A method and system for quickly measuring the thickness of optical glass

Abstract

The present application relates to the technical field of optical glass thickness measurement, and specifically to a method and system for rapid measurement of optical glass thickness, the method comprising: obtaining a beam image; using an edge detection algorithm to obtain the edge of two beam connected domains, and using an edge refinement algorithm to obtain the edge skeleton of the two beam connected domains; then obtaining the light offset value of each beam image; calculating the distortion influence coefficient of each beam image; then obtaining the distortion offset factor of each beam image, and measuring the thickness of the optical glass. The present application improves the accuracy of measuring the thickness of curved optical glass such as optical glass.

Description

Technical Field

[0001] The present application relates to the technical field of optical glass thickness measurement, and in particular to a method and system for rapid measurement of optical glass thickness. Background Art

[0002] Optical glass is a transparent material with specific optical properties, which is widely used in optical instruments, precision measuring equipment and other fields. With the development of science and technology, the demand for accurate measurement of optical glass thickness is increasing. Optical glass not only affects the quality of optical imaging, but also affects the service life and stability of optical instruments. The application of rapid measurement technology can shorten the product development cycle, reduce production costs, and improve product quality, thereby promoting the development of the optical industry. At the same time, high-precision measurement technology also provides strong support for scientific research and technological innovation. The rapid measurement technology of optical glass thickness can not only improve production efficiency, but also is crucial to ensure the stability and reliability of optical system performance. With the advancement of optical technology, non-contact measurement methods have gradually become mainstream, such as laser interferometry, optical microscope measurement, etc. These methods have the advantages of fast measurement speed, high accuracy, and non-destructive testing.

[0003] The existing optical glass thickness measurement is mainly carried out through rapid detection through optical principles, mainly using the linear laser and plane mirror imaging principle, by precisely controlling the laser beam to form imaging features on the glass surface, and then the optical system captures and converts them into analog signals. The computer vision system further processes these signals, extracts image features and calculates the thickness value. The system includes optical design, image acquisition and processing software, which can realize online measurement. This high-precision measurement technology not only improves the quality of optical products, but also provides important support for optical research and precision engineering. However, when optical glass with curved surfaces such as optical glass, concave lenses, etc. are continuously imaged, due to the change in its surface curvature, the focusing points of light at different positions will be different, resulting in imaging distortion and optical distortion. The existing technology cannot effectively solve the problem of inaccurate measurement results caused by optical distortion. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a method and system for quickly measuring the thickness of optical glass. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for quickly measuring the thickness of optical glass, the method comprising the following steps:

[0006] Acquire the light beam image of the optical glass at each acquisition moment;

[0007] An edge detection algorithm is used to obtain the edge of the connected domain of two beams, and an edge refinement algorithm is used to obtain the edge skeleton of the connected domain of two beams;

[0008] Based on the differences in shape and length between edge skeletons of connected domains of each light beam in the light beam image, a light offset value of each light beam image is obtained;

[0009] Obtaining the distortion influence coefficient of each beam image based on the similarity between the beam image and the beam images at adjacent acquisition moments before and after, and the similarity between edge skeletons;

[0010] Obtaining a distortion offset factor of each light beam image based on the light offset value and the distortion influence coefficient;

[0011] Optical glass thickness is measured based on a distortion offset factor.

[0012] Furthermore, the method for obtaining the light offset value is:

[0013] For the beam image at each acquisition moment, the threshold segmentation algorithm is used to divide the beam image to obtain the beam area, and the clustering algorithm is used to divide the beam area to obtain two beam connected domains;

[0014] Obtaining a length difference value between the connected domains of the light beams in each light beam image based on the edge length difference between the connected domains of the light beams;

[0015] Acquire shape difference values ​​between connected domains of the light beams in each light beam image based on shape differences between connected domains of the light beams;

[0016] The product of the length difference value and the shape difference value is taken as the light deviation value of each light beam image.

[0017] Furthermore, the method for obtaining the shape difference value is:

[0018] Compute the normalized first-order difference chain code of the edge skeleton of the connected domains of two beams;

[0019] Calculate the absolute value of the difference between the elements of the same position sequence in the normalized first-order difference chain code of each edge skeleton, and compose the difference chain code of each beam image according to the position sequence;

[0020] For the difference chain code of each light beam image, the frequency of each non-zero number in the difference chain code is counted as the first frequency of each non-zero number, the product between the value of each non-zero number and its first frequency is calculated as the first product of each non-zero number, and the average of all the first products in the difference chain code of each light beam image is taken as the shape difference value between the connected domains of the light beams.

[0021] Further, the method for obtaining the length difference value is:

[0022] For each beam image, the absolute value of the difference in length between the edge skeletons of two beam connected domains in the beam image is taken as the length difference value between the beam connected domains.

[0023] Furthermore, the method for obtaining the distortion influence coefficient is:

[0024] Based on the structural similarity between the beam image and the beam images at adjacent acquisition moments, a front similarity index and a rear similarity index of each beam image are obtained;

[0025] Based on the similarity between the edge skeletons of the beam image and the beam images at adjacent acquisition moments before and after, obtaining a first similarity mean value of each edge skeleton in each beam image;

[0026] For each beam image, the distortion influence coefficient of each beam image is obtained based on the front similarity index, the back similarity index and the first similarity mean of each edge skeleton. The formula is: d=x×α; where d is the distortion influence coefficient of the beam image, x is the mean of the front similarity index and the back adjacent index of the beam image, and α is the mean of the first similarity means of all edge skeletons in the beam image.

[0027] Furthermore, the method for obtaining the front similarity index and the back similarity index is:

[0028] For the beam image at each acquisition moment, the beam image at the previous adjacent acquisition moment is used as its previous adjacent beam image, and the beam image at the next adjacent acquisition moment is used as its next adjacent beam image;

[0029] The structural similarity index algorithm is used to calculate the structural similarity index between the beam image at each acquisition moment and its preceding adjacent beam image as the preceding similarity index, and the structural similarity index between the beam image and its succeeding adjacent beam image as the succeeding similarity index.

[0030] Further, the method for obtaining the first similarity mean is:

[0031] For the beam image at each acquisition moment, calculate the similarity between the normalized first-order difference chain code of each edge skeleton in the beam image and the normalized first-order difference chain code of all edge skeletons in the previous adjacent beam image, and obtain the maximum value of the similarity as the first maximum similarity of each edge skeleton;

[0032] Calculate the similarity between the normalized first-order difference chain code of each edge skeleton in the beam image and the normalized first-order difference chain code of all edge skeletons in the subsequent adjacent beam images, and obtain the maximum value of the similarity as the second maximum similarity of each edge skeleton;

[0033] The mean of the first maximum similarity and the second maximum similarity of each edge skeleton in the beam image at each acquisition moment is calculated as the first similarity mean of each edge skeleton.

[0034] Furthermore, the method for obtaining the distortion offset factor is:

[0035] For the beam image at each acquisition moment, the difference between the light offset value of the beam image and its previous adjacent beam image is calculated as the first difference, and the difference between the light offset value of the beam image and its subsequent adjacent beam image is calculated as the second difference, and the average of the first difference and the second difference is used as the distortion coefficient of the beam image at each acquisition moment;

[0036] The formula of the distortion offset factor is: B=c×d; where B is the distortion offset factor of the beam image; c is the distortion degree coefficient of the beam image, and d is the distortion influence coefficient of the beam image.

[0037] Further, the measuring of the optical glass thickness based on the distortion offset factor includes:

[0038] Obtaining the incident angle of the light beam, the beam spacing, the refractive index of the optical glass, and the magnification of the lens of the light beam image at each acquisition moment;

[0039] The corrected beam spacing is obtained based on the beam spacing, and the formula is: l ’ =ε×B×l, where l ’ To correct the beam spacing, B is the distortion offset factor of the beam image at each acquisition moment, l is the beam spacing, and ε is the preset scaling factor;

[0040] The thickness of the optical glass is calculated based on the corrected beam spacing, the incident angle of the beam, the refractive index of the optical glass, and the magnification of the lens. The formula is: D = (n 2 -sin 2 θ) 1 / 2 / 2×(1- sin 2 θ) 1 / 2 ×β×l ’ ; Where D is the thickness of the optical glass, n is the refractive index of the optical glass, θ is the incident angle, β is the magnification of the lens, l ’ To correct the beam spacing.

[0041] In a second aspect, an embodiment of the present application further provides a system for rapid measurement of optical glass thickness, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0042] This application has at least the following beneficial effects:

[0043] The present application proposes a method and system for quickly measuring the thickness of optical glass. Aiming at the problem of beam shape deviation caused by optical distortion, the Otsu threshold algorithm and chain code method are used to analyze the change of beam shape caused by tangential distortion when the beam passes through the optical glass, thereby solving the problem of deviation between the beam shape and position and the actual existence due to optical distortion. Aiming at the problem of cumulative distortion effect in the process of continuous image acquisition, the structural similarity index and cosine similarity are used to reflect the comprehensive change of beam path and shape caused by optical distortion in the process of continuous image acquisition, thereby eliminating the interference of cumulative distortion effect on measurement accuracy. By calculating the incident angle and beam spacing, the beam spacing is corrected in combination with the distortion offset factor, and the distortion offset factor is scaled by the scaling factor, thereby eliminating the influence caused by the curvature of the optical glass surface, achieving higher-precision glass thickness measurement, and improving the accuracy of thickness measurement of curved optical glass such as optical glass. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A flowchart of a method for quickly measuring the thickness of optical glass provided in one embodiment of the present application;

[0046] Figure 2 A schematic diagram of a collection device provided in one embodiment of the present application;

[0047] Figure 3 A flowchart for obtaining a distortion offset factor is provided for one embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the optical glass thickness rapid measurement method and system proposed in the present application, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0049] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0050] The specific scheme of a method and system for quickly measuring the thickness of optical glass provided by the present application is described in detail below with reference to the accompanying drawings.

[0051] See also Figure 1 , which shows a flowchart of a method for quickly measuring the thickness of optical glass provided by an embodiment of the present application, the method comprising the following steps:

[0052] Step S1, capturing the image of the light beam reflected and refracted at different positions of the optical glass by a CCD camera, and preprocessing the image.

[0053] This application uses a reflection method to measure the thickness of optical glass. The specific measuring device and process are as follows: select a helium-neon laser with a small divergence angle and low noise, emit a light source to illuminate the optical glass, place a magnifying lens in front of a CCD camera to amplify the reflected and refracted light beams, and record the magnification of the magnifying lens according to the lens production parameters, and use a CCD camera to receive the reflected light and refracted light, and use an image acquisition card to transmit the signal to a computer. The reflection method for measuring glass thickness is a technology well known to people in this field and will not be elaborated here.

[0054] Select optical glass and obtain the refractive index of the optical glass according to its production parameters. Install a uniform and slow-moving conveyor belt at the optical glass placement location to ensure that the glass can move slowly after being placed on the conveyor belt. Set the conveyor belt moving speed to 1mm / s and the interval of the image collected by the CCD camera to 1s. After placing the glass on the conveyor belt and maintaining it in a stationary state, start the conveyor belt, laser and CCD camera at the same time to obtain the beam image collected by the CCD camera. The schematic diagram of the acquisition device is shown in the figure below. Figure 2 shown.

[0055] Figure 2 In the figure, T1 is a He-Ne laser, T2 is a CCD camera, T3 is a lens, T4 is an optical glass, T5 is a conveyor belt, E and G are the positions of the reflected light beam and the refracted light beam after being magnified by the lens and reaching the CCD camera, and the dotted line is the normal line.

[0056] After the measuring device is installed, a second CCD camera is installed on the side of the measuring device to ensure that the second CCD camera is parallel to the laser path emitted by the helium-neon laser and that the second CCD camera can fully capture the complete path of the laser emitted by the laser emitter. In addition, the start-up time and acquisition interval of the second CCD camera are consistent with the above.

[0057] Step S2, determining the beam offset caused by tangential distortion by the difference in shape between the incident beam and the reflected beam, and determining the accumulated tangential distortion generated during continuous acquisition of the optical glass by combining the difference in beam shape between images acquired at adjacent acquisition moments.

[0058] When the light beam passes through a lens or a reflective surface, the light fails to focus correctly on the imaging plane due to the change in curvature of the glass thickness to be tested, resulting in tangential distortion. This distortion manifests itself as an unexpected offset of the straight line portion of the image edge. Specifically, the tangential distortion causes the edge of the object to shift along the tangent direction. The offset causes the actual position of the object to be inconsistent with the imaging position, thereby affecting the accuracy of the measurement.

[0059] For the beam image at each acquisition moment, the Otsu threshold algorithm is used to obtain the optimal segmentation threshold of the beam image, and the grayscale value of the pixel point whose grayscale value is greater than the optimal segmentation threshold is set to 255, and the grayscale value of the pixel point whose grayscale value is less than or equal to the optimal segmentation threshold is set to 0, and a binary image of the beam image is obtained to eliminate the influence of scattered light caused by the part of the light that may be scattered when the beam encounters the surface of the object. The area composed of pixels with a grayscale value of 255 in the binary image is used as the beam area.

[0060] Furthermore, the k-means clustering algorithm is used to divide the beam area, k is set to 2, and the Euclidean distance between the pixels is used as the metric distance to obtain the connected domains of the two beams. Among them, the Otsu threshold algorithm and the k-means clustering algorithm are well-known technologies and will not be described in detail here.

[0061] For two beam connected domains, the two beam connected domain regions are respectively used as input, the canny edge detection algorithm is used to obtain the edges of the two beam connected domains, and the ZhangSuen edge thinning algorithm is used to thin all edges to obtain the edge skeletons of the two beam connected domains. Among them, the canny edge detection algorithm and the ZhangSuen edge thinning algorithm are well-known technologies and are not described in detail here.

[0062] Furthermore, for the edge skeletons of the two beam connected domains, each edge skeleton is input, and the normalized first-order differential chain code of each edge skeleton is output using the eight-neighborhood Freeman chain code method. The normalized first-order differential chain code can represent the local direction change of the edge skeleton, and each number represents a direction change. Among them, the eight-neighborhood Freeman chain code method is a well-known technology and will not be described in detail here.

[0063] The absolute values ​​of the differences between the elements of the same bit sequence in the normalized first-order difference chain code of each edge skeleton are calculated, and the difference chain code of each beam image is composed according to the bit sequence. It should be noted that if the lengths of the two normalized first-order difference chain codes are different, zero padding is performed at the end of the shorter normalized first-order difference chain code.

[0064] For the difference chain code of each light beam image, the frequency of each non-zero number in the difference chain code is counted as the first frequency of each non-zero number, the product between the value of each non-zero number and its first frequency is calculated as the first product of each non-zero number, and the average of all the first products in the difference chain code of each light beam image is taken as the shape difference value between the connected domains of the light beams.

[0065] For each beam image, the absolute value of the difference in length between the edge skeletons of two beam connected domains in the beam image is taken as the length difference value between the beam connected domains.

[0066] Since the incident angle of the light beam is the same, the light beams after refraction and reflection through the optical glass show similar characteristics in shape. For each light beam image, the light offset value A of the light beam image is calculated based on the difference in edge length between two light beam connected domains in the light beam image and the difference in edge shape between the two light beam connected domains. The calculation formula is: A=a×b; where A is the light offset value of the light beam image; a is the length difference value between the light beam connected domains in the light beam image, and b is the shape difference value between the light beam connected domains in the light beam image.

[0067] In the formula, the length difference value a reflects the change in the beam path caused by the different refractive indices on the upper and lower surfaces of the glass and the tangential distortion when the light beam passes through the optical glass. The refractive index of the glass does not change much within the thickness variation range, so the impact on the beam path is very small. Therefore, when the length difference value a is larger, it means that the change in the beam path caused by the tangential distortion is greater, and the light offset value A of the beam image increases accordingly; the shape difference value b reflects the change in the beam shape caused by the tangential distortion when the light beam passes through the optical glass. The larger its value is, the greater the difference in shape between the two edge skeletons, the greater the irregularity on the glass surface, and the value of the light offset value A will also increase accordingly.

[0068] The light deviation value A describes the comprehensive change of the light beam caused by tangential distortion when the light beam passes through optical glass. The larger the value, the stronger the influence of tangential distortion on the light beam image.

[0069] When the CCD camera continuously captures the light beam, the continuously collected images will amplify the effect of tangential distortion. As the object moves on the conveyor belt, the focus point of the light at different positions changes, which will cause the straight line part of the edge of the image to accumulate unexpected deviations. This continuous deviation is accumulated in multiple images, making the distortion effect more obvious, thus affecting the accuracy of the measurement.

[0070] For the beam image at each acquisition moment, the beam image at the previous adjacent acquisition moment is used as its previous adjacent beam image, and the beam image at the next adjacent acquisition moment is used as its next adjacent beam image.

[0071] The structural similarity index (SSIM) algorithm was used to calculate the structural similarity index between the beam image at each acquisition moment and its preceding adjacent beam image as the preceding similarity index, and the structural similarity index between the beam image and its succeeding adjacent beam image as the succeeding similarity index.

[0072] For the beam image at each acquisition moment, the normalized first-order difference chain code of each edge skeleton in the beam image is calculated, and the similarity between it and the normalized first-order difference chain code of all edge skeletons in its previous adjacent beam image is calculated, and the maximum value of the similarity is obtained as the first maximum similarity of each edge skeleton; similarly, the normalized first-order difference chain code of each edge skeleton in the beam image is calculated, and the similarity between it and the normalized first-order difference chain code of all edge skeletons in its subsequent adjacent beam image is calculated, and the maximum value of the similarity is obtained as the second maximum similarity of each edge skeleton.

[0073] The mean of the first maximum similarity and the second maximum similarity of each edge skeleton in the beam image at each acquisition moment is calculated as the first similarity mean of each edge skeleton.

[0074] In this embodiment, cosine similarity is selected as the similarity measurement method, and the implementer can select other measurement methods according to actual conditions.

[0075] For the beam image at each acquisition moment, the distortion influence coefficient of each beam image is obtained based on the front similarity index, the back similarity index of the beam image and the first similarity mean of each edge skeleton. The method is: d=x×α; where d is the distortion influence coefficient of the beam image, x is the mean of the front similarity index and the back adjacent index of the beam image, and α is the mean of the first similarity means of all edge skeletons in the beam image.

[0076] It should be noted that x reflects the degree of structural change between the beam image and its two adjacent images due to optical distortion, α reflects the degree of beam edge change between the beam image and its two adjacent beam images, and the distortion influence coefficient d reflects the beam deviation caused by the accumulation of distortion effects caused by the change of the focal point of light at different positions during continuous image acquisition; when x is larger, the greater the structural change caused by optical distortion, and when α is larger, the greater the degree of beam edge change, the more obvious the beam deviation is, and the larger the value of the distortion influence coefficient d is.

[0077] Furthermore, the distortion coefficient c between the beam image and the beam images at adjacent acquisition moments is obtained by:

[0078] For the light beam image at each acquisition moment, the difference between the light offset value of the light beam image and its previous adjacent light beam image is calculated as the first difference, and the difference between the light offset value of the light beam image and its subsequent adjacent light beam image is calculated as the second difference. The average of the first difference and the second difference is taken as the distortion coefficient of the light beam image at each acquisition moment.

[0079] Furthermore, based on the distortion coefficient c and the distortion influence coefficient d between the beam image and the beam image at the adjacent acquisition time, the distortion offset factor B between the two beams of the beam image is calculated, and the formula is: B=c×d; where B is the distortion offset factor of the beam image; c is the distortion coefficient of the beam image, and d is the distortion influence coefficient of the beam image. The flowchart for obtaining the distortion offset factor is as follows: Figure 3 shown.

[0080] The distortion coefficient c reflects the light offset value caused by the change in curvature of the glass surface during the light beam image acquisition process. The larger the value of the distortion coefficient c, the greater the change in the light offset value between adjacent images, that is, the greater the cumulative degree of distortion, and the larger the distortion offset factor B; the larger the distortion influence coefficient d, the more significant the distortion effect in the light beam image, and the distortion offset factor B increases accordingly.

[0081] Among them, the distortion offset factor B describes the comprehensive changes in the beam path and shape caused by optical distortion during the continuous image acquisition process. The larger the value of the distortion offset factor B is, the stronger the optical distortion effect on the beam image is, and the more serious the impact on the accuracy of the measurement result is.

[0082] At this point, the distortion offset factor of the beam image at each acquisition moment is obtained.

[0083] Step S3, correcting the distance between the two light beams by the distortion offset factor between the two light beams of the light beam image, thereby offsetting the influence of the tangential distortion and achieving more accurate optical glass thickness detection.

[0084] For the beam image, the canny edge detection algorithm is used to obtain the edge image of the beam image of the second CCD camera at each acquisition time, the Hough transform contour tracking algorithm is used to process the edge image, the path of the beam is extracted, and the Harris corner point detection algorithm is used to output the corner point of the beam path. The corner point is the intersection between the beam and the lens, and the paths on both sides of the corner point are the incident light and the reflected light, respectively. The vector dot product is used to calculate the angle between the incident light and the reflection, and half of the angle is used as the incident angle of the beam. The method for obtaining the incident angle is a well-known technology. This embodiment provides the above method, and the implementer can select other incident angle calculation methods according to actual conditions.

[0085] However, due to the influence of the section distortion, there is a deviation between the beam spacing received by the CCD camera and the actual spacing. According to the distortion offset factor B between the two beams of the beam image, the measured beam spacing is corrected. The correction formula is l ’ =ε×B×l, where l ’ To correct the beam spacing, B is the distortion offset factor of the beam image at each acquisition moment, l is the beam spacing, and ε is the preset scaling factor, which is 3 in this embodiment. The beam spacing of the beam image at each acquisition moment is the distance between the fitting straight lines of the two edge skeletons, and the calculation method is a well-known technology and will not be described here; in this embodiment, Euclidean distance is selected as the distance measurement method, and the implementer can select other measurement methods according to actual conditions.

[0086] By scaling the distortion offset factor B between the two beams of the light beam image by ε, the influence caused by the curvature of the optical glass surface can be eliminated. When the curvature of the optical glass surface is larger, it means that the edge of the lens is steeper. When the light passes through the edge of the lens, the refraction angle will be larger, causing the focal point of the light to deviate from the ideal position, resulting in more obvious tangential distortion. At this time, the implementer can adjust the scaling factor ε to offset the influence of different curvatures.

[0087] Furthermore, based on the incident angle of the light beam, the refractive index and magnification of the lens, and the corrected beam spacing, the thickness of the optical glass is calculated. The thickness calculation method is a well-known technique, and the specific formula is as follows: D = (n 2 -sin 2 θ) 1 / 2 / 2×(1- sin 2 θ) 1 / 2 ×β×l ’ ; Where D is the thickness of the optical glass, n is the refractive index of the optical glass, θ is the incident angle, β is the magnification of the lens, l ’ To correct the beam spacing.

[0088] Thus, the thickness measurement of optical glass is completed.

[0089] Based on the same inventive concept as the above method, an embodiment of the present application also provides a system for rapid measurement of optical glass thickness, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of any one of the above-mentioned methods for rapid measurement of optical glass thickness are implemented.

[0090] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0091] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0092] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for rapid measurement of optical glass thickness, characterized in that: The method comprises the following steps: Acquire the light beam image of the optical glass at each acquisition moment; An edge detection algorithm is used to obtain the edge of the connected domain of two beams, and an edge refinement algorithm is used to obtain the edge skeleton of the connected domain of two beams; Based on the differences in shape and length between edge skeletons of connected domains of each light beam in the light beam image, a light offset value of each light beam image is obtained; Obtaining the distortion influence coefficient of each beam image based on the structural similarity between the beam image and the beam images at adjacent acquisition moments before and after, and the similarity between edge skeletons; Obtaining the distortion offset factor of each light beam image based on the light offset value and the distortion influence coefficient; and measuring the thickness of the optical glass based on the distortion offset factor; The method for obtaining the distortion influence coefficient is: Based on the structural similarity between the beam image and the beam images at adjacent acquisition moments, a front similarity index and a rear similarity index of each beam image are obtained; Based on the similarity between the edge skeletons of the beam image and the beam images at adjacent acquisition moments before and after, obtaining a first similarity mean value of each edge skeleton in each beam image; For each beam image, based on the front similarity index, the back similarity index and the first similarity mean of each edge skeleton, the distortion influence coefficient of each beam image is obtained, and the formula is: d=x×α; where d is the distortion influence coefficient of the beam image, x is the mean of the front similarity index and the back adjacent index of the beam image, and α is the mean of the first similarity means of all edge skeletons in the beam image; The method for obtaining the front similarity index and the back similarity index is: For the beam image at each acquisition moment, the beam image at the previous adjacent acquisition moment is used as its previous adjacent beam image, and the beam image at the next adjacent acquisition moment is used as its next adjacent beam image; The structural similarity index algorithm is used to calculate the structural similarity index between the beam image at each acquisition moment and its preceding adjacent beam image as the preceding similarity index, and the structural similarity index between the beam image and its succeeding adjacent beam image as the succeeding similarity index; The method for obtaining the first similarity mean is: For the beam image at each acquisition moment, calculate the similarity between the normalized first-order difference chain code of each edge skeleton in the beam image and the normalized first-order difference chain code of all edge skeletons in the previous adjacent beam image, and obtain the maximum value of the similarity as the first maximum similarity of each edge skeleton; Calculate the similarity between the normalized first-order difference chain code of each edge skeleton in the beam image and the normalized first-order difference chain code of all edge skeletons in the subsequent adjacent beam images, and obtain the maximum value of the similarity as the second maximum similarity of each edge skeleton; Calculate the mean of the first maximum similarity and the second maximum similarity of each edge skeleton in the beam image at each acquisition moment as the first similarity mean of each edge skeleton; The method for obtaining the distortion offset factor is: For the beam image at each acquisition moment, the difference between the light offset value of the beam image and its previous adjacent beam image is calculated as the first difference, and the difference between the light offset value of the beam image and its subsequent adjacent beam image is calculated as the second difference, and the average of the first difference and the second difference is used as the distortion coefficient of the beam image at each acquisition moment; The formula of the distortion offset factor is: B=c×d; where B is the distortion offset factor of the beam image; c is the distortion degree coefficient of the beam image, and d is the distortion influence coefficient of the beam image; The method of measuring the thickness of optical glass based on the distortion offset factor includes: Obtaining the incident angle of the light beam, the beam spacing, the refractive index of the optical glass, and the magnification of the lens of the light beam image at each acquisition moment; The corrected beam spacing is obtained based on the beam spacing, and the formula is: l ’ =ε×B×l, where l ’ To correct the beam spacing, B is the distortion offset factor of the beam image at each acquisition moment, l is the beam spacing, and ε is the preset scaling factor; The thickness of the optical glass is calculated based on the corrected beam spacing, the incident angle of the beam, the refractive index of the optical glass, and the magnification of the lens. The formula is D=(n 2 -sin 2 θ) 1 / 2 / 2×(1- sin 2 θ) 1 / 2 ×β×l ’ ; Where D is the thickness of the optical glass, n is the refractive index of the optical glass, θ is the incident angle, β is the magnification of the lens, l ’ To correct the beam spacing.

2. A method for rapid measurement of optical glass thickness as claimed in claim 1, characterized in that: The method for obtaining the light offset value is: For the beam image at each acquisition moment, the threshold segmentation algorithm is used to divide the beam image to obtain the beam area, and the clustering algorithm is used to divide the beam area to obtain two beam connected domains; Obtaining a length difference value between the connected domains of the light beams in each light beam image based on the edge length difference between the connected domains of the light beams; Acquire shape difference values ​​between connected domains of light beams in each light beam image based on shape differences between connected domains of light beams; The product of the length difference value and the shape difference value is taken as the light deviation value of each light beam image.

3. A method for rapid measurement of optical glass thickness as claimed in claim 2, characterized in that: The method for obtaining the shape difference value is: Compute the normalized first-order difference chain code of the edge skeleton of the connected domains of two beams; Calculate the absolute value of the difference between the elements of the same position sequence in the normalized first-order difference chain code of each edge skeleton, and compose the difference chain code of each beam image according to the position sequence; For the difference chain code of each light beam image, the frequency of each non-zero number in the difference chain code is counted as the first frequency of each non-zero number, the product between the value of each non-zero number and its first frequency is calculated as the first product of each non-zero number, and the average of all the first products in the difference chain code of each light beam image is taken as the shape difference value between the connected domains of the light beams.

4. A method for rapid measurement of optical glass thickness as claimed in claim 3, characterized in that: The method for obtaining the length difference value is: For each beam image, the absolute value of the difference in length between the edge skeletons of two beam connected domains in the beam image is taken as the length difference value between the beam connected domains.

5. A system for rapid measurement of optical glass thickness, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method for quickly measuring the thickness of optical glass as described in any one of claims 1 to 4 are implemented.

Images