Method and device for detecting defect depth of plate glass
By using 30° angle reflection and transmitting light sources combined with static and dynamic correction processing in flat glass detection, the difficulty of depth measurement caused by the inability to form a clear virtual image during the transmission of flat glass in traditional technology is solved, and efficient and accurate Depth detection.
Patent Information
- Application Number
- CN202411033795.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-08
AI Technical Summary
Traditional automatic optical measurement technology makes it difficult to accurately measure the depth of defects in the process of transmission of flat glass due to slight vibration, optical refractive changes or complex surface morphology, especially when it is impossible to form a clear virtual image, it is difficult to obtain depth information.
Illumination is performed using the reflective light source and the transmittance light source and the flat glass normal at 30°. After static and dynamic correction processing, combined with the image processing algorithm, the spacing between the reflected real and virtual images is calculated to measure the defect depth. The formula Depth = 0.5*H is used. *tan(90-θ2) or Depth=K*L calculates the defect depth based on the refractive angle and the optical image position distance.
It significantly improves the detection ability of defects with low reflectivity on the glass surface and the accuracy of depth position measurement, solves the problem of difficulty in depth measurement caused by transmission jumps, and realizes automated and accurate defect depth detection.
Smart Images

Figure CN118913131A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of glass detection, and in particular to a method and device for detecting the depth of defects in flat glass by combining reflection with light transmission. Background Art
[0002] With the rapid advancement of science and technology and the increasing sophistication of industrial applications, flat glass, as a key material, is indispensable in display manufacturing and the photovoltaic industry. The requirements for glass quality in these industries are far from what they used to be. They pursue not only visual flawlessness, but also functional excellence and performance stability. Therefore, any tiny flaws on the surface of flat glass, such as scratches, bubbles, inclusions, etc., have become key factors that affect product quality, limit production efficiency, and increase maintenance costs.
[0003] Automated optical measurement technology (AOI) is widely used in the detection of glass surface defects due to its non-contact, high efficiency and high degree of automation. This technology captures images of the glass surface through a high-precision camera, and uses image processing algorithms to identify and analyze defect characteristics. However, in the face of increasingly stringent inspection standards, especially the accurate measurement of defect depth, traditional AOI technology seems to be unable to cope with it. Specifically, when defects cannot form a clear and recognizable virtual image due to factors such as tiny vibrations in the glass transmission process, optical refraction changes, or complex surface morphology, the acquisition of depth information becomes a major problem. How to efficiently and accurately detect flat glass defects, especially the depth measurement of defects where flat glass cannot form a virtual image, has become a problem that needs to be solved urgently in the glass manufacturing industry. Therefore, a method and device for detecting the depth of flat glass defects is urgently needed. Summary of the invention
[0004] The object of the present invention is to provide a method and device for detecting the depth of defects in flat glass, so as to overcome the disadvantage of the prior art that when the defect cannot form a virtual image due to reasons such as transmission jump of the flat glass, its depth measurement becomes particularly difficult.
[0005] In a first aspect, the present application provides a method for detecting the depth of a defect in a flat glass, the method comprising:
[0006] The flat glass is illuminated by a reflected light source and a transmitted light source, and the reflected light source and the transmitted light source are both 30° away from the normal line of the flat glass, and the light signals of the reflected real image and the virtual image are obtained. The reflected light source and the transmitted light source are both 30° away from the normal line of the flat glass, which can significantly improve the defect detection capability and defect depth position size of the glass surface with low reflectivity;
[0007] According to the optical signals of the reflected real image and virtual image, static correction and dynamic correction are first performed on them, and then the noise is removed after filtering, and finally amplified and converted into electrical signals;
[0008] Extract key information reflecting defect characteristics from the electrical signal, and generate a defect imaging schematic diagram based on the key information reflecting defect characteristics;
[0009] The defect depth is calculated by measuring the distance between the reflected real image and the virtual image from the defect imaging diagram.
[0010] Specifically, the reflective light source and the camera are located above the flat glass, and the transmissive light source is located below the flat glass.
[0011] Specifically, the static correction is to multiply the grayscale of each pixel of the camera by a coefficient, multiply the position where the light source has strong luminous ability by a coefficient less than 1, and multiply the position where the light source has weak luminous ability by a coefficient greater than 1; the constant coefficient is fixed to form a static correction curve, and the static correction curve is superimposed on the original signal of the digital image to obtain a static corrected image;
[0012] The dynamic correction is that the camera increases the speed of 1gv / scan row by row in the moving direction of the flat glass, and changes the grayscale of a pixel to obtain a dynamically corrected digital image.
[0013] Specifically, the computer processing and filtering of digital image signals includes image segmentation, feature extraction, recognition and classification, and quality control.
[0014] Specifically, the algorithm for calculating the defect depth is:
[0015] When the reflected real image R and virtual image RE are collected:
[0016] Depth=0.5*H*tan(90-θ2) (1)
[0017] Among them, θ2 is the refraction angle, H is the distance between the reflected real image and the virtual image measured simultaneously by the reflection channel; the glass refractive index is 1.5; Depth is the defect depth;
[0018] When the flat glass vibrates and the position of the defect virtual image RE(D) cannot be measured:
[0019] Depth = K*L (2);
[0020] L is the distance between the position of the transmitted image T(D) and the position of the reflected real image R(D), and the constant K is related to the glass thickness and the size of the defect.
[0021] In a second aspect, the present invention further provides a device for detecting the depth of defects in flat glass:
[0022] Light source device: Light source device: Use reflected light source and projected light source to illuminate the flat glass, the reflected light source and projected light source are both 30° away from the normal line of the flat glass, and obtain the light signals of the reflected real image and virtual image;
[0023] Acquisition device: The camera first performs static correction and dynamic correction processing on the optical signals of the reflected real image and virtual image, then filters and removes noise, and finally amplifies and converts them into electrical signals. The camera is set in the detection width direction of the flat glass;
[0024] Output device: Computer, which extracts key information reflecting defect characteristics from the electrical signal and generates defect imaging schematic diagram based on the key information reflecting defect characteristics.
[0025] Specifically, the reflected light source and the transmitted light source use LED white light lamps, and the angle of the light source is adjusted through a mechanical structure. The reflected light source and the transmitted light source are at an angle of 30° to the normal of the flat glass, which can significantly improve the defect detection capability and defect depth position size of the glass surface with low reflectivity.
[0026] Specifically, the camera is equipped with a GPU image processing system for image acquisition, static and dynamic processing, and signal filtering and amplification.
[0027] Specifically, the number of the cameras is ≥1, and the number is determined according to the number of camera pixels, detection defect resolution, and glass detection range.
[0028] Specifically, the flat glass moves to detect the defect location, size and depth, and the moving speed is monitored by an encoder.
[0029] Compared with the prior art, the present invention has the following beneficial technical effects:
[0030] The present invention discloses a method for detecting the depth of defects in flat glass. The method adopts a reflected light source and a transmitted light source to measure defects by imaging at an angle of 30° with the normal line of the flat glass. After a series of image processing, the defect depth value is determined according to a defect depth algorithm. When the glass transmission is unstable, the defect depth can be obtained by multiplying a distance L between a transmitted image position T(D) and a reflected real image position R(D) by a constant value K. This solves the problem in the prior art that it is difficult to measure the depth of a defect when a virtual image cannot be formed due to reasons such as transmission jump of the flat glass.
[0031] The present invention discloses a device for detecting the depth of defects in flat glass. The device tracks the moving speed of the flat glass according to an encoder to control the operation of a light source device, an acquisition device and an output device. The control system of the device can realize automatic control through programming to improve the detection efficiency and accuracy of the device in actual production. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1Schematic diagram of the process of the method for detecting the depth of defects in flat glass according to an embodiment of the present invention.
[0033] Figure 2 Schematic diagram of a light source device in an embodiment of the present invention.
[0034] Figure 3 Schematic diagram of camera image processing flow in an embodiment of the present invention.
[0035] Figure 4 Schematic diagram of a static correction signal in an embodiment of the present invention.
[0036] Figure 5 Schematic diagram of dynamically correcting a background image in an embodiment of the present invention.
[0037] Figure 6 Schematic diagram of outputting defect imaging layered recognition features in an embodiment of the present invention.
[0038] Figure 7 Schematic diagram of defect depth algorithm in an embodiment of the present invention.
[0039] In the picture, 1. Camera; 2. Reflected light source; 3. Transmitted light source; 4. Flat glass. DETAILED DESCRIPTION
[0040] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0041] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0042] like Figures 1 to 6 As shown, a method for detecting the depth of defects in flat glass comprises the following steps:
[0043] The flat glass is illuminated by a reflected light source and a transmitted light source, and the reflected light source and the transmitted light source are both 30° away from the normal line of the flat glass, and the light signals of the reflected real image and the virtual image are obtained. The reflected light source and the transmitted light source are both 30° away from the normal line of the flat glass, which can significantly improve the defect detection capability and defect depth position size of the glass with low surface reflectivity;
[0044] The camera first performs static and dynamic correction processing on the light signals of the reflected real and virtual images, then filters them to remove noise, and finally amplifies them and converts them into electrical signals.
[0045] The computer extracts key information reflecting the defect characteristics from the electrical signal, and generates a defect imaging schematic diagram based on the key information reflecting the defect characteristics;
[0046] The defect depth Depth is calculated by measuring the distance between the reflected real image and the virtual image from the defect imaging diagram.
[0047] The light source includes a reflective light source and a projected light source; the reflective light source and the camera are located above the flat glass, and the transmissive light source is located below the flat glass.
[0048] In some embodiments of the present invention, the camera 1 adopts a high-resolution dual-line TDI line scanning camera with 8K pixels and a scanning frequency of up to 140KHz / s. The camera 1 is equipped with a GPU image processing system.
[0049] Figure 4 : is a schematic diagram of the static correction. The static correction is because the reflected light source and the projected light source are composed of many LED light points, and the brightness of each LED is uneven, which will cause the camera 1 to collect the original digital image signal unevenly; in order to compensate for this physical difference, the camera performs static correction processing on the received original signal, specifically, the grayscale of each pixel of the camera (the median grayscale value of all pixel grayscale values in a certain neighborhood window of the point) is multiplied by a coefficient, the position where the light source has a strong luminous ability is multiplied by a coefficient less than 1, and the position where the light source has a weak luminous ability is multiplied by a coefficient greater than 1; the constant coefficient (within the range of 1±5%) is fixed to form a static correction curve, and the static correction curve is superimposed on the original digital image signal to obtain a static corrected image; the GPU of camera 1 superimposes the brightness correction curve on the original signal waveform to obtain a flat static corrected image signal waveform with grayscale 128 in the figure.
[0050] Figure 5It is a schematic diagram of the background image of dynamic correction. Static correction eliminates the influence of uneven brightness of the light source. In the detection, there will be uneven glass material and unstable movement. At this time, the digital image background obtained only by static correction is still uneven and cannot be used directly for detection. Dynamic correction is required. The dynamic correction is that the camera increases row by row in the moving direction of the flat glass at a speed of 1gv / scan in an arithmetic progression, changes the grayscale of a pixel, and obtains a dynamically corrected digital image; the dynamically corrected digital image is low-pass filtered and enhanced and amplified to obtain a filtered digital image signal.
[0051] After the computer outputs the defect imaging layered recognition feature diagram, under normal circumstances, the distance H between the real image R(D) and the virtual image RE(D) of the reflected R channel can be used to calculate the depth of the defect in the glass thickness direction:
[0052] Due to the glass transmission jump, the position of the defect virtual image RE(D) cannot be measured, but the uncaptured virtual image RE can still be seen in the R channel imaging. According to empirical data, by comparing the distance between the transmission image T(D) and the reflected real image R(D), when the defect is at the bottom of the glass, the distance will be close to 0, when the defect is at the top of the glass, the distance will reach a larger value (depending on the glass thickness and the size of the defect), and when the defect is in the middle of the glass, the distance will have different values;
[0053] When the reflection channel measures the distance H between the real image and the virtual image at the same time, the defect depth is calculated according to the formula:
[0054] Depth=0.5*H*tan(90-θ2) (1);
[0055] Among them, θ2 is the refraction angle, H is the distance between the reflected real image and the virtual image measured simultaneously by the reflection channel, and the glass refractive index is 1.5;
[0056] When the flat glass transmission is unstable and the position of the defect virtual image RE(D) cannot be measured:
[0057] Depth = K*L (2);
[0058] L is the distance between the position of the transmitted image T(D) and the position of the reflected real image R(D), and the constant K is related to the glass thickness and the size of the defect.
[0059] In some embodiments of the present invention, Figure 6 , 7 As shown, the algorithm for calculating defect depth is as follows:
[0060] If the thickness t of the flat glass 4 used is 0.7 mm, the refractive index is 1.5, and the angle θ1 of the reflected light 2 is 30 degrees, the value of the refraction angle θ2 can be calculated according to the reflectivity formula: N1sinθ1=N2sinθ2, θ2=19.47°; theoretically, the maximum distance H(max) between the real image position R(D) and the virtual image position RE(D) of the surface defect on the 0.7 mm thick glass and the depth value of the defect in the thickness direction of the glass can be calculated:
[0061] H(max)=2*0.7mm*tan19.47=0.49;
[0062] Depth=0.5*H*tan(90-θ2)=1.41*H;
[0063] In fact, during the transmission of glass 4, due to transmission jumps, it is impossible to collect the virtual image position data RD_E(D), and it is also impossible to use formula (1) to calculate the defect depth Depth. At this time, the data of R(D) and T(D) are collected synchronously to obtain the distance L between the position of the transmitted image T(D) and the position of the reflected real image R(D). Formula (2) is used for calculation. Based on the calculated data, it can be further determined at which level the defect is located, top, middle, or bottom.
[0064] In another aspect, the present invention provides a flat glass defect layer detection device, such as Figure 7 As shown, the light source device: uses a reflection light source 2 and a projection light source 3 to illuminate the flat glass to obtain light signals of a reflected real image and a virtual image;
[0065] Acquisition device: The camera 1 first performs static correction and dynamic correction processing on the optical signals of the reflected real image and virtual image, then filters and removes noise, and finally amplifies and converts them into electrical signals. The camera 1 is set in the detection width direction of the flat glass 4;
[0066] Output device: A computer, which is used to receive the filtered and amplified digital image signal collected by the camera, perform processing and analysis, and output a defect imaging layered identification feature schematic diagram, which can intuitively display the position, shape and size of the defect, and accurately calculate the actual depth of the defect by measuring the distance between the reflected real image and the transmitted real image according to the algorithm provided by the present invention.
[0067] The reflected light source 2 and the transmitted light source 3 are LED white light lamps, and the angle of the light source is adjusted by a mechanical structure so that the reflected light source 2 and the transmitted light source 3 are 30° with respect to the normal line of the flat glass.
[0068] The camera 1 is equipped with a GPU image processing system for image acquisition, static and dynamic processing, and signal filtering and amplification.
[0069] The number of the cameras 1 is ≥1, and the number of cameras is determined by the number of camera pixels, the detection defect resolution and the glass detection range.
[0070] The flat glass 4 moves to detect the defect location, size and depth, and the moving speed is monitored by an encoder.
[0071] The entire process of the present invention realizes a closed loop from optical imaging to signal conversion and processing, and then to defect imaging and depth calculation, providing an efficient and accurate defect detection and quality control solution for the flat glass industry.
[0072] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. However, the obvious changes or modifications derived therefrom are still within the scope of protection of the present invention.
Claims
1. A method for detecting the depth of defects in flat glass, It is characterized in that The following steps are involved: The flat glass is illuminated by a reflected light source and a transmitted light source, and the reflected light source and the transmitted light source are both 30° away from the normal of the flat glass to obtain the light signals of the reflected real image and the virtual image; According to the optical signals of the reflected real image and virtual image, static correction and dynamic correction are first performed on them, and then the noise is removed after filtering, and finally amplified and converted into electrical signals; Extract key information reflecting defect characteristics from the electrical signal, and generate a defect imaging schematic diagram based on the key information reflecting defect characteristics; The defect depth is calculated by measuring the distance between the reflected real image and the virtual image from the defect imaging diagram.
2. A method for detecting the depth of defects in flat glass according to claim 1, It is characterized in that The reflective light source and the camera are located above the flat glass, and the transmissive light source is located below the flat glass.
3. A method for detecting the depth of defects in flat glass according to claim 1, It is characterized in that The static correction is to multiply the grayscale of each pixel of the camera by a coefficient, multiply the position where the light source has strong luminous ability by a coefficient less than 1, and multiply the position where the light source has weak luminous ability by a coefficient greater than 1; the constant coefficient is fixed to form a static correction curve, and the static correction curve is superimposed on the original signal of the digital image to obtain a static corrected image; The dynamic correction is that the camera increases the speed of 1gv / scan row by row in the moving direction of the flat glass, and changes the grayscale of a pixel to obtain a dynamically corrected digital image.
4. A method for detecting the depth of defects in flat glass according to claim 1, It is characterized in that The extraction of electrical signals reflects the key information of defect characteristics, including image segmentation, feature extraction, recognition and classification, and quality control.
5. A method for detecting the depth of defects in flat glass according to claim 1, It is characterized in that The algorithm for calculating the defect depth is: When the reflected real image R and virtual image RE are collected: Depth=0.5*H*tan(90-θ2) (1); Among them, θ2 is the refraction angle, H is the distance between the reflected real image and the virtual image measured simultaneously by the reflection channel, the glass refractive index is 1.5, and Depth is the defect depth; When the flat glass transmission is unstable and the position of the defect virtual image RE(D) cannot be measured: Depth = K*L (2); L is the distance between the position of the transmitted image T(D) and the position of the reflected real image R(D), and the constant K is related to the glass thickness and the size of the defect.
6. A device for detecting the depth of defects in flat glass according to any one of claims 1 to 5, It is characterized in that include: Light source device: using a reflection light source (2) and a projection light source (3) to illuminate the flat glass, the reflection light source (2) and the projection light source (3) are both at 30° to the normal line of the flat glass, to obtain light signals of a reflected real image and a virtual image; Processing device: The camera (1) first performs static correction and dynamic correction processing on the optical signals of the reflected real image and virtual image, then performs filtering to remove noise, and finally performs amplification processing to convert them into electrical signals. The camera (1) is arranged in the width direction of the flat glass (4); Output device: Computer, which extracts key information reflecting defect characteristics from the electrical signal and generates defect imaging schematic diagram based on the key information reflecting defect characteristics.
7. A device for detecting the depth of defects in flat glass according to claim 6, It is characterized in that The reflection light source (2) and the transmission light source (3) are LED white light lamps, and the angles of the light sources are adjusted by a mechanical structure so that the reflection light source (2) and the transmission light source (3) are 30 degrees with respect to the normal line of the flat glass.
8. A device for detecting the depth of defects in flat glass according to claim 6, It is characterized in that The camera (1) is equipped with a GPU image processing system for image acquisition, static and dynamic processing, and signal filtering and amplification.
9. The device for detecting the depth of defects in flat glass according to claim 6, It is characterized in that The number of the cameras (1) is ≥1.
10. The device for detecting the depth of defects in flat glass according to claim 6, It is characterized in that The flat glass (4) moves to detect the defect location, size and depth, and the moving speed is monitored by an encoder.
