2024-09-18 | 2024-10-22 | 南通新耀玻璃制品有限公司 | A defect detection platform for glass |
| CN119044122A (en)* | 2024-09-03 | 2024-11-29 | 东海县亚连玻璃有限公司 | Be used for cavity glass lower face detection device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH05322796A (en)* | 1992-05-25 | 1993-12-07 | Sony Corp | Paint film inspecting device |
| JP2006078426A (en)* | 2004-09-13 | 2006-03-23 | Konica Minolta Holdings Inc | Apparatus and method for inspecting defect |
| CN102564740A (en)* | 2011-12-28 | 2012-07-11 | 北京奥博泰科技有限公司 | Variable-angle glass reflecting measurement device and method |
| CN103226108A (en)* | 2012-01-27 | 2013-07-31 | 昭和电工株式会社 | Surface inspection method and surface inspection apparatus |
| CN110542541A (en)* | 2019-08-08 | 2019-12-06 | 歌尔股份有限公司 | Method and device for measuring mirror reflectance |
| CN111189803A (en)* | 2019-12-30 | 2020-05-22 | 彩虹显示器件股份有限公司 | Method for detecting micro-particles on working surface of flat glass plate |
- 2022
- 2022-03-25CNCN202210306997.8Apatent/CN114858805A/enactivePending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH05322796A (en)* | 1992-05-25 | 1993-12-07 | Sony Corp | Paint film inspecting device |
| JP2006078426A (en)* | 2004-09-13 | 2006-03-23 | Konica Minolta Holdings Inc | Apparatus and method for inspecting defect |
| CN102564740A (en)* | 2011-12-28 | 2012-07-11 | 北京奥博泰科技有限公司 | Variable-angle glass reflecting measurement device and method |
| CN103226108A (en)* | 2012-01-27 | 2013-07-31 | 昭和电工株式会社 | Surface inspection method and surface inspection apparatus |
| CN110542541A (en)* | 2019-08-08 | 2019-12-06 | 歌尔股份有限公司 | Method and device for measuring mirror reflectance |
| CN111189803A (en)* | 2019-12-30 | 2020-05-22 | 彩虹显示器件股份有限公司 | Method for detecting micro-particles on working surface of flat glass plate |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115342729A (en)* | 2022-08-26 | 2022-11-15 | 湖南科洛德科技有限公司 | Automotive glass size detection device |
| CN117388314A (en)* | 2023-12-13 | 2024-01-12 | 徐州丰诚新材料科技有限公司 | Intelligent detection method and system for molding temperature of optical glass |
| CN117388314B (en)* | 2023-12-13 | 2024-02-27 | 徐州丰诚新材料科技有限公司 | Intelligent detection method and system for molding temperature of optical glass |
| CN119044122A (en)* | 2024-09-03 | 2024-11-29 | 东海县亚连玻璃有限公司 | Be used for cavity glass lower face detection device |
| CN118817609A (en)* | 2024-09-18 | 2024-10-22 | 南通新耀玻璃制品有限公司 | A defect detection platform for glass |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114858805A (en) | Glass coated surface defect online detection device and defect classification identification method | |
| CN110473179B (en) | Method, system and equipment for detecting surface defects of thin film based on deep learning | |
| CN100483116C (en) | Method for detecting 3D defects on surface of belt material | |
| CN102455171B (en) | A method for detecting the geometric shape of the back side of laser tailor welding seam | |
| JP5031691B2 (en) | Surface flaw inspection device | |
| US9594021B2 (en) | Apparatus of detecting transmittance of trench on infrared-transmittable material and method thereof | |
| KR20170129949A (en) | Subject object imaging device, subject sensing method, surface inspection device, and surface inspection method | |
| CN103983426B (en) | The detection of a kind of defect of optical fiber based on machine vision and sorting technique | |
| JPH0850081A (en) | System and method of lens inspection | |
| WO2020051779A1 (en) | Defect detection method and detection system for curved surface object | |
| Yang et al. | An automatic aperture detection system for LED cup based on machine vision | |
| CN119251230B (en) | Welding quality visual inspection method and system applied to welding machine | |
| CN114092411B (en) | An efficient and fast binocular 3D point cloud solder joint defect detection method | |
| CN116664508A (en) | Weld surface quality detection method and computer readable storage medium | |
| CN119495587B (en) | Wafer film defect detection system and method | |
| KR102433319B1 (en) | Vision inspection method of diffusing lens for led light source | |
| US11657495B2 (en) | Non-lambertian surface inspection system using images scanned in different directions | |
| TW201818065A (en) | Machine direction line film inspection | |
| CN111665259A (en) | Testing equipment and testing methods | |
| JP2005189113A (en) | Surface inspection apparatus and surface inspection method | |
| CN118513323A (en) | Laser cleaning online monitoring system based on deep learning and its method and medium | |
| CN217385248U (en) | Glass coating face defect on-line measuring device | |
| Yang et al. | Laser line scanner based real-time geometry monitoring using Encoder-Decoder network during Directed Energy deposition | |
| JP7207443B2 (en) | Surface defect detection device, surface defect detection method, steel plate manufacturing method, steel plate quality control method, and steel plate manufacturing equipment | |
| JPH09105618A (en) | Method and apparatus for inspection of defect on smooth surface of object as well as method and apparatus for measurement of roughness on surface of object |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |