Method and system for detecting optical defects in glass windshields - Patents.com

A computer-implemented method and system analyze angular distances and ghost images to detect narrow optical defects in glass windshields, addressing the undetectability of threads by conventional systems and ensuring compliance with optical quality standards.

JP2024540925AInactive Publication Date: 2024-11-06SAINT GOBAIN VITRAGE SA
View PDF 10 Cites 0 Cited by

Patent Information

Application Number
JP2024523557
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-10-22
Filing Date
2022-10-20
Publication Date
2024-11-06
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Current inspection systems fail to detect special optical defects called threads in glass windshields, which are narrow and oriented along the longitudinal axis, often remaining undetected by production line inspections but visible to drivers, leading to potential customer rejection and production losses.

Method used

A computer-implemented method and system that utilizes a digital image recording device to capture and process images of a periodic pattern through the windshield, analyzing angular distances and ghost images to identify signal artifacts and shape parameters, enabling detection of optical defects like threads.

Benefits of technology

The method effectively detects optical defects such as threads, which are undetectable by conventional systems, while requiring minimal adaptation to existing manufacturing processes, and uses available equipment for image acquisition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024540925000001
    Figure 2024540925000001
  • Figure 2024540925000002
    Figure 2024540925000002
  • Figure 2024540925000003
    Figure 2024540925000003
Patent Text Reader

Abstract

A novel, easy-to-implement method is provided for detecting optical defects in glass windshields, and in particular, a special type of optical defect called a thread that may not be detected by current inspection systems but remains visible to a human driver. [Solution] The system takes as input an angular distance map between a primary image and a ghost image of a periodic pattern (1003) viewed through a glass windshield (1002), an image of the periodic pattern viewed through the glass windshield, and at least one shape parameter of individual features of the periodic pattern, and provides as output an image of optical defects in the glass windshield (1002).
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to methods and systems for detecting optical defects in glass windshields. [Background technology]

[0002] Glass windshields are widely used, for example, in the automotive, rail transportation and aviation transportation industries. They are the glazing through which the driver sees what is ahead, such as the road, the railway, the scenery, etc. They are usually made of two curved inorganic glass sheets laminated together with a polymer interlayer.

[0003] For safety reasons, distortion of objects when viewed through the windshield must be as small as possible, or at least must not be so distracting to the driver. In this regard, the optical quality of the windshield must meet certain requirements detailed in paragraph 9.2 of Annex 3 to Article 43 of the United Nations Economic Commission for Europe (UN / ECE) Regulations.

[0004] In the art several methods and instruments are described which may be intended to measure the optical distortion of the windshield within the framework of Article 43 of the Agreement Regulation.

[0005] US Patent Application Publication No. 2009 / 282871 describes an apparatus and method for measuring the transmitted optical distortion of a glass sheet from a dot-matrix image transmitted through said glass sheet.

[0006] WO 2014 / 191401 describes a method for determining the refractive power of a windshield from the geometric deformations that occur in an image of a line grating acquired through said windshield.

[0007] WO 2010 / 037958 describes an apparatus and method for analysing the optical distortion of a windshield from a phase shift measured on a grating image acquired through said windshield.

[0008] WO 2009 / 065740 describes a deflectometry system and method for the optical inspection of a phase and amplitude object placed in front of a periodic pattern, which is based on a two-step Fourier transform and inverse Fourier transform of an image of said periodic pattern through said object.

[0009] EP 0 463 940 A1 describes a process for measuring the optical quality of a windshield based on the illuminance of the shadows.

[0010] WO 98 / 17993, GB 2152210 and EP 1061357 describe methods for measuring the optical distortion of a windshield through image analysis of the transmission or reflection pattern.

[0011] WO 2017 / 008159 describes a method for detecting optical defects in a windshield through analysis of a composite image of chromatic aberration.

[0012] WO 2018 / 024550 describes an apparatus and method for determining the angular distance between the positions of primary and secondary images of multiple point-like light sources viewed through a windshield. [Prior art documents] [Patent documents]

[0013] [Patent Document 1] US Patent Application Publication No. 2009 / 282871 [Patent Document 2] International Publication No. 2014 / 191401 [Patent Document 3] International Publication No. 2010 / 037958 [Patent Document 4] International Publication No. 2009 / 065740 [Patent Document 5] European Patent Application Publication No. 0463940 [Patent Document 6] International Publication No. 98 / 17993 [Patent Document 7] GB 2152210 [Patent Document 8] European Patent Application Publication No. 1061357 [Patent Document 9] International Publication No. 2017 / 008159 [Patent Document 10] International Publication No. 2018 / 024550 Summary of the Invention [Problem to be solved by the invention]

[0014] No matter how effective they are, methods described in the art may fail to detect one particular optical defect called a thread, which may resemble what is called a ream defect in the glass industry, but is much thinner and often oriented substantially along the longitudinal axis of the windshield.

[0015] These special optical defects may only be visible to the driver and / or passengers at certain angles or different viewing angles, and may be missed entirely by production line inspection systems.

[0016] In particular, deflectometry-based systems and methods may fail to detect these defects because they may not leave a signature in the overall windshield optical distortion map.

[0017] In particular, the sled appears to be able to disrupt electromagnetic signals picked up by current inspection systems such that the inspection systems no longer detect the signals. The inspection system is then unable to distinguish whether this lack of signal is due to an actual optical defect or due to some of the elements that often decorate the windshield, such as gaskets, masking enamels, etc., or even masking marks, such as ink marks, that are made on the surface of the windshield to activate the inspection system.

[0018] The immediate adverse effect is that a windshield that was originally deemed to meet technical specifications may later be rejected by the customer. At that time, claims and production losses may occur.

[0019] In this regard, there is a need for novel, easy-to-implement methods for detecting optical defects in glass windshields, particularly special optical defects called sleds that may not be detected by current inspection systems but remain visible to a human driver. [Means for solving the problem]

[0020] In a first aspect of the present disclosure there is provided a computer implemented method for detecting optical defects in a windshield, as set forth in claim 1, the dependent claims being advantageous embodiments.

[0021] In a second aspect of the present invention there is provided a data processing apparatus, a computer program and a computer readable medium for implementing the method of the first aspect as set out in claims 6 to 8.

[0022] In a third aspect of the present invention there is provided a process for detecting optical defects in a glass windshield as claimed in claim 9.

[0023] In a fourth aspect of the present invention there is provided an inspection system for detecting optical defects in a glass windshield, as described in claim 10, the dependent claims being advantageous embodiments.

[0024] The methods, processes and inspection systems according to the present disclosure may be used during the manufacturing process of windshields. Effect of the Invention

[0025] A first outstanding benefit of the present invention is that it enables the detection of optical defects, particularly threads, in glass windshields that may remain undetected by most common inspection systems but may still be visible to the human eye, e.g., the driver and / or passengers.

[0026] A second advantage is that the present invention can be implemented relatively easily into existing manufacturing processes, requiring little, if any, adaptation. More precisely, computer-implemented inventions and processes according to the present invention can benefit from equipment already available in the manufacturing line and / or quality control processes for acquiring digital image maps of optical power. [Brief description of the drawings]

[0027] [Figure 1] FIG. 1 is a schematic diagram of an inspection system for acquiring images of a periodic pattern through a glass windshield. [Diagram 2] 1 is an example of a superposition of primary and ghost images of individual features of a periodic pattern acquired through a windshield. [Diagram 3] An example of the superposition of primary and ghost images of distorted individual features of a periodic pattern acquired through a windshield. [Figure 4] FIG. 2 is a data flow diagram of a computer-implemented method for detecting optical defects in a glass windshield according to an embodiment of the first aspect of the present invention. [Diagram 5] 4 is an example of an image map of optical imperfections in a windshield provided as output by the method according to the first aspect of the invention; [Figure 6] 2 is a data processing system according to an embodiment of the second aspect of the present invention; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0028] In the context of this disclosure, "signal artifact" should be understood according to its general definition in the field of signal processing. A signal artifact can be understood as an unexpected, unwanted, or accidental modification of a signal that represents activity or phenomena other than that intended or expected. For example, it may be a signal gap or an outlier, ie a part of a signal that differs significantly from the rest of the signal.

[0029] Referring to FIG. 1, an inspection system 1000 for detecting optical defects in a glass windshield includes a digital image recording device 1001 positioned behind a glass windshield 1002. The digital image recording device is configured to capture an image of a periodic pattern 1003, which is positioned in front of and viewed through the windshield 1002. Once the image is acquired, it is processed to find the optical distortions caused by the windshield when an electromagnetic signal, e.g. light, passes through the windshield.

[0030] Most processing methods rely on a comparison of individual features 1003 a - n of the periodic pattern 1003 in the acquired image, ie the image as viewed through the windshield 1002 , with individual features of the original periodic pattern 1003 . This comparison makes it possible to detect individual features 1003a-n that appear distorted and to calculate the associated optical parameters, e.g. phase shift, optical power, etc., to generate an optical distortion map or angular distance map of the windshield 1002.

[0031] Current inspection equipment and methods are unable to detect a particular optical defect in glass windshields called a thread. The threads may resemble ream defects, but are much thinner and often oriented substantially along the longitudinal axis (B) of the windshield 1002. It appears that the thread can interfere with incident electromagnetic signals, e.g., light, that are later acquired by the digital image recording device 1001, so that, for example, angular distances are not calculated at that position and no optical distortions are then reported.

[0032] The immediate negative impact is that current inspection systems and methods can be confused by their inability to identify the cause of a missing signal and often attribute the cause to other elements decorating the windshield that are unrelated to optical defects, such as threads.

[0033] As an illustrative example, referring to FIG. 2, the angular distance at one position on the windshield can be calculated from the distance between an individual feature of a periodic pattern 1003, e.g., 1003a, which appears at 2001, on a primary image of the pattern 1003 as viewed through the windshield 1002, and the same individual feature in a ghost image of the pattern, e.g., 1003a, which appears at a shifted 2002. Angular distance calculations are performed for each individual feature 1003a-n of the periodic pattern 1003 to generate an angular distance map for all or a portion of the windshield.

[0034] Ghost images are a well-known phenomenon in optics that arise from the internal reflection of light between surfaces of an optical system, such as a windshield, causing the object to appear twice with a spatial displacement that depends on the refractive index of the optical system and the angle at which the object is viewed.

[0035] Referring to FIG. 3, optical imperfections in the windshield 1002, such as threads, can disrupt the optical path such that ghost images of individual features of the periodic pattern 1003, such as 1003a, go undetected while the individual features, such as 1003a, remain highly distorted 3001. Such signal artifacts, e.g., non-detection or outliers of ghost images, may result from the optical system, e.g., windshield, being assumed to be perfect due to the absence of ghost images, from the misalignment between the primary and ghost images being too small to resolve overlapping non-shifted and shifted features, or from the misalignment being too large to resolve the shifted features outside the field of view of the acquisition system.

[0036] The immediate adverse effect is that the angular distance at the location of the optical defect approaches zero. Thus, the angle map shows no signs of any defects, while the solid is still distorted, clearly indicating the presence of optical defects.

[0037] In this regard, and with reference to FIGS. 1-4, in a first aspect of the present invention, a computer-implemented method 4000 for detecting optical defects in a glass windshield 1002 is provided, the method comprising: The method takes as input an angular distance map I4001 between a primary image and a ghost image of a periodic pattern 1003 as viewed through the glass windshield, an image I4002 of the periodic pattern 1003 as viewed through the glass windshield 1002, and at least one shape parameter I4003 of individual features 1003a-1003n of the periodic pattern 1003; The method 4000 provides as output an image O2001 of the optical defect; The method 4000 includes the following steps. (a) a step S4001 of searching for signal artifacts in the input angular distance map I4001; (b) a step S4002 of searching for individual features 1003a-1003n of the periodic pattern 1003 within zones of an input image I4002 of the periodic pattern 1003, where said zones coincide with the zones of signal artifacts searched for in step (a); (c) Step S4003 of defining the shape of each of the retrieved individual features 1003a-1003n by image processing; (d) a step S4004 of comparing the input shape parameters I4003 of the individual features 1003a-1003n of the periodic pattern 1003 with the corresponding shape parameters of the defined shape of step (c); (e) a step S4005 of generating an image from an input image I4002 of the periodic pattern 1003, where the input shape parameters I4003 of individual features 1003a-1003n of the periodic pattern 1003 do not match the corresponding shape parameters of a defined shape, and said generated image is an image of an optical defect.

[0038] In a preferred embodiment, in step S4001, the signal artifacts looked for are signal outliers or signal non-detections of ghost images in the angular distance map I4001. A non-signal detection or signal outlier may be due to the absence of ghost images such that the optical system, e.g., windshield, may be considered perfect, or the shift between the primary and ghost images may be too small to resolve overlapping non-shifted and shifted features, or the shift may be so large that the shifted feature is outside the field of view of the acquisition system.

[0039] The shape parameters I4003 of the individual features 1003a-1003n and the parameters of the corresponding defined shapes may be any morphological parameters for measuring or quantifying at least one morphological characteristic of the individual features and the defined shapes.

[0040] Thus, in certain embodiments, the shape parameter I4003 is selected from area, circularity, area envelopment, minor axis, major axis, perimeter, Feret diameter, or a combination thereof.

[0041] In step S4003, the definition operation can be any definition operation available in the field of image processing, such as image segmentation. For example, it can be thresholding segmentation, watershed segmentation, or random walker segmentation.

[0042] In step S4005, and referring to FIG. 5 as an illustrative example, an image map O2001 of the optical defect 5001 is generated from the defined area of ​​step S4003. In FIG. 5, the detected glass defect 5001 is a so-called thread, which is an elongated glass defect oriented along the longitudinal axis (B) of the windshield. While the method according to the first aspect of the present invention is capable of detecting such fine glass defects, it is unable to detect non-optically transparent glass defects, such as decorative elements 5002a-c, such as ink marks, enamel, gaskets, etc., made on the surface of the glass windshield.

[0043] In a preferred embodiment, the periodic pattern 1003 may be a dot matrix, as shown in FIG. In a more preferred embodiment, the dots of the dot matrix may be point light sources, for example point LEDs. The most accurate results, especially for elongated optical defects such as threads, can be obtained with a point light source.

[0044] In a second aspect of the present invention, referring to FIG. 6, there is provided a data processing system 6000 comprising means for performing the method 4000 according to any one of the embodiments of the first aspect of the present invention, and a computer program I6001 comprising instructions which, when executed by a computer, cause the computer to perform the method according to any one of the embodiments of the first aspect of the present invention.

[0045] The data processing system 6000 comprises means 6001 for performing the method according to any of the embodiments of the first and second aspects of the invention. An example of a means for performing a method 6001 may be a device that can be instructed to automatically perform a sequence of arithmetic or logical operations to perform a task or action. Such devices, also referred to as computers, may comprise one or more central processing units (CPUs) and at least one controller adapted to perform these operations.

[0046] It may further comprise other electronic components such as an input / output interface 6003, a non-volatile or volatile storage device 6002, and a bus, which is a communication system for transferring data between components within a computer or between computers. One of the input / output devices may be a user interface for human-machine interaction, for example a graphical user interface for displaying information that can be understood by humans.

[0047] Because calculations can require a lot of computing power to process significant amounts of data, the data processing system may advantageously be equipped with one or more graphics processing units (GPUs), whose parallel structure makes them more efficient than CPUs, especially for image processing in ray tracing.

[0048] With regard to the computer program I6001, any kind of programming language, compiled or interpreted, can be used to implement the steps of the method of the invention. The computer program I6001 may be part of a software solution, ie a collection of executable instructions, code, scripts or the like and / or part of a database.

[0049] In certain embodiments, a computer readable storage device or medium 6002 is also provided that includes instructions that, when executed by a computer, cause the computer to perform a method according to any of the embodiments of the first aspect of the invention.

[0050] The computer readable storage device 6002 is preferably a non-volatile storage device or memory, such as a hard disk drive or solid state drive. The computer readable storage device may be a removable storage medium or a non-removable storage medium that is part of the computer.

[0051] Alternatively, the computer readable storage device may be a volatile memory in a removable medium.

[0052] The computer readable storage device 6002 may be part of a computer used as a server from which executable instructions can be downloaded which, when executed by the computer, cause the computer to perform a method according to any of the embodiments described herein.

[0053] Alternatively, program I6001 may be implemented in a distributed computing environment, such as cloud computing. The instructions may be executed on a server to which client computers can connect to provide encoded data as input to the method. Once the data has been processed, the output may be downloaded and encoded to a client computer or may be transmitted directly, for example as instructions. This type of implementation can be advantageously realizable in distributed computing environments, such as cloud computing solutions.

[0054] In a third aspect of the present invention, a process for detecting optical defects in a glass windshield 1002 is provided, the process comprising the following steps. (a) providing a glass windshield 1002; (b) using an inspection system 1000 to acquire an angular distance map I4001 between a primary image and a ghost image of a periodic pattern 1003 viewed through the glass windshield 1002, and an image I4002 of the periodic pattern 1003 viewed through the glass windshield 1002; (c) defining at least one shape parameter (I4003) of individual features (1003a-1003n) of the periodic pattern (1003); (d) performing a method according to any embodiment of the first aspect of the invention, in which as inputs said angular distance map (I4001), said image (I4002) of a periodic pattern (1003) and at least one shape parameter (I4003) are provided.

[0055] The technical aspects and features of the various embodiments detailed within the context of the first and second aspects of the invention may also be applied to the third aspect of the invention. It is within the skill of the art to modify, convert or adapt them in the process according to the third aspect of the invention.

[0056] In a fourth aspect of the present invention, and referring to FIGS. 1 to 6, there is provided an inspection system 1000 for detecting optical defects in a glass windshield, said inspection system 1000 comprising: a periodic pattern 1003, preferably a dot matrix of point light sources, an acquisition device 1001 configured to acquire an angular distance map I4001 between a primary image and a ghost image of a periodic pattern 1003 seen through said glass windshield 1002 and an image I4002 of the periodic pattern 1003 seen through said glass windshield 1002; A data processing system 6000 comprising means for performing the method 4000 according to any embodiment of the first aspect of the invention.

[0057] The acquisition module 1001 may be a digital image recording device such as a matrix camera, for example a CCD or CMOS camera system.

[0058] In one particular embodiment, the inspection system may further include a display module, such as a digital display screen, for displaying an image map of the optical defects.

[0059] The methods, processes and inspection systems according to the present disclosure may be advantageously used in the manufacturing process of windshields.

[0060] All embodiments described herein, whether relating to the first, second, third or fourth aspect of the invention, may be combined by the person skilled in the art, unless they are deemed technically incompatible.

[0061] Furthermore, while the present invention has been described in connection with preferred embodiments, it will be understood that various modifications, additions and variations may be made thereto by those skilled in the art without departing from the spirit and scope of the invention as defined in the claims. [Explanation of symbols]

[0062] 1000 Inspection Systems 1001 Digital image recording device 1002 Glass Windshield 1003 Periodic Pattern 4000 Computer Implementation Methods 6000 Data Processing System

Claims

1. 1. A computer-implemented method (4000) for detecting optical defects in a glass windshield (1002), comprising: The method takes as input an angular distance map (I4001) between a primary image and a ghost image of a periodic pattern (1003) viewed through the glass windshield, an image (I4002) of the periodic pattern (1003) viewed through the glass windshield (1002), and at least one shape parameter (I4003) of individual features (1003a-1003n) of the periodic pattern (1003); The method (4000) provides as output an image (O2001) of the optical defect; The method (4000) comprises the steps of: (a) searching for signal artifacts (S4001) in the input angular distance map (I4001); (b) searching (S4002) for individual features (1003a-1003n) of the periodic pattern (1003) within zones of an input image (I4002) of the periodic pattern (1003), where said zones coincide with the zones of signal artifacts searched for in step (a); (c) defining the shape of each of the retrieved individual features (1003a-1003n) by image processing (S4003); (d) comparing (S4004) the input shape parameters (I4003) of the individual features (1003a-1003n) of the periodic pattern (1003) with the corresponding shape parameters of the defined shapes of step (c); (e) generating an image (S4005) from an input image (I4002) of a periodic pattern (1003), where the input shape parameters (I4003) of individual features (1003a-1003n) of the periodic pattern (1003) do not match the corresponding shape parameters of a defined shape, said generated image being an image of an optical defect; The method according to claim 1,

2. 2. The computer-implemented method (4000) of claim 1, wherein the signal artifacts sought are signal outliers or signal non-detections of ghost images in the angular distance map (I4001).

3. 3. The computer-implemented method of claim 1 or 2, wherein the shape parameter (I4003) is selected from area, circularity, area envelopment, minor axis, major axis, perimeter, Feret diameter, or a combination thereof.

4. The computer-implemented method (4000) of any one of claims 1 to 3, wherein the periodic pattern (1003) is a dot matrix.

5. 5. The computer-implemented method (4000) of claim 4, wherein the dots of the dot matrix are point-like light sources.

6. A data processing system (6000) comprising means for performing the method according to any one of claims 1 to 5.

7. A computer program (I6001) comprising instructions which cause a computer to carry out the method according to any one of claims 1 to 5 when said program is executed by a computer.

8. A computer readable medium (6002) comprising instructions that, when said instructions of a program are executed by a computer, cause the computer to perform the method according to any one of claims 1 to 5.

9. A process for detecting optical defects in a glass windshield (1002), said process comprising the steps of: (a) providing a glass windshield (1002); (b) using an inspection system (1000) to obtain an angular distance map (I4001) between a primary image and a ghost image of a periodic pattern (1003) viewed through said glass windshield (1002) and an image (I4002) of the periodic pattern (1003) viewed through said glass windshield (1002); (c) defining at least one shape parameter (I4003) of the individual features (1003a-1003n) of the periodic pattern (1003); (d) implementing the method according to any one of claims 1 to 5, in which the angular distance map (I4001), the image (I4002) of a periodic pattern (1003) and at least one shape parameter (I4003) are provided as input, A process having

10. 1. An inspection system (1000) for detecting optical defects in a glass windshield, the inspection system (1000) comprising: a periodic pattern (1003), preferably a dot matrix of point light sources; an acquisition device (1001) configured to acquire an angular distance map (I4001) between a primary image and a ghost image of a periodic pattern (1003) seen through said glass windshield (1002) and an image (I4002) of the periodic pattern (1003) seen through said glass windshield (1002); A data processing system (6000) according to claim 6, An inspection system comprising:

11. The inspection system of claim 10 further comprising a display module for displaying an image map of the optical defects.

Citation Information

Patent Citations

  • Process and apparatus for measuring the optical quality of a glass sheet

    EP0463940A1

  • Method and apparatus for determining reflective optical quality

    EP1061357A2

  • Distortion inspection apparatus for a glass sheet

    GB2152210A

  • Method and apparatus for measuring transmitted optical distortion in glass sheets

    US20090282871A1

  • Method for detecting optical errors in large surface panels

    WO1998017993A2