Float glass defect detection method and device

Abstract

The invention discloses a defect detection method and device for float glass. The defect detection method for the float glass comprises the following steps: acquiring a fitting curve of a deformation distance and an optical deformation angle; obtaining a target image of the float glass to be detected, and obtaining a deformation distance of the float glass to be detected according to the target image; and calibrating the fitted curve, and substituting the deformation distance of the float glass to be measured into the calibrated fitted curve to obtain the optical deformation angle of the float glass to be measured. According to the method, the deformation distance of the float glass is detected on line, and then the deformation distance is substituted into the calibrated fitting curve of the deformation distance and the optical deformation angle, so that on-line detection of the optical deformation angle of the deformation defect on the float glass can be realized; the step of performing offline detection on the float glass by using a measuring tool such as a zebra angle detector is omitted, so that a user can quickly and accurately obtain the deformation data of the float glass and evaluate the deformation defect.

Description

Technical Field

[0001] The present invention relates to the field of glass detection, and in particular to a method and device for detecting defects of float glass. Background Art

[0002] Float glass is made by melting evenly molten glass through a runner into a tin bath. Due to its own gravity and surface tension, it is naturally flattened into a glass ribbon on the surface of the molten tin, and then pulled by external force and adjusted in temperature. As float glass has a relatively uniform thickness and high transparency, it is increasingly used.

[0003] Especially now in response to the requirements of environmental protection and energy conservation, most of the float glass has begun to be tempered and coated. In this process, it is necessary to place metal fixing supports on the upper surface of the float glass to ensure the parallelism of the glass. However, due to the different material properties of the glass material and the metal fixing supports, the expansion coefficients of the two are very different, resulting in the difficulty in ensuring the consistency of the flatness of the float glass itself after the tempering treatment. The surface of the float glass is prone to defects such as edge collapse, bumps, white spots, bruises, scratches and deformation. Therefore, it is very important to detect the surface of the float glass.

[0004] The existing online detection technology for float glass deformation detection can usually only obtain the deformation range (i.e. deformation distance) and use it to evaluate the deformation defect. If you want to further detect the optical deformation angle (i.e. zebra angle) of the float glass, you need to use a measuring tool such as a zebra angle detector for offline detection, and the detection process is very complicated. Summary of the invention

[0005] The main purpose of the present invention is to provide a method for detecting defects of float glass, aiming to solve the problem that the existing online monitoring technology cannot detect the optical deformation angle of float glass.

[0006] To achieve the above object, the present invention provides a float glass defect detection method, the defect detection method comprising:

[0007] Obtaining a fitting curve of deformation distance and optical deformation angle;

[0008] Acquire a target image of the float glass to be tested, and acquire a deformation distance of the float glass to be tested according to the target image;

[0009] The fitting curve is calibrated, and the deformation distance of the float glass to be measured is substituted into the calibrated fitting curve to obtain the optical deformation angle of the float glass to be measured.

[0010] In some embodiments, obtaining a fitting curve of the deformation distance and the optical deformation angle includes:

[0011] Placing a plurality of lenses with different refractive powers at a detection position and collecting a calibration image of each lens; obtaining a deformation distance of the lens according to the calibration image, thereby performing data fitting between the refractive power and the deformation distance of the lens to obtain a first functional relationship;

[0012] Acquiring the optical deformation angle of the lens, thereby performing data fitting on the diopter and the optical deformation angle of the lens to obtain a second functional relationship;

[0013] A fitting curve of the deformation distance and the optical deformation angle is calculated according to the first functional relationship and the second functional relationship.

[0014] In some embodiments, placing a plurality of lenses with different refractive powers at a detection position and collecting a calibration image of each lens; obtaining the deformation distance of the lens according to the calibration image, thereby performing data fitting between the refractive power and the deformation distance of the lens to obtain the first functional relationship includes:

[0015] Acquiring a calibration image of the lens by a line scan camera;

[0016] Selecting a number of sample areas at intervals on the calibration image along the length direction, and obtaining deformation distances of the calibration image located in the number of sample areas;

[0017] Fitting the deformation distances of the plurality of sample areas with the refractive power to generate a plurality of initial models;

[0018] The plurality of initial models are fitted to obtain a first functional relationship.

[0019] In some embodiments, the plurality of lenses include lenses having refractive powers of 0 dpt, ±20 mdpt, ±40 mdpt, ±60 mdpt, ±80 mdpt, and ±100 mdpt, respectively.

[0020] In some embodiments, the calibration fit curve comprises:

[0021] Place three lenses with diopters of 0 and ±100 mdpt at the detection position to obtain the corresponding deformation distances of the three lenses;

[0022] The fitting curve is calibrated according to the deformation distances corresponding to the three lenses.

[0023] In some embodiments, acquiring a target image of the float glass to be tested includes:

[0024] Collecting the transmitted light and reflected light of the float glass surface to be tested;

[0025] Generate a target image of the float glass to be tested.

[0026] In some embodiments, after generating the target image of the float glass to be tested, the method further includes:

[0027] Acquire defects on the float glass to be tested according to the target image, and calculate the size of the defects;

[0028] The defects include one or more of cracks, dust, bubbles, inclusions, tin spots and scratches.

[0029] The present invention further provides a float glass defect detection device, comprising:

[0030] A standard calibration module, comprising a plurality of lenses with different refractive powers, for obtaining a deformation distance of the lens and an optical deformation angle of the lens;

[0031] An imaging light source module, used for irradiating the float glass to be tested to generate a target image;

[0032] An image acquisition module, used for acquiring the target image;

[0033] A data processing module is connected to the standard calibration module and the image acquisition module, and is used to generate and calibrate a fitting curve according to the deformation distance of the lens and the optical deformation angle of the lens, and to obtain the deformation distance in the target image, and substitute the deformation distance in the target image into the fitting curve to obtain the optical deformation angle of the float glass to be tested.

[0034] In some embodiments, the imaging light source module includes:

[0035] A reflecting light source, wherein the reflecting light source can illuminate the float glass to be tested to generate reflected light;

[0036] A transmission light source, which can emit light of two different wavelengths;

[0037] A filter is arranged between the transmitted light source and the float glass to be tested, and a plurality of light-blocking sections are arranged at intervals on the filter, and the light-blocking sections can only allow one of the two different wavelengths of light to pass through, so that the two different wavelengths of light form stripe light and uniform backlight after passing through the filter, and the stripe light and the uniform backlight can illuminate the float glass to be tested to generate transmitted light.

[0038] In some embodiments, the data processing module includes:

[0039] A data preprocessor, connected to the image acquisition module, used to set the photographing parameters of the image acquisition module and process the target image to generate fringe data;

[0040] A data processor, connected to the standard calibration module and the data preprocessor, for receiving the deformation distance of the lens and the optical deformation angle of the lens and generating the fitting curve; and the data processor can receive the fringe data and output the deformation distance of the float glass to be measured according to the fringe data;

[0041] The industrial computer is connected to the data processor, and is used for receiving the deformation distance of the float glass to be tested, substituting the deformation distance of the float glass to be tested into the fitting curve, and outputting the optical deformation angle of the float glass to be tested to an external device.

[0042] The technical solution provided by the present invention can realize online detection of the optical deformation angle of deformation defects on the float glass by online detecting the deformation distance of the float glass and then substituting the deformation distance into the fitting curve of the calibrated deformation distance and the optical deformation angle, thereby eliminating the step of using a measuring tool such as a zebra angle detector to perform offline detection on the float glass, allowing users to quickly and accurately obtain deformation data of the float glass and evaluate deformation defects, thereby facilitating users to timely adjust production plans and solve problems arising in the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of an embodiment of a method for defect detection of float glass of the present invention;

[0044] Figure 2 A flow chart of another embodiment of a method for detecting defects in float glass according to the present invention;

[0045] Figure 3 A flow chart of another embodiment of a method for detecting defects in float glass according to the present invention;

[0046] Figure 4 A flow chart of another embodiment of a method for detecting defects in float glass according to the present invention;

[0047] Figure 5 A flow chart of another embodiment of a method for detecting defects in float glass according to the present invention;

[0048] Figure 6 A flow chart of another embodiment of a method for detecting defects in float glass according to the present invention;

[0049] Figure 7 It is a structural schematic diagram of a defect detection device for float glass of the present invention;

[0050] Figure 8 It is a structural schematic diagram of a data processing module in a float glass defect detection device of the present invention;

[0051] Fig. 9A schematic diagram of a calibration image in an embodiment of a defect detection device for float glass of the present invention;

[0052] Fig.10 for Fig. 9 Schematic diagram of the cosine function of the fringe data.

[0053] Reference numerals:

[0054] 100, standard calibration module; 200, imaging light source module; 300, image acquisition module; 400, data processing module; 410, data preprocessor; 420, data processor; 430, industrial computer. DETAILED DESCRIPTION

[0055] The scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a 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 are within the scope of protection of the present invention.

[0056] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0057] It should also be noted that when an element is referred to as being "fixed on" or "disposed on" another element, it may be directly on the other element or there may be an intermediate element at the same time. When an element is referred to as being "connected to" another element, it may be directly connected to the other element or there may be an intermediate element at the same time.

[0058] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0059] The present invention provides a float glass defect detection method, referring to Figure 1 , the defect detection method comprises:

[0060] Step S100: obtaining a fitting curve of deformation distance and optical deformation angle.

[0061] The actual measured data of deformation distance and optical deformation angle can be obtained through experiments, and then the obtained measured data can be fitted to generate a fitting curve. For example, a group of float glass samples can be selected, and these samples can be deformed to different degrees. The deformation treatment should refer to the deformation that may occur in the actual production process of float glass as much as possible, so that the data has sufficient accuracy; then the float glass samples are measured using an optical measuring instrument to obtain the optical deformation angle and deformation distance after deformation; then a mathematical model (such as polynomial fitting, linear fitting or nonlinear fitting) is used to generate a fitting curve between the deformation distance and the optical deformation angle. It can also be simulated by using finite element analysis (FEA) and optical simulation software to simulate float glass with different deformation degrees, then obtain the deformation distance of float glass, and simulate the influence of different deformation acceptances on the optical deformation angle, and finally fit the simulated data to obtain a fitting curve of deformation distance and optical deformation angle. It can be understood that the historical data of deformation distance and optical deformation angle corresponding to float glass with different deformation degrees can also be collected, and then a regression model (such as linear regression, support vector regression, neural network, etc.) is used to fit the data to generate a fitting curve of deformation distance and optical deformation angle.

[0062] Step S200: obtaining a target image of the float glass to be tested, and obtaining a deformation distance of the float glass to be tested according to the target image;

[0063] In this embodiment, machine vision technology is used to detect the surface defects of float glass. Machine vision technology is used to use machines instead of people to make measurements and judgments. Further, this embodiment usually uses a high-definition lens and a line scan camera to collect the target image. The deformation distance can be obtained by the deviation between the target image and the standard image. For example, a stripe light source (which can emit stripe light) can be used to illuminate the float glass to be tested. If the surface of the float glass to be tested is flat, the target image obtained is neat stripes. If the surface of the float glass to be tested is deformed, the stripes in the target image obtained will be distorted. The deformation distance of the float glass to be tested can be obtained by calculating the degree of distortion.

[0064] Step S300: calibrate the fitting curve, and substitute the deformation distance of the float glass to be tested into the calibrated fitting curve to obtain the optical deformation angle of the float glass to be tested.

[0065] Since the experimental environment is different from the actual production environment, the position of the light source, float glass, and the position and model of the camera will affect the test results. Therefore, before the fitting curve is applied specifically, the fitting curve needs to be calibrated to improve the accuracy of the test results. Specifically, several float glass samples can be tested in the actual production environment, and then the data obtained after the test (that is, the deformation distance and the optical deformation angle) are substituted into the fitting curve, the error of the fitting curve is analyzed, and the error between the data of the float glass sample is reduced by adjusting the parameters of the fitting curve. The correction of the parameters can be completed by linear regression correction, polynomial correction, local weighted regression, least squares method, machine learning calibration, etc., and this application does not limit this.

[0066] The technical solution provided by the present invention can realize online detection of the optical deformation angle of deformation defects on the float glass by online detecting the deformation distance of the float glass and then substituting the deformation distance into the fitting curve of the calibrated deformation distance and the optical deformation angle, thereby eliminating the step of using a measuring tool such as a zebra angle detector to perform offline detection on the float glass, allowing users to quickly and accurately obtain deformation data of the float glass and evaluate deformation defects, thereby facilitating users to timely adjust production plans and solve problems arising in the production process.

[0067] like Figure 2 As shown, in some embodiments, step S100 includes:

[0068] Step S110: placing a plurality of lenses with different diopters at a detection position, and collecting a calibration image of each lens; obtaining a deformation distance of the lens according to the calibration image, thereby performing data fitting between the diopters and the deformation distance of the lens to obtain a first functional relationship;

[0069] Step S120: acquiring the optical deformation angle of the lens, thereby performing data fitting on the diopter and the optical deformation angle of the lens to obtain a second functional relationship;

[0070] Step S130: Calculate and obtain a fitting curve of the deformation distance and the optical deformation angle according to the first functional relationship and the second functional relationship.

[0071] In this embodiment, since the refractive power of the lens itself (used to measure the ability of the lens to refract light) is known and fixed, the lens is used as a sample, a calibration image of the lens is collected and the deformation distance corresponding to the lens is calculated, and then data fitting is performed with the refractive power parameters corresponding to the lens to generate the relationship between the deformation distance and the refractive power, that is, the first functional relationship; then the optical deformation angle of the lens can be measured by a zebra angle tester or other equipment, or the optical deformation angle of the lens can be obtained by a transmission deformation test method, a stress birefringence method, an optical profilometer, etc., and then data fitting is performed with the refractive power parameters corresponding to the lens to generate the relationship between the refractive power and the optical deformation angle, that is, the second functional relationship; finally, the first functional relationship and the second functional relationship can be fitted into a new model, that is, a fitting curve of the deformation distance and the optical deformation angle.

[0072] It is understandable that a lens with a diopter within a suitable range can be selected according to actual needs. For example, if the general deformation of float glass during the production process is large, multiple lenses with large diopters can be selected as samples, thereby improving the accuracy of the final fitting curve. In this embodiment, by selecting lenses with known diopters as samples, and then using diopters as intermediate parameters to calculate the fitting curve of deformation distance and optical deformation angle, the experimental process is simple, the workload of calculation is reduced, and the efficiency of online testing of float glass light deformation angle is further improved.

[0073] like Figure 3 As shown, in some embodiments, step S110 includes:

[0074] Step S111: acquiring a calibration image of the lens through a line scan camera;

[0075] Step S112: selecting a number of sample areas at intervals on the calibration image along the length direction, and obtaining deformation distances of the calibration image located in the number of sample areas;

[0076] Step S113: fitting the deformation distances and diopters of several sample areas to generate several initial models;

[0077] Step S114: Fitting a number of initial models to obtain a first functional relationship.

[0078] The calibration image is usually a fringe image, and the deformation distance can be obtained by calculating the distance the fringe boundary moves. Fig. 9 and Fig.10 As shown, the stripes in the calibration image can be expressed as a cosine function, set to F(x), which is a fixed intrinsic value of the light source and does not change; the function formed by the detection is expressed as f(x), which will change according to the detection glass. In order to amplify the change of x, it can be multiplied with another cosine function R(x) with a smaller period, and the final functions are t(x) and T(x):

[0079] t(x)=f(x)·R(x)

[0080] T(x)=F(x)·R(x)

[0081] By comparing t(x) and T(x), the displacement of the stripes in the x direction can be calculated.

[0082] In actual testing, the deformation distance may be very small. At this time, the change of x cannot be accurately detected and calculated because the camera has limited pixels and cannot achieve high-precision segmentation. However, y is the energy value, and a smaller resolution can be obtained through energy control. The phase change of the cosine function is not only reflected in the horizontal axis x, but also in the vertical axis y. Therefore, the phase shift change can also be calculated by calculating the offset of the y value. The specific calculation process is:

[0083] Amplify changes in the y direction by multiplying it by itself,

[0084] G(x)=F(x)·F(x)

[0085] T'(x)=f(x)·F(x)

[0086] By comparing G(x) and T'(x), the offset in the y direction can be obtained.

[0087] Since the y direction is more reflected in small deformations, it is impossible to calculate larger deformations, and x cannot reflect small deformations. We calculate the deformation of x as dx and the deformation of y as dy, and the final deformation distance is expressed as d=dx+dy.

[0088] Due to the existence of lens distortion, the calibration image taken by the line scan camera will produce different degrees of deformation and distortion. Therefore, in this embodiment, the deformation distance is calculated at each location of the calibration image, and then the relationship between the deformation distance and the refractive power of multiple lenses is obtained, that is, the initial model. Then, multiple initial models are fitted to obtain the first functional relationship, so as to achieve correction of lens distortion to a certain extent and improve the accuracy of the first functional relationship.

[0089] For example, the initial model can be established at one sixth, one half, and five sixths of the calibration image, and the first functional relationship can be finally obtained by fitting. Since the image distortion is more serious at the edge, the sample area is set at the edge of the calibration image to further improve the accuracy of the first functional relationship.

[0090] In some embodiments, the plurality of lenses include lenses having refractive powers of 0 dpt, ±20 mdpt, ±40 mdpt, ±60 mdpt, ±80 mdpt, and ±100 mdpt, respectively.

[0091] The positive refractive power represents the refractive power of a convex lens, which is used to simulate a float glass with convex deformation, and the negative refractive power represents the refractive power of a concave lens, which is used to simulate a float glass with concave deformation. Since the deformation degree of float glass is usually small, this embodiment uses a lens within 100mdpt as a sample for testing, so as to better simulate the actual deformation situation, so as to make the fitting curve of the deformation distance and the optical deformation angle more accurate.

[0092] like Figure 4 As shown, in some embodiments, the calibration fitting curve includes:

[0093] Step S310: placing three lenses with diopters of 0 and ±100 mdpt at a detection position, and obtaining deformation distances corresponding to the three lenses;

[0094] Step S320: calibrating the fitting curve according to the deformation distances corresponding to the three lenses.

[0095] Among them, the lens with a refractive power of 0dpt (i.e., flat glass) has no focusing ability and will not cause any optical deformation or deviation. It can be regarded as an ideal reference sample without optical deformation. Therefore, when calibrating the fitting curve, the lens with a refractive power of 0dpt provides a reference point for judging the measurement results of the system when there is no optical deviation, thereby verifying the initial accuracy of the fitting curve and ensuring that the fitting curve can start from zero. The refractive power of ±100mdpt can cover more extreme optical deformation scenarios, and significant optical deformation effects can be detected during testing, thereby obtaining results with obvious optical deformation, ensuring that the fitting curve has reliable reference points in a wider range of refractive powers. At the same time, using lenses with a larger refractive power span as samples for calibrating the fitting curve can more effectively reduce fitting errors, enhance the robustness of the fitting curve, further balance the deviations of the fitting curve in the positive and negative directions, and improve the accuracy of the fitting curve.

[0096] like Figure 5 As shown, in some embodiments, obtaining a target image of the float glass to be tested includes:

[0097] Step S210: collecting the transmitted light and reflected light of the float glass surface to be tested;

[0098] Step S220: generating a target image of the float glass to be tested.

[0099] In this embodiment, the target image of the float glass to be tested is generated and acquired in real time by real-time acquisition of the transmitted light and reflected light on the surface of the float glass to be tested. Compared with only acquiring the transmitted light, the light intensity of the acquired target image is increased, thereby improving the image quality of the acquired target image.

[0100] like Figure 6As shown, in some embodiments, after generating the target image of the float glass to be tested, the method further includes:

[0101] Step S230: acquiring defects on the float glass to be tested according to the target image, and calculating the size of the defects;

[0102] The defects include one or more of cracks, dust, bubbles, inclusions, tin spots and scratches.

[0103] Defects on float glass can be identified based on image processing algorithms. For example, edge detection algorithms (such as Sobel, Canny, etc.) can be used to identify obvious cracks, scratches and other defects on the glass surface; or a defect template (such as an image of float glass with bubbles) can be set in advance, and the image to be tested can be matched with the defect template to identify defects that match the template; the system can also be trained using a large amount of data based on deep learning technologies such as convolutional neural networks to achieve automatic recognition and classification of defects.

[0104] There are also many ways to calculate the defect size. For example, the pixel count of the defect area can be performed to calculate the total number of pixels in the defect area, and then the number of pixels can be converted into the actual length or area according to the image resolution (the number of pixels per unit length, usually in pixels / mm). For example, if it is known that the image contains 100 pixels per mm and the width of a defect area is 50 pixels, then its actual width is 0.5 mm. For irregularly shaped defects (such as cracks and scratches), the size of the defect can be approximately calculated by fitting a minimum circumscribed rectangle.

[0105] This embodiment integrates deformation detection and defect detection of float glass into the same detection process, and can complete accurate detection of float glass in real time on the production line, reducing time cost, allowing users to quickly discover problems in the production process of float glass, and thus improving the yield rate of float glass.

[0106] The present invention further provides a float glass defect detection device, referring to Figure 7 ,include:

[0107] The standard calibration module 100 includes a plurality of lenses with different refractive powers, and is used to obtain the deformation distance and the optical deformation angle of the lens;

[0108] An imaging light source module 200 is used to illuminate the float glass to be tested to generate a target image;

[0109] An image acquisition module 300 is used to acquire a target image;

[0110] The data processing module 400 is connected to the standard calibration module 100 and the image acquisition module 300, and is used to generate and calibrate a fitting curve according to the deformation distance of the lens and the optical deformation angle of the lens, and to obtain the deformation distance in the target image, and substitute the deformation distance in the target image into the fitting curve to obtain the optical deformation angle of the float glass to be tested.

[0111] In this embodiment, the deformation distance of the float glass is detected online through the image acquisition module 300, and the deformation distance is substituted into the fitting curve of the calibrated deformation distance and the optical deformation angle, so that the online detection of the optical deformation angle of the deformation defect on the float glass can be realized, eliminating the step of using a measuring tool such as a zebra angle detector to perform offline detection on the float glass, allowing the user to quickly and accurately obtain the deformation data of the float glass and evaluate the deformation defects, so that the user can adjust the production plan in time and solve the problems arising in the production process.

[0112] In some embodiments, the imaging light source module 200 includes:

[0113] A reflected light source, which can illuminate the float glass to be tested to generate reflected light;

[0114] Transmitted light source, which can emit light of two different wavelengths;

[0115] The filter is arranged between the transmitted light source and the float glass to be tested, and a plurality of light-blocking sections are arranged at intervals on the filter. The light-blocking sections can only allow one of two different wavelengths of light to pass through, so that the two different wavelengths of light form stripe light and uniform backlight after passing through the filter. The stripe light and the uniform backlight can illuminate the float glass to be tested to generate transmitted light.

[0116] In this embodiment, the transmission light source cooperates with the filter so that the light emitted by the transmission light source becomes stripe light and uniform backlight after passing through the filter, wherein the stripe light is used to calculate the deformation distance of the float glass, and the uniform backlight is used for defect detection and defect size calculation of the float glass.

[0117] The filter can be laser defilmed to form spaced light-blocking segments, which are covered with a film layer that allows only a certain monochromatic light to pass through and cuts off light of other wavelengths, while other areas of the filter can pass light of any wavelength.

[0118] Exemplarily, a bicolor LED that can emit light with wavelengths A and B is used as a transmission light source, wherein the two wavelengths of light can be lit separately, so as to form two images using time division multiplexing technology. A filter is covered on the top of the bicolor LED, wherein the light blocking section can cut off the light with wavelength B, so that when the bicolor LED only emits light with wavelength A, a uniform backlight is formed, and when the bicolor LED only emits light with wavelength B, a stripe light is formed.

[0119] like Figure 8 As shown, in some embodiments, the data processing module 400 includes:

[0120] The data preprocessor 410 is connected to the image acquisition module 300 and is used to set the photographing parameters of the image acquisition module 300 and process the target image to generate fringe data;

[0121] The data processor 420 is connected to the standard calibration module 100 and the data preprocessor 410, and is used to receive the deformation distance of the lens and the optical deformation angle of the lens and generate a fitting curve; and the data processor 420 can receive the fringe data, and output the deformation distance of the float glass to be measured according to the fringe data;

[0122] The industrial computer 430 is connected to the data processor 420, and is used to receive the deformation distance of the float glass to be tested, substitute the deformation distance of the float glass to be tested into the fitting curve, and output the optical deformation angle of the float glass to be tested to an external device.

[0123] The data preprocessor 410 may be one or more of a field programmable gate array, a digital signal processor, and a graphics processing unit, and the data processor 420 may be a high-performance central processing unit or a graphics processor. The external device may be a display device, a mobile terminal, etc., so that the user can check the optical deformation angle of the float glass to be tested in time.

[0124] The above description is only a partial or preferred embodiment of the present invention. Neither the text nor the drawings can limit the scope of protection of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the overall concept of the present invention, or direct / indirect application in other related technical fields are included in the scope of protection of the present invention.

Claims

1. A method for detecting defects in float glass, characterized in that: include Obtaining a fitting curve of deformation distance and optical deformation angle; Acquire a target image of the float glass to be tested, and acquire a deformation distance of the float glass to be tested according to the target image; The fitting curve is calibrated, and the deformation distance of the float glass to be measured is substituted into the calibrated fitting curve to obtain the optical deformation angle of the float glass to be measured.

2. The float glass defect detection method according to claim 1, characterized in that: The step of obtaining a fitting curve of deformation distance and optical deformation angle comprises: Placing a plurality of lenses with different refractive powers at a detection position and collecting a calibration image of each lens; obtaining a deformation distance of the lens according to the calibration image, thereby performing data fitting between the refractive power and the deformation distance of the lens to obtain a first functional relationship; Acquiring the optical deformation angle of the lens, thereby performing data fitting on the diopter and the optical deformation angle of the lens to obtain a second functional relationship; A fitting curve of the deformation distance and the optical deformation angle is calculated according to the first functional relationship and the second functional relationship.

3. The float glass defect detection method according to claim 2, characterized in that: Placing a plurality of lenses with different refractive powers at a detection position and collecting a calibration image of each lens; The deformation distance of the lens is obtained according to the calibration image, so as to perform data fitting between the diopter and the deformation distance of the lens, and the first functional relationship is obtained, including: Acquiring a calibration image of the lens by a line scan camera; Selecting a number of sample areas at intervals on the calibration image along the length direction, and obtaining deformation distances of the calibration image located in the number of sample areas; Fitting the deformation distances of the plurality of sample areas with the refractive power to generate a plurality of initial models; The plurality of initial models are fitted to obtain a first functional relationship.

4. The float glass defect detection method according to claim 3, characterized in that: The plurality of lenses include lenses having refractive powers of 0 dpt, ±20 mdpt, ±40 mdpt, ±60 mdpt, ±80 mdpt, and ±100 mdpt, respectively.

5. The float glass defect detection method according to claim 1, characterized in that: The calibration fitting curve comprises: Place three lenses with diopters of 0 and ±100 mdpt at the detection position to obtain the corresponding deformation distances of the three lenses; The fitting curve is calibrated according to the deformation distances corresponding to the three lenses.

6. The float glass defect detection method according to any one of claims 1 to 5, characterized in that: The step of obtaining a target image of the float glass to be tested comprises: Collecting the transmitted light and reflected light from the surface of the float glass to be tested; Generate a target image of the float glass to be tested.

7. The float glass defect detection method according to claim 6, characterized in that: After generating the target image of the float glass to be tested, the method further includes: Acquire defects on the float glass to be tested according to the target image, and calculate the size of the defects; The defects include one or more of cracks, dust, bubbles, inclusions, tin spots and scratches.

8. A float glass defect detection device, characterized in that: include: A standard calibration module, comprising a plurality of lenses with different refractive powers, for obtaining a deformation distance of the lens and an optical deformation angle of the lens; An imaging light source module, used for irradiating the float glass to be tested to generate a target image; An image acquisition module, used for acquiring the target image; A data processing module is connected to the standard calibration module and the image acquisition module, and is used to generate and calibrate a fitting curve according to the deformation distance of the lens and the optical deformation angle of the lens, and to obtain the deformation distance in the target image, and substitute the deformation distance in the target image into the fitting curve to obtain the optical deformation angle of the float glass to be tested.

9. The float glass defect detection device according to claim 8, characterized in that: The imaging light source module comprises: A reflecting light source, wherein the reflecting light source can illuminate the float glass to be tested to generate reflected light; A transmission light source, which can emit light of two different wavelengths; A filter is arranged between the transmitted light source and the float glass to be tested, and a plurality of light-blocking sections are arranged at intervals on the filter, and the light-blocking sections can only allow one of the two different wavelengths of light to pass through, so that the two different wavelengths of light form stripe light and uniform backlight after passing through the filter, and the stripe light and the uniform backlight can illuminate the float glass to be tested to generate transmitted light.

10. The float glass defect detection device according to claim 9, characterized in that: The data processing module comprises: A data preprocessor, connected to the image acquisition module, used to set the photographing parameters of the image acquisition module and process the target image to generate fringe data; A data processor, connected to the standard calibration module and the data preprocessor, for receiving the deformation distance of the lens and the optical deformation angle of the lens and generating the fitting curve; and the data processor can receive the fringe data and output the deformation distance of the float glass to be measured according to the fringe data; The industrial computer is connected to the data processor, and is used for receiving the deformation distance of the float glass to be tested, substituting the deformation distance of the float glass to be tested into the fitting curve, and outputting the optical deformation angle of the float glass to be tested to an external device.

Images