Glass linear optical deformation defect detection method based on stripe background
By projecting stripe patterns on the glass surface and analyzing stripe changes using high-resolution cameras and image processing systems, the problems of strong subjectivity and low efficiency of traditional manual inspection methods are solved, and high-precision, non-contact defects of glass linear optical deformation are achieved. For large-scale production line inspection.
Patent Information
- Application Number
- CN202411097234.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional manual inspection of linear optical deformation defects of glass depends on the operator's experience and visual judgment, and there are problems of strong subjectivity and low efficiency.
Using a detection method based on stripe background, by projecting a stripe pattern with a predetermined frequency and amplitude on the glass surface, combining a high-resolution camera and an image processing system, the changes in the stripe pattern are analyzed, linear optical deformation defects are identified, and the linear optical deformation defects are graphicized. Report or alert output results.
Improve the accuracy and efficiency of detection, reduce the operator's dependence on experience, realize sensitive detection of tiny linear optical deformation, and does not need to directly contact the glass surface, avoid secondary damage or contamination, and is suitable for large-scale production lines. Scale inspection.
Smart Images

Figure CN118883571A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of glass surface defect detection, and in particular relates to a glass linear optical deformation defect detection method based on a stripe background. Background Art
[0002] Glass is manufactured through a process of raw material batching, furnace melting, tin bath molding, and other process steps. Many defects will inevitably appear in the manufacturing process, such as optical deformation, flatness, point defects, etc. Among them, linear optical deformation is a transparent optical deformation defect with a length of about 1-5 mm. Traditional manual inspection methods often rely on the operator's experience and visual judgment, which has the problems of strong subjectivity and low efficiency. In order to improve the accuracy and efficiency of inspection, it is necessary to provide a new glass linear optical deformation defect detection method based on stripe background to solve the above technical problems. Summary of the invention
[0003] The technical problem solved by the present invention is to provide a glass linear optical deformation defect detection method based on a striped background that can improve the accuracy and efficiency of inspection, thereby solving the problem that traditional manual inspection methods often rely on the operator's experience and visual judgment, have strong subjectivity and low efficiency.
[0004] In order to solve the above technical problems, the present invention provides a glass linear optical deformation defect detection method based on stripe background, which comprises the following steps:
[0005] S1. Project a fringe pattern with a predetermined frequency and amplitude onto the surface of the glass to be tested and place it under appropriate lighting conditions to ensure that the fringe pattern is clearly visible;
[0006] S2. Place the glass to be inspected parallel to the striped pattern background at a certain distance, and the operator slowly moves the glass and performs inspection based on the following two methods:
[0007] The first detection method is as follows:
[0008] (1) Preprocessing: Use a high-resolution camera to photograph the glass surface with the stripe pattern projected on it;
[0009] (2) Edge detection: The collected image is input into the image processing system. The system uses the image processing algorithm to analyze the changes in the fringe pattern and identify the fringe distortion caused by the deformation of the glass surface.
[0010] (3) Fringe analysis: Based on the characteristics of fringe distortion, determine whether there are linear optical deformation defects on the glass surface. For example, by quantitatively analyzing the fringe offset and deformation degree, determine the location, nature and severity of the defect.
[0011] (4) Deformation detection: Output the defect detection results in the form of a graphical report or alarm to facilitate the operator to handle. The report should include the location, type and recommended treatment measures of the defect;
[0012] The second detection method is as follows:
[0013] (1) When moving the glass, the operator observes the overlap between the glass surface and the striped background with the naked eye, paying special attention to the projection changes of the striped pattern on the glass surface;
[0014] (2) When there is a linear optical deformation of 1-5 mm on the glass surface, these deformations will change the reflection and refraction paths of light on the glass surface, causing the projection of the fringe pattern on the glass surface to be deformed;
[0015] (3) By carefully observing these deformation features, such as twisting, breaking, misalignment or blurring of stripes, the operator can preliminarily determine whether the glass has deformed properly and its severity.
[0016] As a further solution of the present invention, the stripe pattern can be generated by a laser projector or other stripe generating device, and the stripe pattern should have clear boundaries and high contrast so that it can be clearly displayed during the inspection process.
[0017] As a further solution of the present invention, the parameters of the spacing, direction and density of the stripe pattern should be adjusted and optimized according to the size, shape and expected deformation type of the glass to be inspected.
[0018] As a further solution of the present invention, the camera should be a camera with high frame rate and high resolution to ensure that subtle changes in the stripes can be captured.
[0019] As a further solution of the present invention, the steps of analyzing the input image by the image processing system are as follows:
[0020] S1. De-noising and contrast adjustment are performed on the original image to improve the image quality and make the stripe pattern clearer;
[0021] S2. Use Canny edge detection algorithm to process the stripes in the image;
[0022] S3. Detect the periodicity and phase change of the stripes by Fourier transform frequency domain analysis method, compare the difference between the actual shape of the stripes and the ideal stripes, and thus identify the deformed area;
[0023] S4. According to the distortion of the stripes, a specific algorithm is used to quantify the offset and deformation degree of the stripes to determine whether there is a linear optical deformation defect on the glass surface.
[0024] As a further solution of the present invention, the Canny edge detection algorithm is specifically as follows:
[0025]
[0026] Among them, G x and G y are the gradients of the image in the x and y directions, respectively.
[0027] As a further solution of the present invention, the Canny edge detection algorithm further includes the following:
[0028] (1) Gaussian filtering: used to remove noise;
[0029] (2) Gradient calculation: calculate the gradient magnitude and direction of the image while identifying the edges in the image;
[0030] (3) Non-maximum suppression: Detect local maxima based on the gradient direction while refining the edges;
[0031] (4). Double threshold processing: divide the edges into strong edges and weak edges and determine the true edges.
[0032] As a further solution of the present invention, the analysis of the fringe pattern is to perform Fourier transform on the fringe image to analyze the frequency and phase of the fringe, and the formula is as follows:
[0033]
[0034] Among them, F(μ, ν) is the Fourier transform in the image, and f(x, y) is the original image. Therefore, the periodicity and phase change of the fringes can be detected through Fourier transform:
[0035] Magnitude(F(μ,ν)) and Phase(F(μ,ν)).
[0036] As a further solution of the present invention, the detection of the fringe distortion is the difference between the actual fringe and the ideal fringe. For periodic fringe, the distortion can be quantified by the phase difference, and the formula is as follows:
[0037]
[0038] in, is the phase of the actual fringe, is the phase of the theoretical fringe.
[0039] As a further solution of the present invention, the deformation is detected as follows:
[0040] (1) Quantification of stripe distortion: The offset and deformation of stripes can be quantified by calculating the displacement of stripes:
[0041] Displacement = x measured -x ideal
[0042] Among them, x measured is the displacement of the actual fringe, x ideal is the displacement under ideal conditions;
[0043] (2) Defect identification: Based on the offset and distortion of the stripes, a specific algorithm is used, and the detected defect area and information are output in the form of a graphical report or alarm.
[0044] Compared with the related art, the glass linear optical deformation defect detection method based on stripe background provided by the present invention has the following beneficial effects:
[0045] 1. The present invention improves the detection accuracy of tiny linear optical deformation by utilizing fringe background technology, and the periodicity and uniformity of the fringe make it possible for even tiny deformation to cause obvious fringe distortion, thereby improving the detection sensitivity;
[0046] 2. The present invention adopts the combination of optical detection technology, high frame rate camera and efficient image processing algorithm, so that it does not need to directly contact the glass surface, avoiding the secondary damage or pollution that may be caused by contact detection, and can also realize real-time detection of glass surface deformation, thereby improving accuracy;
[0047] 3. The present invention uses a stripe background to make the linear optical deformation defects on the glass surface more intuitive and easy to see, reducing the operator's reliance on experience. Then, by quickly identifying the deformation characteristics of the stripe pattern, the operator can quickly complete the inspection process, improving the inspection efficiency. The design of the stripe background can be adjusted and optimized according to actual needs to adapt to the inspection of glass products of different sizes, shapes and optical deformation types. Therefore, compared with complex detection equipment and methods, this method has the advantages of low cost and easy implementation, and is suitable for large-scale inspection on the production line. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.
[0049] Figure 1 It is a schematic diagram of the process of the present invention;
[0050] Figure 2 This is a diagram of part of the first detection method in the present invention;
[0051] Figure 3 Schematic diagram of (a) a fringe diagram before deformation and (b) a fringe diagram after deformation in the present invention;
[0052] Figure 4It is a schematic diagram of the deformation amount in the present invention. DETAILED DESCRIPTION
[0053] Please refer to Figures 1 to 4 ,in, Figure 1 It is a schematic diagram of the process of the present invention; Figure 2 This is a diagram of part of the first detection method in the present invention; Figure 3 Schematic diagram of (a) a fringe diagram before deformation and (b) a fringe diagram after deformation in the present invention; Figure 4 The method for detecting glass linear optical deformation defects based on stripe background includes the following steps:
[0054] S1. Project a fringe pattern with a predetermined frequency and amplitude onto the surface of the glass to be tested and place it under appropriate lighting conditions to ensure that the fringe pattern is clearly visible;
[0055] S2. Place the glass to be inspected parallel to the striped pattern background at a certain distance, and the operator slowly moves the glass and performs inspection based on the following two methods:
[0056] The first detection method is as follows:
[0057] (1) Preprocessing: Use a high-resolution camera to photograph the glass surface with the stripe pattern projected on it;
[0058] (2) Edge detection: The collected image is input into the image processing system. The system uses the image processing algorithm to analyze the changes in the fringe pattern and identify the fringe distortion caused by the deformation of the glass surface.
[0059] (3) Fringe analysis: Based on the characteristics of fringe distortion, determine whether there are linear optical deformation defects on the glass surface. For example, by quantitatively analyzing the fringe offset and deformation degree, determine the location, nature and severity of the defect.
[0060] (4) Deformation detection: Output the defect detection results in the form of a graphical report or alarm to facilitate the operator to handle. The report should include the location, type and recommended treatment measures of the defect;
[0061] The second detection method is as follows:
[0062] (1) When moving the glass, the operator observes the overlap between the glass surface and the striped background with the naked eye, paying special attention to the projection changes of the striped pattern on the glass surface;
[0063] (2) When there is a linear optical deformation of 1-5 mm on the glass surface, these deformations will change the reflection and refraction paths of light on the glass surface, causing the projection of the fringe pattern on the glass surface to be deformed;
[0064] (3) By carefully observing these deformation features, such as twisting, breaking, misalignment or blurring of stripes, the operator can preliminarily determine whether the glass has deformed properly and its severity.
[0065] The fringe pattern may be generated by a laser projector or other fringe generating device, and the fringe pattern should have clear boundaries and high contrast so as to be clearly displayed during the inspection process.
[0066] The parameters of the spacing, direction and density of the stripe pattern should be adjusted and optimized according to the size, shape and expected deformation type of the glass to be inspected.
[0067] The camera should be selected with high frame rate and high resolution capabilities to ensure that subtle changes in the fringes can be captured.
[0068] The steps of analyzing the input image by the image processing system are as follows:
[0069] S1. De-noising and contrast adjustment are performed on the original image to improve the image quality and make the stripe pattern clearer;
[0070] S2. Use Canny edge detection algorithm to process the stripes in the image;
[0071] S3. Detect the periodicity and phase change of the stripes by Fourier transform frequency domain analysis method, compare the difference between the actual shape of the stripes and the ideal stripes, and thus identify the deformed area;
[0072] S4. According to the distortion of the stripes, a specific algorithm is used to quantify the offset and deformation degree of the stripes to determine whether there is a linear optical deformation defect on the glass surface.
[0073] The Canny edge detection algorithm is as follows:
[0074]
[0075] Among them, G x and G y are the gradients of the image in the x and y directions, respectively.
[0076] The Canny edge detection algorithm also includes the following:
[0077] (1) Gaussian filtering: used to remove noise;
[0078] (2) Gradient calculation: calculate the gradient magnitude and direction of the image while identifying the edges in the image;
[0079] (3) Non-maximum suppression: Detect local maxima based on the gradient direction while refining the edges;
[0080] (4). Double threshold processing: divide the edges into strong edges and weak edges and determine the true edges.
[0081] The analysis of the fringe pattern is to perform Fourier transform on the fringe image to analyze the frequency and phase of the fringe, and the formula is as follows:
[0082]
[0083] Among them, F(μ, ν) is the Fourier transform in the image, and f(x, y) is the original image. Therefore, the periodicity and phase change of the fringes can be detected through Fourier transform:
[0084] Magnitude(F(μ,ν)) and Phase(F(μ,ν)).
[0085] The detection of the fringe distortion is the difference between the actual fringe and the ideal fringe. For periodic fringe, the distortion can be quantified by the phase difference, as shown in the following formula:
[0086]
[0087] in, is the phase of the actual fringe, is the phase of the theoretical fringe.
[0088] The deformation is detected as follows:
[0089] (1) Quantification of stripe distortion: The offset and deformation of stripes can be quantified by calculating the displacement of stripes:
[0090] Displacement = x measured -x ideal
[0091] Among them, x measured is the displacement of the actual fringe, x ideal is the displacement under ideal conditions;
[0092] (2) Defect identification: Based on the offset and distortion of the stripes, a specific algorithm is used, and the detected defect area and information are output in the form of a graphical report or alarm.
[0093] The advantages of the first detection method are as follows:
[0094] (1) High precision: The fringe background technology is used to improve the detection accuracy of small linear optical deformations. The periodicity and uniformity of the fringe make it possible for even small deformations to cause obvious fringe distortion, thereby improving the detection sensitivity.
[0095] (2) Non-contact detection: This method uses optical detection technology, which does not require direct contact with the glass surface, avoiding secondary damage or contamination that may be caused by contact detection;
[0096] (3) Strong real-time performance: High frame rate camera and efficient image processing algorithm are used to achieve real-time detection of glass surface deformation;
[0097] (4) High degree of automation: This method can be integrated with the automated inspection system to achieve fully automatic defect detection, improve inspection efficiency, and reduce operational errors.
[0098] The advantages of the second detection method are as follows:
[0099] (1) Intuitiveness: The use of striped background makes the linear optical deformation defects on the glass surface more intuitive and easy to see, reducing the operator's dependence on experience;
[0100] (2) Efficiency: By quickly identifying the deformation characteristics of the stripe pattern, the operator can quickly complete the inspection process and improve inspection efficiency;
[0101] (3) Flexibility: The design of the striped background can be adjusted and optimized according to actual needs to adapt to the inspection of glass products of different sizes, shapes and optical deformation types;
[0102] (4) Cost-effectiveness: Compared with complex testing equipment and methods, this method has the advantages of low cost and easy implementation, and is suitable for large-scale inspection on the production line.
[0103] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A method for detecting glass linear optical deformation defects based on stripe background, characterized in that: The following steps are involved: S1. Project a fringe pattern with a predetermined frequency and amplitude onto the surface of the glass to be tested and place it under appropriate lighting conditions to ensure that the fringe pattern is clearly visible; S2. Place the glass to be inspected parallel to the striped pattern background at a certain distance, and the operator slowly moves the glass and performs inspection based on the following two methods: The first detection method is as follows: (1) Preprocessing: Use a high-resolution camera to photograph the glass surface with the stripe pattern projected on it; (2) Edge detection: The collected image is input into the image processing system. The system uses the image processing algorithm to analyze the changes in the fringe pattern and identify the fringe distortion caused by the deformation of the glass surface. (3) Stripe analysis: Based on the characteristics of stripe distortion, determine whether there are linear optical deformation defects on the glass surface. For example, by quantitatively analyzing the stripe offset and deformation degree, determine the location, nature and severity of the defect. (4) Deformation detection: Output the defect detection results in the form of a graphical report or alarm to facilitate the operator to handle. The report should include the location, type and recommended treatment measures of the defect; The second detection method is as follows: (1) When moving the glass, the operator observes the overlap between the glass surface and the striped background with the naked eye, paying special attention to the projection changes of the striped pattern on the glass surface; (2) When there is a linear optical deformation of 1-5 mm on the glass surface, these deformations will change the reflection and refraction paths of light on the glass surface, causing the projection of the fringe pattern on the glass surface to be deformed; (3) By carefully observing these deformation features, such as twisting, breaking, misalignment or blurring of stripes, the operator can preliminarily determine whether the glass has deformed properly and its severity.
2. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The fringe pattern may be generated by a laser projector or other fringe generating device, and the fringe pattern should have clear boundaries and high contrast so as to be clearly visible during the inspection process.
3. The glass linear optical deformation defect detection method based on stripe background according to claim 1, characterized in that: The parameters of the spacing, direction and density of the stripe pattern should be adjusted and optimized according to the size, shape and expected deformation type of the glass to be inspected.
4. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The camera should be selected with high frame rate and high resolution capabilities to ensure that subtle changes in the fringes can be captured.
5. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The steps of analyzing the input image by the image processing system are as follows: S1. De-noising and contrast adjustment are performed on the original image to improve the image quality and make the stripe pattern clearer; S2. Use Canny edge detection algorithm to process the stripes in the image; S3. Detect the periodicity and phase change of the stripes by Fourier transform frequency domain analysis method, compare the difference between the actual shape of the stripes and the ideal stripes, and thus identify the deformed area; S4. According to the distortion of the stripes, a specific algorithm is used to quantify the offset and deformation degree of the stripes to determine whether there is a linear optical deformation defect on the glass surface.
6. The glass linear optical deformation defect detection method based on stripe background according to claim 1, characterized in that: The Canny edge detection algorithm is as follows: in, and G y are the gradients of the image in the x and y directions, respectively.
7. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The Canny edge detection algorithm also includes the following: (1) Gaussian filtering: used to remove noise; (2) Gradient calculation: calculate the gradient magnitude and direction of the image while identifying the edges in the image; (3) Non-maximum suppression: Detect local maxima based on the gradient direction while refining the edges; (4). Double threshold processing: divide the edges into strong edges and weak edges and determine the true edges.
8. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The analysis of the fringe pattern is to perform Fourier transform on the fringe image to analyze the frequency and phase of the fringe, and the formula is as follows: where F(μ, ν) is the Fourier transform of the image, is the original image, so through Fourier transform, the periodicity and phase change of the fringes can be detected: Magnitude(F(μ,ν)) and Phase(F(μ,ν)).
9. The method for detecting glass linear optical deformation defects based on stripe background according to claim 1, characterized in that: The detection of the fringe distortion is the difference between the actual fringe and the ideal fringe. For periodic fringe, the distortion can be quantified by the phase difference, as shown in the following formula: in, is the phase of the actual fringe, is the phase of the theoretical fringe.
10. The glass linear optical deformation defect detection method based on stripe background according to claim 1, characterized in that: The deformation is detected as follows: (1) Quantification of stripe distortion: The offset and deformation of stripes can be quantified by calculating the displacement of stripes: in, is the displacement of the actual fringe, is the displacement under ideal conditions; (2) Defect identification: Based on the offset and distortion of the stripes, a specific algorithm is used, and the detected defect area and information are output in the form of a graphical report or alarm.
Citation Information
Patent Citations
Method of detecting glass defects based on phase image processing
CN103344651A
Glass defect detection method
CN107024488A
Method for extracting and identifying surface defects of transparent body
CN108665458A
Omnibearing stereoscopic vision detection method for workpiece with bright surface and application
CN115060198A
Device for detecting flatness of glass
CN204128517U
