Classification system, method, apparatus and electronic device for glass internal defects

Abstract

The present application relates to the technical field of glass manufacturing, and discloses a glass internal defect classification system, method, device and electronic equipment, the system comprising: a light source module comprising a light transmission zone and a light emitting zone, the light emitting zone being configured to emit light to a glass to be measured; a backlight module, a diffuse reflection surface of the backlight module being close to the glass to be measured, and being configured to diffuse the light emitted by the light emitting zone and passing through the glass to be measured; the glass to be measured being arranged between the light source module and the backlight module, a first gap being formed between the glass to be measured and the light source module, and a second gap being formed between the glass to be measured and the backlight module; a camera module being arranged corresponding to the light transmission zone of the light source module, and being configured to collect an image of the glass to be measured through the light transmission zone; and a processing module being connected to the camera module, and being configured to classify defects in the glass to be measured based on the collected image.

Description

Technical Field

[0001] This invention relates to the field of glass manufacturing technology, and more specifically to a classification system, method, apparatus, and electronic equipment for internal defects in glass. Background Technology

[0002] During the production of float glass, various point defects inevitably occur inside the glass. These defects mainly include bubbles, inclusions, sodium sulfate bubbles, and nickel sulfide particles. Related technologies often rely on manual re-inspection or destructive testing to determine the types of defects in the glass, but this method is inefficient and cannot meet the needs of industrial-scale online full inspection. Summary of the Invention

[0003] This invention provides a classification system, method, apparatus, and electronic device for internal defects in glass, to solve the problem of low efficiency in determining the types of defects in glass through manual re-inspection or destructive testing in related technologies.

[0004] In a first aspect, the present invention provides a classification system for internal defects in glass, comprising: a light source module including a light-transmitting area and a light-emitting area, wherein the light-emitting area is used to emit light toward the glass under test; a backlight module, wherein the diffuse reflective surface of the backlight module is close to the glass under test, and is used to diffusely reflect the light emitted by the light-emitting area through the glass under test; the glass under test is disposed between the light source module and the backlight module, wherein a first gap is formed between the glass under test and the light source module, and a second gap is formed between the glass under test and the backlight module; a camera module, corresponding to the light-transmitting area of ​​the light source module, for acquiring an image of the glass under test through the light-transmitting area; and a processing module, connected to the camera module, for classifying defects contained in the glass under test based on the acquired image.

[0005] In one optional embodiment, the diffuse reflective surface of the backlight module is a light-transmitting surface, and the backlight module further includes a light-emitting unit and a light-diffusing unit disposed on the back side of the diffuse reflective surface. The light-emitting unit is used to emit light, and the light-diffusing unit is used to disperse the light emitted by the light-emitting unit.

[0006] In one alternative embodiment, the system further includes a motion positioning module for moving the camera module.

[0007] According to an embodiment of the present invention, an image of the target in the glass under test is acquired by a camera module, and then handed over to a processing module to determine the type of defect in the glass under test. This allows for rapid and non-destructive determination of the defect type. Furthermore, by using a backlight module to diffuse the light, the amount of reflected light collected by the camera module is increased, enabling the camera module to obtain a clear image with only a short exposure time, thereby further improving the speed of defect classification.

[0008] In a second aspect, the present invention provides a method for classifying internal defects in glass, applied to the system of the first aspect above. The method includes: when an acquired image contains defects, performing feature extraction on the defects to obtain feature parameters of the defects; classifying the defects based on the feature parameters to determine the type of defects in the glass to be tested corresponding to the image to be tested.

[0009] In one optional implementation, the feature parameters of the defect include the defect shape and the color features within the defect; the feature extraction of the image to be tested to obtain the feature parameters of the defect includes: using a contour extraction algorithm to extract the outer contour and inner contour of the defect from the image to be tested; determining the defect shape based on the outer contour and the inner contour; and extracting the color information in the inner contour as the color features within the defect.

[0010] In one optional implementation, the above-mentioned classification of defects based on the aforementioned feature parameters to determine the defect type of the glass to be tested corresponding to the image to be tested includes: determining the outer contour radius and outer contour elliptic radius based on the outer contour of the defect, wherein the outer contour radius characterizes the similarity between the outer contour of the defect and a standard circle, and the outer contour elliptic radius characterizes the similarity between the outer contour of the defect and a standard ellipse; determining the inner contour parameter based on the inner contour of the defect, wherein the inner contour parameter characterizes the morphological features within the defect; and determining the defect type based on the outer contour radius, the outer contour elliptic radius, the inner contour parameter, and / or the color feature.

[0011] In one optional embodiment, the inner contour parameters include hole parameters, the hole parameters include the number of holes, and the inner contour parameters also include inner contour curvature, inner contour elliptic curvature, and inner contour smoothness. Determining the defect type based on the outer contour curvature, the outer contour elliptic curvature, and the inner contour parameters includes: when the outer contour curvature and the outer contour elliptic curvature are both greater than or equal to a preset outer contour curvature threshold, the inner contour curvature and the inner contour elliptic curvature are both greater than or equal to a preset inner contour curvature threshold, the inner contour smoothness is greater than or equal to a preset smoothness threshold, and the number of holes is a preset number of holes, the defect type is determined to be a bubble defect.

[0012] In an optional implementation, the hole parameters further include hole similarity. The determination of the defect type based on the outer contour arcuateness, the outer contour elliptical arcuateness, and the inner contour parameters includes: when the outer contour arcuateness and the outer contour elliptical arcuateness are both greater than or equal to a preset outer contour arcuateness threshold, the hole similarity is greater than or equal to a preset similarity threshold, and the number of holes is greater than or equal to a preset number of holes, the defect type is determined to be a Glauber's salt bubble defect.

[0013] In an optional embodiment, the inner contour parameters further include inner contour smoothness. The determination of the defect type based on the outer contour radii, the outer contour elliptic radii, the inner contour parameters, and the color features includes: determining the defect type as nickel sulfide defect when the outer contour radii and the outer contour elliptic radii are both less than a preset outer contour radii threshold, the inner contour radii and the inner contour elliptic radii are both less than a preset inner contour radii threshold, the inner contour smoothness is less than a preset smoothness threshold, and the color features meet preset color determination conditions.

[0014] In one optional implementation, classifying the defects based on the feature parameters to determine the defect type of the glass to be tested corresponding to the image to be tested includes: inputting the feature parameters of the defects into a pre-trained neural network to obtain the defect type of the glass to be tested corresponding to the image to be tested.

[0015] Thirdly, the present invention provides a classification device for internal defects in glass, comprising: an extraction module for extracting features of defects from an image containing defects to obtain feature parameters of the defects; and a classification module for classifying the defects based on the feature parameters to determine the type of defects in the glass to be tested corresponding to the image to be tested.

[0016] Fourthly, the present invention provides an electronic device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the glass internal defect classification method of the second aspect above or any corresponding embodiment thereof.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the glass internal defect classification method of the second aspect or any corresponding embodiment described above.

[0018] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the method for classifying internal defects in glass according to the second aspect above or any corresponding embodiment. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of a glass internal defect classification system and a glass to be tested according to an embodiment of the present invention. Figure 2 A schematic diagram comparing typical defect image features in related technologies; Figure 3 This is a schematic diagram of another classification system for internal defects in glass according to an embodiment of the present disclosure and the glass to be tested; Figure 4 This is a schematic diagram of a glass internal defect classification system including a motion positioning module according to an embodiment of the present invention. Figure 5A This is an image of the imaging results from the related technology; Figure 5B An image showing the imaging result according to an embodiment of the present invention; Figure 6 A flowchart illustrating a method for classifying internal defects in glass according to an embodiment of the present invention; Figure 7 This is a structural block diagram of a glass internal defect classification device according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0023] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0024] During the production of float glass, various point defects inevitably occur inside the glass. These defects mainly include bubbles, inclusions, sodium sulfate bubbles, and nickel sulfide particles. Among them, nickel sulfide particles pose a significant threat to the safety of tempered glass products due to their unique physicochemical properties.

[0025] Nickel sulfide particles are typically small and do not wet glass, resulting in a clear interface between them. Consequently, a transition layer with optically gradient characteristics cannot be formed between them. Due to the lack of this transition layer, nickel sulfide particles are difficult to detect with the naked eye and also difficult to distinguish using current optical inspection equipment.

[0026] During the glass tempering process, if the original glass sheet contains nickel sulfide inclusions, these inclusions may be sealed inside the tempered glass as a high-temperature metastable phase during rapid cooling. Subsequently, with the passage of time, temperature changes, or external environmental influences, the high-temperature phase gradually transforms into a low-temperature stable phase. During this phase transformation, the nickel sulfide particles typically undergo a volume expansion of approximately 2% to 4%, thereby generating significant localized tensile stress within the glass. When this stress exceeds the glass's allowable strength limit, it may trigger spontaneous breakage of the tempered glass, commonly referred to as "spontaneous breakage." Even after obtaining the precise three-dimensional spatial coordinates of point defects inside and on the surface of the glass sheet, it is difficult to quickly determine the specific properties of these point defects. Current methods for differentiation largely rely on manual re-inspection or destructive testing, which are inefficient and cannot meet the needs of industrial-scale online full inspection.

[0027] In view of this, the present invention proposes a classification system for internal defects in glass, comprising: a light source module including a light-transmitting area and a light-emitting area, the light-emitting area being used to emit light onto the glass under test; a backlight module, the diffuse reflective surface of the backlight module being close to the glass under test, for diffuse reflection of the light emitted from the light-emitting area through the glass under test; the glass under test being disposed between the light source module and the backlight module, a first gap being formed between the glass under test and the light source module, and a second gap being formed between the glass under test and the backlight module; a camera module, corresponding to the light-transmitting area of ​​the light source module, for acquiring images of the glass under test through the light-transmitting area; and a processing module, connected to the camera module, for classifying defects contained in the glass under test based on the acquired images.

[0028] Figure 1 This is a schematic diagram of a glass internal defect classification system and a glass to be tested according to an embodiment of the present invention.

[0029] like Figure 1 As shown, the system includes a camera unit 101, a lens unit 102, a light source module 103, a glass under test 104, a defect 105, a backlight module 106, and a processing module 107. The camera unit 101 can be a high-speed color area array camera, and the lens unit 102 can be a low-distortion lens. The camera's optical axis is preferably perpendicular to the surface of the glass under test to ensure consistent field of view across different imaging modes. The light source module 103 can be a ring-shaped (including circular or other polygonal) light source with a central hole or a surface light source (corresponding to the light-transmitting and light-emitting areas). The light source can be an LED, an OLED matrix, or a ring-shaped fluorescent lamp, positioned above the glass under test to provide oblique incident illumination to the target area in reflective imaging mode. It is important to note that the preferred incident angle of the light source module is 15°–45°. By controlling the incident angle, the outer contour, internal texture, and metallic color difference characteristics of the defect can be more clearly presented. The backlight module 106 can be a white diffuse reflective substrate, used to provide uniform reflected background light for the target defect in reflective imaging mode; the white reflective substrate can be, but is not limited to, a flat plate coated with white reflective paint, white plastic sheet such as acrylic, PC, PVC, PET, PP, white ceramic plate, white enamel plate, or white fabric. The aforementioned defects can include bubble defects, stone defects, sodium sulfate bubble defects, spherical silicon defects, and nickel sulfide defects. The processing module 107 can be a standalone industrial control computer, embedded industrial control computer, PLC controller, edge computing device, or a computing unit deployed on a local server or cloud server. The processing module 107 includes at least a processor and a memory, the memory storing executable computer program instructions, which, when executed by the processor, implement the defect classification function.

[0030] For ease of understanding, the following will be explained... Figure 2 Compare the defect characteristics.

[0031] Figure 2 This is a schematic diagram comparing the image features of typical defects in related technologies.

[0032] like Figure 2As shown, typical defects include bubble defects, stone defects, sodium sulfate bubble defects, spherical silicon defects, and nickel sulfide defects. The characteristics of each typical defect are described below. Bubble: The outer contour is round or elliptical, with a smooth outline and an inner contour similar to the outer contour. The hollow part is the same color as the background. Stone: It has an irregular shape. Sodium sulfate bubble: The outer contour is similar to that of a bubble, but the interior shows irregular spots of the same color as the background. Spherical silicon defect: The outer contour is round or elliptical, with a bright inner ring similar to the outer contour in the middle. There is often a dark spot in the middle of the ring. Nickel sulfide defect: The outer contour is generally smooth, round or elliptical, but some defects have microcracks adjacent to the glass due to expansion, resulting in an irregular state of the outer contour. The interior of the defect shows irregular bright spots with a lower brightness than spherical silicon defects, and the bright spots have a metallic copper color different from the reflected background.

[0033] Furthermore, the backlight module can also be used to transmit light through the glass under test, providing uniformly transmitted background light for the camera module.

[0034] Figure 3 This is a schematic diagram of another classification system for internal defects in glass according to an embodiment of the present disclosure, and of the glass to be tested.

[0035] like Figure 3 As shown, the system includes a camera unit 101 of the camera module, a lens unit 102 of the camera module, a light source module 103, a glass under test 104, a defect 105, a backlight module 106, and a processing module 107. The backlight module 106 further includes a light-emitting unit 1061 and a light-diffusing unit 1062. The backlight module provides uniformly transmitted background light to the defect in the transmission illumination function. The light-diffusing unit 1062 can be made of a material that can both diffuse and transmit light and function as a reflector, and can be selected from, but is not limited to, milky white high-transmittance materials such as acrylic sheets, PC sheets, PVC sheets, PET sheets, lightbox fabric, or PTFE high-temperature fabric. The light-emitting unit can use a device similar to the light source module, which will not be described in detail here.

[0036] In one feasible embodiment of the present invention, the system may further include a motion positioning module for moving the camera module.

[0037] Figure 4 This is a schematic diagram of a glass internal defect classification system including a motion positioning module, according to an embodiment of the present invention.

[0038] like Figure 4As shown, the system demonstrates the camera unit 101, lens unit 102, light source module 103, glass under test 104, Z-axis linear guide module 401, Y-axis linear guide module 402, and X-axis linear motor module 403 of the motion positioning module. When photographing the glass under test, the motion positioning module first moves the image acquisition module to the target defect location based on three-dimensional spatial coordinates, and adjusts the Z-axis linear guide module 401 to ensure a clear image of the target defect. Next, the light source module 103 is activated to acquire a reflected image of the target defect in reflection imaging mode. If necessary, the light-emitting unit 1061 and the light-diffusing unit 1062 in the backlight module 106 are activated to acquire a transmitted image of the same target defect in transmission imaging mode (the transmitted image is not mandatory). Then, image processing is performed on the image under test, and based on the processed image, the contour morphology features, edge features, internal texture features, color features, and light transmittance features of the defect are extracted. Finally, the defects are classified based on the extracted features to output the defect types.

[0039] When the system is working, the light-emitting units 1061 in the light source module 103 and the backlight module 106 can be triggered separately. This invention does not restrict the triggering order and can determine the triggering order according to the actual situation. In reflection mode, only the light source module 103 is triggered, and in transmission mode, the light-emitting unit 1061 is triggered, so that light passes through the glass under test.

[0040] It should be noted that, because this diagram is a top view, the above-mentioned features are not shown. Figure 1 and Figure 3 The system includes a backlight module 106 and a processing module 107 connected via a network or wired connection, but both the system and the backlight module are present.

[0041] In one feasible embodiment of the present invention, the light source module emits light at a small angle, with most of the light projected onto a white diffuse reflector (backlight module) beneath the glass under test. Due to diffuse reflection, the direction of light propagation is redistributed, allowing some light to enter the lens of the area array camera (camera module). In defect-free areas, light enters the camera directly without obstruction, thus forming a bright background in the image. When transparent or translucent defects (such as bubbles or sodium sulfate bubbles) are present, these defects partially block the reflected light from the diffuse reflector to the camera, causing the corresponding area to exhibit brightness differences in the image. For completely opaque defects (such as silicon spheres), they completely block the background light reflected from the diffuse reflector to the camera. However, at the same time, the light directly illuminating the surface of the defect particles by the ring light source is received by the camera due to the reflective properties of the defect itself, thus forming an image with specific reflective characteristics within the defect boundary in the image. Silicon sphere defects belong to this type.

[0042] It should be noted that, compared to related technologies, the backlight module in this invention plays a decisive role. Since the glass itself is transparent, without a diffuser, the camera can collect very little reflected light, resulting in an image with extremely low brightness, almost unusable for subsequent image processing. However, by adding a highly reflective white diffuser under the glass, a clear image can be obtained with only a very short exposure time.

[0043] Figure 5A This is an image of the imaging results from a related technology.

[0044] Figure 5B This is an image showing the imaging result according to an embodiment of the present invention.

[0045] like Figure 5A and Figure 5B As shown, it can be seen that, compared with related technologies, the system provided by the embodiments of the present invention has better imaging results.

[0046] According to an embodiment of the present invention, the exposure time for capturing a clear bubble image as shown in Figure 5 above is only 80 μs. In related technologies lacking a backlight module, a high-brightness light source and a longer exposure time (e.g., 300 μs) are required to capture large-sized defects, while for small-sized defects (e.g., 0.2 mm), it is difficult to form a clear and distinguishable image. Therefore, the backlight module of the present invention, by providing diffused light, at least partially reduces the exposure time of the camera module, thereby improving the defect classification speed.

[0047] According to an embodiment of the present invention, when transmitting light through the glass under test, the processing module 107 can receive the three-dimensional spatial coordinate information of the point-like defects to be determined sent by the upstream defect detection equipment. The three-dimensional spatial coordinate information includes at least X, Y, and Z coordinates. The motion positioning module then uses the three-dimensional coordinate information to move the camera module to the target location containing the defect and completes focusing. The motion positioning module can employ a translation mechanism that moves along the width and length directions of the glass, and cooperate with a Z-axis focusing mechanism to achieve fixed-point imaging of the target defect. Furthermore, since the defect coordinates are known, only the target defect needs to be re-examined at a fixed point, eliminating the need for repeated scanning of the entire plate, resulting in higher detection efficiency.

[0048] This embodiment provides a method for classifying internal defects in glass, which can be used in the aforementioned processing module. Figure 6 A flowchart of a method for classifying internal defects in glass according to an embodiment of the present invention is shown below. Figure 6 As shown, the process includes the following steps: Step S601: When the acquired image contains defects, feature extraction is performed on the defects to obtain their feature parameters.

[0049] According to embodiments of the present invention, traditional machine vision algorithms (such as color space conversion, edge detection, morphological processing, etc.) can be used to extract defect features, and the present invention does not limit this. Before feature extraction from the image, Gaussian filtering can be applied to the image to reduce noise, and then binarization processing can be performed on the denoised image to separate the defect from the background of the glass under test.

[0050] Step S602: Classify defects based on feature parameters to determine the type of defect in the glass to be tested corresponding to the image to be tested.

[0051] Specifically, the defect classification can be implemented using a traditional classifier or a deep learning model. The traditional classifier classifies based on manually extracted feature parameters, while the deep learning model learns and classifies features based on the image under test or the feature parameters, thereby outputting the corresponding defect category. According to an embodiment of the present invention, when the acquired image contains defects, feature parameters of the defects are obtained by feature extraction, and then the defects are classified based on these feature parameters to determine the defect type of the glass under test. This avoids destructive sampling or physical contact with the glass under test, achieving non-destructive testing. Simultaneously, both feature extraction and classification processes are automatically executed by the processing module, eliminating the need for manual re-inspection, shortening the defect determination cycle, and achieving rapid detection. Furthermore, by using the feature parameters as the classification basis, the feature differences between different defect types are distinguished, obtaining accurate defect type determination results and avoiding the risk of glass spontaneous breakage due to misjudgment of defects.

[0052] According to an embodiment of the present invention, the feature parameters of the above-mentioned defect include the defect shape and the color features within the defect; feature extraction is performed on the image to be tested to obtain the feature parameters of the defect, including: using a contour extraction algorithm to extract the outer contour and inner contour of the defect from the image to be tested; determining the defect shape based on the outer contour and inner contour; and extracting the color information in the inner contour of the defect as the color features within the defect.

[0053] In one embodiment of the present invention, connected component analysis can be performed on the binary image to extract various feature parameters of the connected components. Specifically, a two-pass scanning method can be used to obtain the total area and centroid of the connected components, and the outer and inner contours of the defects can be extracted using a contour extraction algorithm. The aforementioned contour extraction algorithm may include the Suzuki contour tracking algorithm, contour extraction algorithms based on edge detection operators, etc., and the present invention does not limit it to these algorithms.

[0054] Furthermore, the color information of the inner contour-defined region of the image under test can be converted into color values ​​in the Lab color space to serve as color features within the defect.

[0055] According to an embodiment of the present invention, classifying defects based on feature parameters and determining the defect type of the glass to be tested corresponding to the image to be tested includes: determining the outer contour radius and outer contour elliptic radius based on the outer contour of the defect, wherein the outer contour radius characterizes the similarity between the outer contour of the defect and a standard circle, and the outer contour elliptic radius characterizes the similarity between the outer contour of the defect and a standard ellipse; determining the inner contour parameters based on the inner contour of the defect, wherein the inner contour parameters characterize the morphological features within the defect; and determining the defect type based on the outer contour radius, outer contour elliptic radius, inner contour parameters, and / or color features.

[0056] Based on such Figure 2 The typical defect image features shown are described below, and the criteria for determining the type of defect in this invention are detailed below.

[0057] In bubble defects, both the outer and inner contours are smooth circles or ellipses, with similar shapes and approximately translated geometric relationships, resulting in a hollow interior in the transmission image. In spherical silicon defects, the outer contour is a smooth circle or ellipse, and the inner contour contains two smooth circles or ellipses. These two inner contours and the outer contour form a similar shape and approximately translated geometric relationship, resulting in a hollow interior in the transmission image. In stone defects, the outer contour is irregular, and the transmission image shows a hollow interior. In sodium sulfate bubble defects, the outer contour conforms to the characteristics of a circle or ellipse; its inner contour contains multiple cavities, and the similarity between the shapes of these cavities is low, with significant differences in geometric parameters such as area and perimeter. The transmission characteristics show irregular and varying numbers of cavities inside the transmission image. In nickel sulfide defects, the overall outer contour is circular or elliptical. Expansion of some nickel sulfide defects may cause cracks in the surrounding glass, resulting in localized irregularities in the contour. The interior of the defect displays irregular bright spots with low brightness. If the image resolution is high enough, pits similar to those on a golf ball can be observed, and the bright spots exhibit the characteristic color of metallic copper—a reddish-yellow to brownish-yellow metallic luster under reflected light, located within the range defined by the red and yellow vectors on the color coordinate system. In the transmission image, the spherical portion is completely opaque. Therefore, defect judgment criteria can be further designed based on these defect characteristics.

[0058] According to an embodiment of the present invention, the above-mentioned defect classification based on feature parameters to determine the defect type of the glass to be tested corresponding to the image to be tested includes: determining the outer contour radius and outer contour elliptic radius based on the outer contour of the defect, wherein the outer contour radius characterizes the similarity between the outer contour of the defect and a standard circle, and the outer contour elliptic radius characterizes the similarity between the outer contour of the defect and a standard ellipse; determining the inner contour parameters based on the inner contour of the defect, wherein the inner contour parameters characterize the morphological features within the defect; and determining the defect type based on the outer contour radius, outer contour elliptic radius, inner contour parameters, and / or color features.

[0059] In one feasible embodiment of the present invention, the curvature of the outer contour of the defect can be fitted by the least squares method or the curvature of the outer contour of the defect can be calculated based on the area and perimeter of the defect. Furthermore, the fitted ellipse can be obtained by the least squares method based on algebraic distance, and then the elliptic curvature can be calculated based on the fitted ellipse.

[0060] According to an embodiment of the present invention, the aforementioned inner contour parameters include hole parameters, which include the number of holes. The inner contour parameters also include inner contour curvature, inner contour elliptic curvature, and inner contour smoothness. Determining the defect type based on the outer contour curvature, outer contour elliptic curvature, and inner contour parameters includes: when the outer contour curvature and outer contour elliptic curvature are both greater than or equal to a preset outer contour curvature threshold, the inner contour curvature and inner contour elliptic curvature are both greater than or equal to a preset inner contour curvature threshold, the inner contour smoothness is greater than or equal to a preset smoothness threshold, and the number of holes is a preset number of holes, the defect type is determined to be a bubble defect.

[0061] In one feasible embodiment of the present invention, the aforementioned preset outer contour curvature threshold can be set to 0.85, for example, and the aforementioned preset inner contour curvature threshold can be set to 0.8, for example. The outer contour circular curvature and the outer contour elliptical curvature can correspond to different preset curvature thresholds, and the present invention does not limit this.

[0062] Due to the morphological characteristics of void defects, judgment criteria can be set by the number of holes and the curvature, ellipticity, and inner curvature of the outer and inner contours. The Suzuki contour tracking algorithm can also record the hierarchical information of the contour to determine the number of holes. The smoothness of the inner contour can be determined by performing curvature analysis on the inner contour and calculating the distance residual from the contour points to the fitted curve. Therefore, when judging void defects, we can determine whether the outer contour of the defect conforms to the characteristics of a circle or an ellipse, and then determine whether the inner contour of the defect meets the conditions of having only a single hole, smooth contour, and a shape that is circular, elliptical, or a local approximation thereof.

[0063] According to an embodiment of the present invention, the above-mentioned hole parameters also include hole similarity. The defect type is determined based on the outer contour arc, outer contour elliptical arc, and inner contour parameters, including: when the outer contour arc and outer contour elliptical arc are both greater than or equal to a preset outer contour arc threshold, the hole similarity is greater than or equal to a preset similarity threshold, and the number of holes is greater than or equal to a preset number of holes, the defect type is determined to be a Glauber's salt bubble defect.

[0064] Due to the defect characteristics of Glauber's salt bubbles, we can first use the algorithm described above to determine whether there are multiple holes within their inner contour, and then use the Hu invariant moment algorithm to determine the similarity of the holes. Therefore, when determining the defects of Glauber's salt bubbles, we can determine whether the outer contour of the defect conforms to the characteristics of a circle or ellipse and whether there are multiple holes in the inner contour. If multiple holes exist, we can further determine the similarity between the holes and whether there are significant differences in their geometric parameters such as area and perimeter.

[0065] Due to the defective characteristics of the stone, it can be determined whether its shape is neither round nor elliptical, that is, the arc degree and elliptical arc degree of the defect outside the contour are both less than the preset outer contour arc degree threshold.

[0066] Due to the defect characteristics of spherical silicon, it can be determined whether the outer contour of the defect is a smooth circle or ellipse; then, the inner contour is fitted to determine whether it contains two smooth circles or ellipses. If it does, it is further determined whether the two inner contours and the outer contour form a geometric relationship of similar shape and approximately translation of each other, which is used as a condition for determining the type of defect.

[0067] According to an embodiment of the present invention, the aforementioned inner contour parameters further include inner contour smoothness. Determining the defect type based on outer contour radii, outer contour elliptic radii, inner contour parameters, and color features includes: when both outer contour radii and outer contour elliptic radii are less than a preset outer contour radii threshold, both inner contour radii and inner contour elliptic radii are less than a preset inner contour radii threshold, inner contour smoothness is less than a preset smoothness threshold, and color features meet preset color determination conditions, the defect type is determined to be a nickel sulfide defect.

[0068] Due to the defect characteristics of nickel sulfide, it can be determined whether the outer contour of the defect is a smooth circle or ellipse; then the inner contour is fitted to determine whether its lines are not smooth, whether the shape is neither a circle nor an ellipse, and whether the overall shape is irregular; if the above inner contour characteristics are met, the color information of the area defined by the inner contour in the original color image is extracted and converted into color values ​​in the Lab color space to determine whether the color value is similar to the preset nickel sulfide.

[0069] According to embodiments of the present invention, the unique circular or elliptical contour of nickel sulfide, its luster similar to that of metallic copper, its opacity, and the presence of microcracks are used as judgment criteria. The outer contour curvature, outer contour elliptical curvature, and color characteristics are obtained to jointly determine the defects of nickel sulfide, thereby achieving accurate determination of the defect type.

[0070] This embodiment also provides a glass internal defect classification device, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0071] This embodiment provides a device for classifying internal defects in glass, such as... Figure 7 As shown, it includes: Extraction module 701 is used to extract features of defects from the acquired image to obtain feature parameters of the defects. The classification module 702 is used to classify defects based on feature parameters in order to determine the type of defect in the glass to be tested corresponding to the image to be tested.

[0072] The glass internal defect classification device provided in this embodiment of the invention can execute the glass internal defect classification method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0073] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0074] The following is a detailed reference. Figure 8 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 801, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 802 or a program loaded from memory 808 into random access memory (RAM) 803. The RAM 803 also stores various programs and data required for the operation of the electronic device. The processor 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0075] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 808 including, for example, magnetic tapes, hard disks, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0076] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 809, or installed from a memory 808, or installed from a ROM 802. When the computer program is executed by the processor 801, it performs the functions defined in the glass internal defect classification method of the embodiments of the present invention.

[0077] Figure 8 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0078] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium after being downloaded via a network. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the method for classifying internal glass defects shown in the above embodiments is implemented.

[0079] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0080] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A classification system for internal defects in glass, characterized in that, The system includes: The light source module includes a light-transmitting area and a light-emitting area, wherein the light-emitting area is used to emit light onto the glass under test; A backlight module, wherein the diffuse reflective surface of the backlight module is close to the glass under test, for diffuse reflection of the light emitted by the light-emitting area through the glass under test; The glass under test is disposed between the light source module and the backlight module, a first gap is formed between the glass under test and the light source module, and a second gap is formed between the glass under test and the backlight module; A camera module is configured to correspond to the light-transmitting area of ​​the light source module, and is used to acquire images of the glass under test through the light-transmitting area; The processing module, connected to the camera module, is used to classify defects contained in the glass under test based on the acquired images.

2. The system according to claim 1, characterized in that, The diffuse reflective surface of the backlight module is a light-transmitting surface. The backlight module also includes a light-emitting unit and a light-diffusing unit, which are disposed on the back side of the diffuse reflective surface. The light-emitting unit is used to emit light, and the light-diffusing unit is used to disperse the light emitted by the light-emitting unit.

3. The system according to claim 1, characterized in that, The system also includes a motion positioning module, which is used to move the camera module.

4. A method for classifying internal defects in glass, characterized in that, A classification system for internal defects in glass according to any one of claims 1-3; the method comprising: When the acquired image contains defects, feature extraction is performed on the defects to obtain the feature parameters of the defects; Based on the feature parameters, the defects are classified to determine the type of defect in the glass to be tested corresponding to the image to be tested.

5. The method according to claim 4, characterized in that, The feature parameters of the defect include the defect shape and the color features within the defect; the feature extraction of the image to be tested to obtain the feature parameters of the defect includes: Using a contour extraction algorithm, the outer contour and inner contour of the defect are extracted from the image to be tested. The defect shape is determined based on the outer contour of the defect and the inner contour of the defect; The color information in the inner contour of the defect is extracted as the color feature of the defect.

6. The method according to claim 5, characterized in that, The step of classifying defects based on the feature parameters to determine the type of defect in the glass to be tested corresponding to the image to be tested includes: The outer contour radius and outer contour elliptic radius are determined based on the outer contour of the defect. The outer contour radius represents the degree of similarity between the outer contour of the defect and a standard circle, and the outer contour elliptic radius represents the degree of similarity between the outer contour of the defect and a standard ellipse. The inner contour parameters are determined based on the inner contour of the defect, and the inner contour parameters characterize the morphological features within the defect. The defect type is determined based on the outer contour radius, the outer contour elliptic radius, the inner contour parameters, and / or the color characteristics.

7. The method according to claim 6, characterized in that, The inner contour parameters include hole parameters, which include the number of holes. The inner contour parameters also include inner contour roundness, inner contour ellipticity, and inner contour smoothness. Determining the defect type based on the outer contour roundness, outer contour ellipticity, and inner contour parameters includes: If the outer contour radius and the outer contour elliptical radius are both greater than or equal to a preset outer contour radius threshold, the inner contour radius and the inner contour elliptical radius are both greater than or equal to a preset inner contour radius threshold, the inner contour smoothness is greater than or equal to a preset smoothness threshold, and the number of holes is a preset number of holes, then the defect type is determined to be a bubble defect.

8. The method according to claim 7, characterized in that, The hole parameters also include hole similarity. Determining the defect type based on the outer contour curvature, the outer contour elliptic curvature, and the inner contour parameters includes: If both the outer contour radius and the outer contour elliptical radius are greater than or equal to a preset outer contour radius threshold, the hole similarity is greater than or equal to a preset similarity threshold, and the number of holes is greater than or equal to a preset number of holes, then the defect type is determined to be a Glauber's salt bubble defect.

9. The method according to claim 7, characterized in that, The inner contour parameters also include inner contour smoothness. Determining the defect type based on the outer contour roundness, the outer contour ellipticity, the inner contour parameters, and the color features includes: If the outer contour radius and the outer contour elliptical radius are both less than a preset outer contour radius threshold, the inner contour radius and the inner contour elliptical radius are both less than a preset inner contour radius threshold, the inner contour smoothness is less than a preset smoothness threshold, and the color feature meets a preset color determination condition, then the defect type is determined to be a nickel sulfide defect.

10. The method according to claim 4, characterized in that, The step of classifying the defects based on the feature parameters to determine the type of defect in the glass to be tested corresponding to the image to be tested includes: The feature parameters of the defect are input into a pre-trained neural network to obtain the defect type of the glass to be tested corresponding to the image to be tested.

11. A device for classifying internal defects in glass, characterized in that, The device includes: The extraction module is used to extract features of the defects when the acquired image contains defects, thereby obtaining the feature parameters of the defects. The classification module is used to classify the defects based on the feature parameters in order to determine the type of defect in the glass to be tested corresponding to the image to be tested.

12. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the method for classifying internal defects in glass as described in any one of claims 4 to 10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the method for classifying internal defects in glass according to any one of claims 4 to 10.

14. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the method for classifying internal defects in glass according to any one of claims 4 to 10.

Images