A method and system for evaluating the quality of a car glass screen in combination with multispectral imaging

Abstract

The application discloses a kind of automobile glass screen quality evaluation method and system combined with multispectral imaging, the method comprises the following steps: collecting glass screen image of different wave band and carrying out spectral response calibration, obtain calibrated multispectral data cube;Based on the calibrated multispectral data cube, the spectral transmittance distribution characteristics of each wave band in the region divided on glass screen are calculated, and spectral transmission characteristic curve is constructed;Based on the spatial frequency distribution of spectral transmittance, the optical uniformity evaluation characteristics of each wave band on the surface of glass screen are calculated, and optical uniformity evaluation index is constructed;By calculating the spectral reflectance distribution of each wave band, the spectral characteristic value of the region divided on glass screen is obtained and material characteristic fingerprint matrix is constructed, the similarity of to-be-measured product and standard sample on material characteristic fingerprint is calculated;Based on the similarity of spectral transmission characteristic curve, optical uniformity evaluation index and material characteristic fingerprint, the quality grade of glass screen is multidimensional comprehensive score.

Description

Technical Field

[0001] This invention relates to the field of quality assessment technology, and in particular to a method and system for assessing the quality of automotive glass screens that incorporates multispectral imaging. Background Technology

[0002] With the rapid development of the automotive industry and the continuous advancement of intelligent driving technology, automotive glass screens not only bear basic protective functions but also need to meet the optical requirements of intelligent functions such as head-up displays and augmented reality navigation. As an important component of the automotive system, the quality of the glass screen directly affects driving safety and the effectiveness of intelligent functions. Therefore, a comprehensive and accurate quality assessment of automotive glass screens is of great significance.

[0003] Currently, quality inspection of automotive glass screens mainly includes manual visual inspection, single optical inspection, and machine vision inspection. While manual visual inspection can leverage the experience of inspectors, it suffers from low efficiency, inconsistent standards, and susceptibility to subjective factors. Single optical inspection, although capable of automation, offers only one dimension of inspection, making it difficult to comprehensively assess the overall performance of the glass screen. While commonly used machine vision inspection methods possess a certain level of intelligence, they primarily focus on identifying surface defects, with limited ability to evaluate the glass's material properties and optical performance.

[0004] Existing patent CN202510088582.1 discloses a detection method after glass polishing. Although this method uses multi-source sensor data and feature extraction technology, it still targets defects on the glass surface and fails to fully consider the optical characteristics of the glass screen in different spectral bands. At the same time, the feature extraction and region segmentation process of this method is computationally complex, lacks real-time performance, and lacks in-depth analysis of the intrinsic properties of the material, making it difficult to meet the comprehensive quality assessment needs of modern automotive glass screens.

[0005] Furthermore, existing testing methods generally suffer from the following problems: First, they lack systematic analysis of the multi-band spectral characteristics of glass screens, making it impossible to comprehensively evaluate their optical performance; second, the means of assessing material consistency are limited, making it difficult to effectively monitor the quality stability of different batches of products; third, the evaluation results lack quantifiable standards, making it difficult to provide clear guidance for production optimization and quality improvement. Therefore, given the rapid development of intelligent driving technology, traditional testing methods are insufficient to meet the increasingly stringent quality requirements for automotive glass screens in terms of optical performance and material properties.

[0006] In conclusion, there is an urgent need in the industry to develop a testing method that can comprehensively evaluate the optical performance and material properties of automotive glass screens, in order to meet the pressing needs of the modern automotive industry for all-round quality monitoring of glass screens. Summary of the Invention

[0007] In view of this, the present invention provides a technical solution for quality assessment of automotive glass screens, aiming to solve the technical problems of existing glass screen quality inspection, such as single dimension, incomplete assessment results, and general level of intelligence. By comprehensively assessing the optical performance and material properties of the glass screen, it provides a scientific basis for product quality control.

[0008] To achieve the above objectives, this invention discloses a method for evaluating the quality of automotive glass screens using multispectral imaging, comprising the following steps:

[0009] S1: Acquire images of the glass screen in different bands and perform spectral response calibration to obtain a calibrated multispectral data cube;

[0010] S2: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution characteristics of each band in the divided regions on the glass screen, and construct the spectral transmittance characteristic curves.

[0011] S3: Based on the spatial frequency distribution of spectral transmittance, calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, and construct an optical uniformity evaluation index.

[0012] S4: Based on the projection position of the light source and the intensity of the incident light, calculate the spectral reflectance distribution of each band on the surface of the glass screen, and then obtain the spectral characteristic values ​​of the divided regions on the glass screen. By weighting the contributions of different bands, construct the material property fingerprint matrix. By calculating the Euclidean distance between the material property fingerprint matrices of the product under test and the standard sample, obtain the similarity of the material property fingerprints.

[0013] S5: Based on the similarity of spectral transmission characteristic curves, optical uniformity evaluation indicators, and material characteristic fingerprints, the quality level of the glass screen is comprehensively scored from multiple dimensions.

[0014] Preferably, step S1 includes the following steps:

[0015] S11: Data acquisition using a multispectral camera Images of the glass screen at different spectral bands were used to obtain the original multispectral image sequence. ,in For the first Images of the glass screen acquired in each band. For band number, ;

[0016] S12: Perform spectral response calibration on the original multispectral image sequence to obtain the calibrated multispectral data cube. Specifically:

[0017] ;

[0018] in, and These are the horizontal and vertical coordinates of the image on the glass screen, respectively. For the first Images after band calibration for In pixel coordinates Pixel value at that location, For the first Images of the glass screen acquired in each band In pixel coordinates Pixel value at that location, and These are the mean and variance of the standard spectral response, respectively. and The first In each band image, pixel coordinates Center Mean and variance of pixel values ​​within the range It is a natural constant.

[0019] Preferably, step S2 includes the following steps:

[0020] S21: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution of each band on the glass screen surface, specifically:

[0021] ;

[0022] in, For the first Each band in pixel coordinates Spectral transmittance at that location For reference, the whiteboard is in the first... bands, pixel coordinates The reflectance value at that location, For the incident light at the th bands, pixel coordinates Strength at that location, For standard incident light intensity, This is the transmittance correction factor;

[0023] S22: Calculate the spectral transmittance distribution characteristics of the divided regions on the glass screen, specifically:

[0024] ;

[0025] in, For the first The band in the first Spectral transmittance distribution characteristics of each region For area code, , For the total number of regions, and The first The range of x and y coordinates for each region pixel coordinates Spatial weight function at the location, These are weight parameters;

[0026] S23: Construct the spectral transmission characteristic curve, specifically as follows:

[0027] ;

[0028] in, For the first Spectral transmission characteristic curve values ​​for each band, For the first Spectral smoothing function for each band, and These are the center band and bandwidth parameters of the smoothing function, respectively.

[0029] Preferably, step S3 includes the following steps:

[0030] S31: Calculate the spatial frequency distribution of spectral transmittance, specifically:

[0031] ;

[0032] in, For the first Two-dimensional Fourier transform results for each band. for The zero-mean result, and These are the frequency coordinates in the horizontal and vertical directions, respectively. and These represent the number of rows and columns of the image on the glass screen, respectively. The imaginary unit;

[0033] S32: Calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, specifically:

[0034] ;

[0035] in, For the first Optical uniformity evaluation characteristics for each band. For the frequency analysis window size, For frequency weighting function, and These are the frequency weighting parameters for the horizontal and vertical directions, respectively;

[0036] S33: Construct an optical uniformity evaluation index, specifically:

[0037] ;

[0038] in, It serves as an evaluation index for the optical uniformity of the glass screen surface.

[0039] Preferably, step S4 includes the following steps:

[0040] S41: Calculate the spectral reflectance distribution of each band on the glass screen surface, specifically:

[0041] ;

[0042] in, For the first Each band in pixel coordinates Spectral reflectance at that location For the first Each band in pixel coordinates The intensity of reflected light at that location For the first Each band in pixel coordinates The intensity of incident light at that point and These are the horizontal and vertical coordinates of the center of the light source's projection on the glass screen, respectively. The vertical distance from the light source to the glass screen;

[0043] S42: Calculate the spectral characteristic values ​​of the divided regions on the glass screen, specifically:

[0044] ;

[0045] in, For the first The region in the first Spectral characteristic values ​​of each band, For the first The set of pixel coordinates contained in a region;

[0046] S43: Constructing a material property fingerprint matrix Specifically:

[0047] ;

[0048] in, For the first The region in the first Material property fingerprint values ​​for each wavelength band;

[0049] S44: Calculate the similarity of material property fingerprints:

[0050] ;

[0051] in, The similarity of material property fingerprints, The material property fingerprint matrix of the standard sample. For the standard sample The region in the first Material property fingerprint values ​​for each wavelength band.

[0052] Preferably, step S5 includes the following steps:

[0053] S51: Calculate the spectral transmission characteristic score, specifically:

[0054] ;

[0055] in, For the standard sample at the 1st Spectral transmission characteristic curve values ​​for each band, The weighting coefficients for spectral transmission characteristics. For the standard sample at the 1st The mean spectral transmittance at all pixel coordinates in each band;

[0056] S52: Calculate the optical uniformity score, specifically:

[0057] ;

[0058] in, The optical uniformity of standard samples is used as an evaluation index. This is the optical uniformity weighting coefficient;

[0059] S53: Calculate the material property parameter score, specifically:

[0060] ;

[0061] in, These are the weighting coefficients for material property parameters;

[0062] S54: Calculate the overall score And determine the quality level:

[0063] ;

[0064] when When, it is judged as Grade A; when When, it is judged as Grade B; when When, it is judged as Grade C; when If so, it is deemed unqualified.

[0065] This invention also discloses a quality assessment system for automotive glass screens that combines multispectral imaging, comprising:

[0066] Data acquisition and calibration module: Acquires glass screen images of different spectral bands and performs spectral response calibration to obtain a calibrated multispectral data cube;

[0067] Spectral transmission characteristics analysis module: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution characteristics of each band in the divided regions on the glass screen, and construct spectral transmission characteristic curves;

[0068] Optical uniformity analysis module: Based on the spatial frequency distribution of spectral transmittance, calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, and construct optical uniformity evaluation index.

[0069] Material consistency analysis module: Based on the projection position of the light source and the intensity of the incident light, the spectral reflectance distribution of each band on the glass screen surface is calculated, and then the spectral characteristic values ​​of the divided regions on the glass screen are obtained. By weighting the contributions of different bands, a material property fingerprint matrix is ​​constructed. By calculating the Euclidean distance between the material property fingerprint matrices of the product under test and the standard sample, the similarity of the material property fingerprints is obtained.

[0070] Quality Grade Comprehensive Evaluation Module: Based on the similarity of spectral transmission characteristic curves, optical uniformity evaluation indicators, and material characteristic fingerprints, the quality grade of the glass screen is comprehensively scored from multiple dimensions.

[0071] Compared with the prior art, the present invention has at least the following beneficial effects:

[0072] This invention uses multi-band spectral analysis technology to detect glass screens. Based on spatial and frequency dual-domain analysis, it innovatively constructs an optical uniformity evaluation method based on two-dimensional Fourier transform. This method can not only effectively identify subtle optical defects that are difficult for the human eye to detect, but also highlight optical features in a specific frequency range through a frequency weighting function, thus achieving accurate evaluation of local high-frequency defects and large-scale low-frequency non-uniformity on the surface of the glass screen.

[0073] This invention innovatively proposes a material property fingerprint construction method based on spectral reflectance. By analyzing the spectral characteristics of glass screens in multiple bands, a complete material property evaluation system is established. This method introduces an incident angle compensation mechanism when calculating spectral reflectance, which effectively eliminates measurement errors. It also extracts features through regional statistical methods, which not only preserves local information but also has strong noise resistance.

[0074] This invention establishes a multi-dimensional comprehensive evaluation system. By reasonably setting weighting coefficients, it organically combines spectral transmission characteristics, optical uniformity, and material characteristic parameters to achieve a comprehensive evaluation of the quality of glass screens. The evaluation system uses a standard sample comparison method to ensure the accuracy of the evaluation and enables rapid decision-making through clear grading standards, making it highly practical and operable. Attached Figure Description

[0075] Figure 1 This is a flowchart of a method for evaluating the quality of automotive glass screens that combines multispectral imaging, according to Embodiment 1 of the present invention.

[0076] Figure 2 This is a spectral transmission characteristic curve obtained in step S2 of Embodiment 1 of the present invention. Detailed Implementation

[0077] The present invention will be further described below with reference to the accompanying drawings, but this is not intended to limit the present invention in any way. Any modifications or substitutions made based on the teachings of the present invention shall fall within the protection scope of the present invention.

[0078] Example 1:

[0079] like Figure 1 As shown, this embodiment provides a method for evaluating the quality of automotive glass screens by combining multispectral imaging, including the following steps:

[0080] S1: Acquire images of the glass screen at different wavelengths and perform spectral response calibration to obtain the calibrated multispectral data cube:

[0081] S11: Data acquisition using a multispectral camera Images of the glass screen at different spectral bands were used to obtain the original multispectral image sequence. ,in For the first Images of the glass screen acquired in each band. For band number, In this embodiment, the multispectral camera has a wavelength range of 380nm-780nm, and the interval between adjacent wavelength bands is [missing information]. , ;

[0082] S12: Perform spectral response calibration on the original multispectral image sequence to obtain the calibrated multispectral data cube. Specifically:

[0083] ;

[0084] in, and These are the horizontal and vertical coordinates of the image on the glass screen, respectively. For the first Images after band calibration for In pixel coordinates Pixel value at that location, For the first Images of the glass screen acquired in each band In pixel coordinates Pixel value at that location, and These represent the mean and variance of the standard spectral response, respectively; in this embodiment, they are 128 and 32. and The first In each band image, pixel coordinates Center Mean and variance of pixel values ​​within the range It is a natural constant.

[0085] This step involves acquiring images of the glass screen in different spectral bands using a multispectral camera and performing spectral response calibration to obtain data that accurately reflects the spectral characteristics of the glass screen. This step uses local regional statistical features for calibration, which can effectively eliminate image distortion caused by factors such as uneven illumination and nonlinear camera response, thereby improving the accuracy of subsequent analysis.

[0086] S2: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution characteristics of each band in the divided regions on the glass screen, and construct the spectral transmittance characteristic curves:

[0087] S21: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution of each band on the glass screen surface, specifically:

[0088] ;

[0089] in, For the first Each band in pixel coordinates Spectral transmittance at that location For reference, the whiteboard is in the first... bands, pixel coordinates The reflectance value at that location, For the incident light at the th bands, pixel coordinates Strength at that location, The standard incident light intensity is 1000 in this embodiment. This is the transmittance correction factor, which is 0.2 in this embodiment;

[0090] S22: Calculate the spectral transmittance distribution characteristics of the divided regions on the glass screen, specifically:

[0091] ;

[0092] in, For the first The band in the first Spectral transmittance distribution characteristics of each region For area code, , The total number of regions is 64 in this embodiment, meaning the glass screen image is divided into 8 equal parts. 8 grid, and The first The range of x and y coordinates for each region pixel coordinates Spatial weight function at the location, The weight parameter is 1 / 4 of the minimum side length of the region in this embodiment;

[0093] The spatial weighting function is specifically as follows:

[0094] ;

[0095] in, For the first The center pixel coordinates of each region;

[0096] S23: Construct spectral transmission characteristic curves, such as Figure 2 As shown, specifically:

[0097] ;

[0098] in, For the first Spectral transmission characteristic curve values ​​for each band, For the first Spectral smoothing function for each band, and These are the center band and bandwidth parameters of the smoothing function, respectively, in this embodiment they are respectively and ;

[0099] The spectral smoothing function is specifically:

[0100] ;

[0101] This step employs reference whiteboard calibration and incident light intensity compensation to effectively eliminate the influence of ambient light variations and equipment response differences, ensuring the accuracy and repeatability of the measurement results. By introducing the calculation of regional spectral transmittance distribution characteristics, the differences in optical properties of different regions of the glass screen can be meticulously characterized. Simultaneously, the design of the spatial weighting function reduces the interference of regional boundary effects and improves the stability of feature extraction.

[0102] S3: Based on the spatial frequency distribution of spectral transmittance, calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, and construct an optical uniformity evaluation index:

[0103] S31: Calculate the spatial frequency distribution of spectral transmittance, specifically:

[0104] ;

[0105] in, For the first Two-dimensional Fourier transform results for each band. for The zero-mean result, and These are the frequency coordinates in the horizontal and vertical directions, respectively. and These represent the number of rows and columns of the image on the glass screen, respectively. The imaginary unit;

[0106] S32: Calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, specifically:

[0107] ;

[0108] in, For the first Optical uniformity evaluation characteristics for each band. The size of the frequency analysis window is, in this embodiment, . And it is an integer power of 2. To return and The smaller value, For frequency weighting function, and These are the frequency weighting parameters for the horizontal and vertical directions, respectively; in this embodiment, both are... ;

[0109] S33: Construct an optical uniformity evaluation index, specifically:

[0110] ;

[0111] in, It serves as an evaluation index for the optical uniformity of the glass screen surface.

[0112] This step utilizes a two-dimensional Fourier transform to convert the spectral transmission characteristics of the glass screen from the spatial domain to the frequency domain for analysis, effectively capturing subtle optical changes and periodic defects that are difficult for the human eye to perceive. By designing a frequency weighting function, this step can highlight optical features within a specific frequency range, enabling the detection of localized high-frequency defects as well as the assessment of large-scale low-frequency optical inhomogeneities.

[0113] S4: Based on the projection position of the light source and the intensity of the incident light, calculate the spectral reflectance distribution of each band on the glass screen surface, and then obtain the spectral characteristic values ​​of the divided regions on the glass screen. By weighting the contributions of different bands, construct a material property fingerprint matrix. By calculating the Euclidean distance between the material property fingerprint matrices of the product under test and the standard sample, obtain the similarity of the material property fingerprints.

[0114] S41: Calculate the spectral reflectance distribution of each band on the glass screen surface, specifically:

[0115] ;

[0116] in, For the first Each band in pixel coordinates Spectral reflectance at that location For the first Each band in pixel coordinates The intensity of the reflected light at the point is obtained in this embodiment by measuring a light intensity sensor installed coaxially with the incident light source. For the first Each band in pixel coordinates The incident light intensity at a given location is calculated in this embodiment using the output power of a calibrated light source. and These are the horizontal and vertical coordinates of the center of the light source's projection on the glass screen, respectively. The vertical distance from the light source to the glass screen;

[0117] S42: Calculate the spectral characteristic values ​​of the divided regions on the glass screen, specifically:

[0118] ;

[0119] in, For the first The region in the first Spectral characteristic values ​​of each band, For the first The set of pixel coordinates contained in a region;

[0120] S43: Constructing a material property fingerprint matrix Specifically:

[0121] ;

[0122] in, For the first The region in the first Material property fingerprint values ​​for each wavelength band;

[0123] S44: Calculate the similarity of material property fingerprints:

[0124] ;

[0125] in, The similarity of material property fingerprints, This is the material property fingerprint matrix of the standard sample. In this embodiment, the standard sample is a glass screen sample that has undergone rigorous quality inspection and has excellent optical performance. The matrix was obtained by averaging multiple measurements. For the standard sample The region in the first Material property fingerprint values ​​for each wavelength band.

[0126] This step introduces an incident angle compensation mechanism when calculating spectral reflectance, which effectively eliminates measurement errors caused by different light source illumination angles, improves the accuracy of reflectance calculation, and thus comprehensively characterizes the material properties of the glass screen material.

[0127] S5: Based on the similarity of spectral transmission characteristic curves, optical uniformity evaluation indicators, and material property fingerprints, a multi-dimensional comprehensive score is given to the quality grade of the glass screen:

[0128] S51: Calculate the spectral transmission characteristic score, specifically:

[0129] ;

[0130] in, For the standard sample at the 1st Spectral transmission characteristic curve values ​​for each band, This is the weighting coefficient for spectral transmission characteristics, which is 0.4 in this embodiment. For the standard sample at the 1st The mean spectral transmittance at all pixel coordinates in each band;

[0131] S52: Calculate the optical uniformity score, specifically:

[0132] ;

[0133] in, The optical uniformity of standard samples is used as an evaluation index. This is the optical uniformity weighting coefficient, which is 0.3 in this embodiment;

[0134] S53: Calculate the material property parameter score, specifically:

[0135] ;

[0136] in, This is the weighting coefficient for material property parameters, which is 0.3 in this embodiment;

[0137] S54: Calculate the overall score And determine the quality level:

[0138] ;

[0139] when When, it is judged as Grade A; when When, it is judged as Grade B; when When, it is judged as Grade C; when If so, it is deemed unqualified.

[0140] This step considers three key dimensions simultaneously: spectral transmission characteristics, optical uniformity, and material properties. By appropriately setting weighting coefficients, a comprehensive evaluation of the glass screen quality is achieved. In the spectral transmission characteristics and optical uniformity scoring, a deviation comparison method with standard samples is used to accurately reflect the differences in optical performance between the tested product and the standard sample. In the material property scoring, similarity analysis based on material property fingerprints accurately determines the material consistency level between the tested product and the standard sample. This step weights and integrates the scoring results of these three dimensions, maintaining the comprehensiveness of the evaluation while highlighting the importance of different characteristics. Finally, through clear grading standards, the evaluation results are transformed into intuitive quality levels, facilitating rapid decision-making and quality control on the production floor.

[0141] Example 2:

[0142] This embodiment provides a quality assessment system for automotive glass screens that combines multispectral imaging, including the following modules:

[0143] Data acquisition and calibration module: Acquires glass screen images of different spectral bands and performs spectral response calibration to obtain a calibrated multispectral data cube;

[0144] Spectral transmission characteristics analysis module: Based on the calibrated multispectral data cube, calculate the spectral transmittance distribution characteristics of each band in the divided regions on the glass screen, and construct spectral transmission characteristic curves;

[0145] Optical uniformity analysis module: Based on the spatial frequency distribution of spectral transmittance, calculate the optical uniformity evaluation characteristics of each band on the glass screen surface, and construct optical uniformity evaluation index.

[0146] Material consistency analysis module: Based on the projection position of the light source and the intensity of the incident light, the spectral reflectance distribution of each band on the glass screen surface is calculated, and then the spectral characteristic values ​​of the divided regions on the glass screen are obtained. By weighting the contributions of different bands, a material property fingerprint matrix is ​​constructed. By calculating the Euclidean distance between the material property fingerprint matrices of the product under test and the standard sample, the similarity of the material property fingerprints is obtained.

[0147] Quality Grade Comprehensive Evaluation Module: Based on the similarity of spectral transmission characteristic curves, optical uniformity evaluation indicators, and material characteristic fingerprints, the quality grade of the glass screen is comprehensively scored from multiple dimensions.

[0148] The automotive glass screen quality assessment system provided in this embodiment is used to implement the automotive glass screen quality assessment method in Embodiment 1 above. The functions implemented by each functional module of the automotive glass screen quality assessment system correspond one-to-one with the steps of the automotive glass screen quality assessment method; therefore, they will not be described in detail here.

[0149] It should be noted that the sequence numbers of the above embodiments of the present invention are merely for descriptive purposes and do not represent the superiority or inferiority of the embodiments. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0150] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0151] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for evaluating the quality of a car glass screen in combination with multi-spectrum imaging, characterized in that, The method comprises the following steps: S1: collecting images of the glass screen in different wave bands and performing spectral response calibration to obtain a calibrated multispectral data cube; S2: based on the calibrated multispectral data cube, calculating spectral transmittance distribution characteristics of each wave band in the divided regions on the glass screen, and constructing a spectral transmittance characteristic curve; comprising: S21: based on the calibrated multispectral data cube, calculating the spectral transmittance distribution of each wave band on the surface of the glass screen, specifically: ； wherein, is the waveband number, , is the total number of wavebands, and are the horizontal and vertical coordinates of the image on the glass screen, respectively, is the calibrated image of the waveband, is the pixel value at the pixel coordinate , is the spectral transmittance of the waveband at the pixel coordinate , is the reflectance value of the reference whiteboard at the waveband, at the pixel coordinate , is the intensity of the incident light at the waveband, at the pixel coordinate , is the standard incident light intensity, is the transmittance correction coefficient; S22: calculating the spectral transmittance distribution characteristics of the divided regions on the glass screen, specifically: ； wherein, is the spectral transmittance distribution of the first band in the first region, is the region number, , is the total number of regions, and are the horizontal and vertical coordinate ranges of the first region, respectively, is the spatial weight function at pixel coordinate , is the weight parameter; S23: constructing the spectral transmittance characteristic curve, specifically: ； wherein, is the spectral transmission characteristic curve value for the th wavelength band, is the spectral smoothing function for the th wavelength band, and are the center wavelength and bandwidth parameters of the smoothing function, respectively. S3: based on the spatial frequency distribution of the spectral transmittance, calculating the optical uniformity evaluation characteristics of each wave band on the surface of the glass screen, and constructing an optical uniformity evaluation index; comprising: S31: calculating the spatial frequency distribution of the spectral transmittance, specifically: ； wherein, is the two-dimensional Fourier transform result of the first band, is the zero-mean result of the , and are the horizontal and vertical frequency coordinates, respectively, and are the number of rows and columns of the glass screen image, respectively, is the imaginary unit; S32: calculating the optical uniformity evaluation characteristics of each wave band on the surface of the glass screen, specifically: ； wherein, is the optical uniformity evaluation feature for the th wavelength band, is the frequency analysis window size, is the frequency weight function, and are the horizontal and vertical frequency weight parameters, respectively. S33: constructing the optical uniformity evaluation index, specifically: ； wherein is an index of optical uniformity of the glass screen surface; S4: based on the light source projection position and the incident light intensity, calculating the spectral reflectance distribution of each wave band on the surface of the glass screen, and then obtaining the spectral characteristic values of the divided regions on the glass screen, weighting the spectral characteristic values through the contributions of different wave bands, constructing a material characteristic fingerprint matrix, and obtaining the similarity of the material characteristic fingerprint by calculating the Euclidean distance between the material characteristic fingerprint matrices of the to-be-tested product and the standard sample; S5: based on the spectral transmittance characteristic curve, the optical uniformity evaluation index and the similarity of the material characteristic fingerprint, performing multi-dimensional comprehensive scoring on the quality grade of the glass screen.

2. The method for automotive glazing screen quality assessment in conjunction with multispectral imaging according to claim 1, characterized in that, The step S1 comprises the following steps: S11: acquiring a glass screen image of different wave bands using a multispectral camera, obtaining a raw multispectral image sequence wherein is a glass screen image acquired for the th wave band;​ S12: calibrating the spectral response of the original multispectral image sequence to obtain a calibrated multispectral data cube Specifically: ； wherein is the pixel value at pixel coordinates , is the glass screen image acquired for the th wavelength band the pixel value at pixel coordinates , and are the mean and variance of the standard spectral response, respectively, and are the mean and variance of the pixel values in the range for the pixel coordinates centered in the image acquired for the th wavelength band, is the natural constant.

3. The method for automotive glazing screen quality assessment in conjunction with multispectral imaging according to claim 2, characterized in that, The step S4 comprises the following steps: S41: calculating the spectral reflectance distribution of each wave band on the surface of the glass screen, specifically: ； wherein, Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), Ri is the spectral reflectance of the i-th waveband at pixel coordinate (x, y), and respectively the horizontal and vertical coordinates of the projection center of the light source on the glass screen, is the vertical distance of the light source to the glass screen; S42: calculating the spectral characteristic values of the divided regions on the glass screen, specifically: ； wherein, is the spectral feature value of the first region in the first waveband, is the spectral feature value of the first region in the second waveband, is the spectral feature value of the first region in the third waveband, is the spectral feature value of the first region in the fourth waveband, is the set of pixel coordinates contained in the first region. S43: Constructing a matrix of build material property fingerprints , in particular: ； wherein, is the material property fingerprint value of the nth region in the mth wavelength band; is the material property fingerprint value of the nth region in the mth wavelength band; is the material property fingerprint value of the nth region in the mth wavelength band; S44: calculating the similarity of the material characteristic fingerprint: ； wherein, is a similarity of material property fingerprints, is a matrix of material property fingerprints of the standard sample, is a material property fingerprint value of the standard sample for the i-th region at the j-th wavelength band. is a material property fingerprint value of the standard sample for the i-th region at the j-th wavelength band. is a material property fingerprint value of the standard sample for the i-th region at the j-th wavelength band.

4. The method for automotive glazing screen quality assessment in conjunction with multispectral imaging according to claim 3, characterized in that, The step S5 comprises the following steps: S51: calculating the spectral transmittance characteristic score, specifically: ； wherein, is a spectral transmission characteristic curve value of the standard sample at the i-th wavelength band, is a spectral transmission characteristic curve value of the standard sample at the i-th wavelength band, is a spectral transmission characteristic weight coefficient, is a spectral transmission characteristic curve value of the standard sample at the i-th wavelength band, is a mean value of the spectral transmission at all pixel coordinates of the standard sample at the i-th wavelength band. S52: calculating the optical uniformity score, specifically: ； wherein, is an optical uniformity evaluation index of the standard sample, is an optical uniformity weight coefficient; S53: calculating the material characteristic parameter score, specifically: ； wherein, is a material property parameter weight coefficient; S54: Calculate the comprehensive score and determine the quality grade: ； When , it is determined as A level; when , it is determined as B level; when , it is determined as C level; and when , it is determined as unqualified.

5. A system for quality assessment of automotive glazing screens incorporating multispectral imaging for implementing the method for quality assessment of automotive glazing screens incorporating multispectral imaging according to any one of claims 1 to 4, characterized in that, Comprising: The data acquisition and calibration module acquires images of the glass screen in different wave bands and performs spectral response calibration to obtain a calibrated multispectral data cube; The spectral transmittance characteristic analysis module, based on the calibrated multispectral data cube, calculates the spectral transmittance distribution characteristics of each wave band in the divided regions on the glass screen, and constructs a spectral transmittance characteristic curve; The optical uniformity analysis module, based on the spatial frequency distribution of the spectral transmittance, calculates the optical uniformity evaluation characteristics of each wave band on the surface of the glass screen, and constructs an optical uniformity evaluation index; The material consistency analysis module, based on the light source projection position and the incident light intensity, calculates the spectral reflectance distribution of each wave band on the surface of the glass screen, and then obtains the spectral characteristic values of the divided regions on the glass screen, weights the spectral characteristic values through the contributions of different wave bands, constructs a material characteristic fingerprint matrix, and obtains the similarity of the material characteristic fingerprint by calculating the Euclidean distance between the material characteristic fingerprint matrices of the to-be-tested product and the standard sample; Quality grade comprehensive evaluation module: based on the spectral transmission characteristic curve, the optical uniformity evaluation index and the similarity of the material characteristic fingerprint, the glass screen quality grade is comprehensively scored in multiple dimensions.

Images