A method for detecting the quality of glass bottle surface spraying

Abstract

The present application relates to the technical field of optical detection, in particular to a glass bottle surface spraying quality detection method, comprising the following steps: under a specified wavelength, target incident light is obtained by adjusting a polarization angle, a glass bottle spraying layer is irradiated, a transmission spectrum signal is collected, parallel and vertical polarization transmittance of the spraying layer and a ratio are calculated, and a polarization transmittance ratio is generated. In the present application, the spraying layer is irradiated by light with a polarization angle, a polarization component of the transmitted light is obtained, the uniformity of thickness and the internal structure are analyzed, the weak area, particle defects and excessive spraying of the spraying layer are identified, the defect detection precision is improved, multi-angle transmittance data collection and change rate calculation are performed, color deviation and thickness anomaly are accurately positioned, cross analysis of multiple optical parameters is performed, adhesion, curing quality and color consistency evaluation are more refined, spraying abnormalities are classified, multiple detection methods are synergistically used, environmental light, material reflectivity and other interferences are reduced, and the detection accuracy of the spraying quality is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of optical detection technology, and particularly relates to a glass bottle surface spraying quality detection method. BACKGROUND

[0002] The field of optical detection technology includes methods for detecting the surface, internal structure and material properties of objects, mainly relying on the analysis of light propagation, reflection, scattering, interference, diffraction, absorption and other physical phenomena, which is widely used in industrial production, medical diagnosis, material science and other fields for non-contact detection and high-precision measurement. In industrial quality detection, optical detection technology is often used for surface defect recognition, thickness measurement, color analysis and coating uniformity detection. This field mainly includes laser interference detection, spectral analysis, polarized light detection, fluorescence imaging and machine vision detection methods. Each method realizes high-precision quality control according to specific optical characteristics. Among them, spectral analysis mainly uses the absorption or scattering characteristics of light to detect material composition, and machine vision detection combines image processing algorithms to analyze surface defects or spraying uniformity.

[0003] Among them, the glass bottle surface spraying quality detection method refers to using optical detection technology to evaluate the quality of the spraying layer on the surface of the glass bottle. This patent subject mainly detects the uniformity, adhesion, thickness and color consistency of the spraying layer, uses spectral analysis and image recognition technology to obtain detection data, the spectral analysis method measures the reflection or transmission characteristics of the spraying layer to specific wavelengths of light to judge the thickness and uniformity of the spraying layer, the image recognition technology uses a high-resolution industrial camera to collect images of the glass bottle surface, and through color space segmentation and texture analysis methods, it identifies defects of the spraying layer such as color difference, spots or uneven spraying phenomenon. In addition, polarized light detection is used to analyze the adhesion of the spraying layer, and the curing quality of the coating is evaluated by the change of the polarization state. The entire detection process combines multiple optical means to ensure the stability and consistency of the glass bottle spraying quality.

[0004] In the existing glass bottle spraying quality detection process, the spectral analysis method is relied on to measure the light absorption and reflection characteristics, but the sensitivity to the surface micro thickness change is limited, the local small deviation of the spraying layer is easily ignored, in the color detection process, the image recognition technology is easily affected by the environmental light, resulting in insufficient detection accuracy of color deviation, especially on high reflectivity materials, the risk of misjudgment is high, for defect recognition, the existing method mainly relies on texture analysis of surface image, it is difficult to distinguish the particle defects and curing problems in the spraying layer, leading to that part of the hidden defects are not accurately detected, the evaluation of the adhesion of the spraying layer does not fully consider the polarization light characteristics, only relies on the optical contrast analysis, it is difficult to quantify the uniformity of the curing state, affects the overall judgment of the spraying quality, lacks systematic analysis of the multi-angle transmittance change, resulting in limitations in judging the spraying thickness and color deviation, it is difficult to comprehensively identify the abnormal area of the spraying layer, the above shortcomings lead to great uncertainty in the spraying quality control, affecting the product consistency and the accuracy of subsequent process adjustment. SUMMARY

[0005] The purpose of the present application is to solve the problems existing in the prior art and to provide a glass bottle surface spraying quality detection method.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: a glass bottle surface spraying quality detection method, comprising the following steps:

[0007] S1: at a specified wavelength, the target incident light is obtained by adjusting the polarization angle, the glass bottle spraying layer is irradiated, the transmittance signal is collected, the parallel and vertical polarization transmittance and the ratio of the spraying layer are calculated, and the polarization transmittance ratio is generated.

[0008] S2: based on the polarization transmittance ratio, the vertical incident narrowband laser scattering signal is obtained, the coherent decay rate is calculated, the threshold of the normal area is set to judge whether the decay rate is out of range, and the local abnormal area identification record of the spraying layer is generated.

[0009] S3: combined with the local abnormal area identification record of the spraying layer, the differential angle transmittance change rate is calculated, the color deviation and thickness abnormal area are analyzed, and the spraying layer thickness distribution and color deviation data are obtained.

[0010] S4: based on the local abnormal area identification record of the spraying layer and the spraying layer thickness distribution and color deviation data, the position and density of the surface defect area are determined, the thickness non-uniformity and color deviation are analyzed, the correlation is calculated, and the spraying layer quality evaluation data is generated.

[0011] S5: according to the spraying layer quality evaluation data, the color deviation, thickness abnormality and defect type are distinguished, the out-of-limit category is marked, the deviation range and deviation ratio are calculated, the thickness offset and uniformity are analyzed, and the spraying layer defect positioning classification data is generated.

[0012] As a further aspect of the present invention, the polarization transmittance includes parallel polarization transmittance, perpendicular polarization transmittance, and transmittance ratio; the local abnormal area identification record of the sprayed coating specifically includes the coordinates of the abnormal area, the area of ​​the abnormal area, and the abnormal value of the coherence attenuation rate; the thickness distribution and color deviation data of the sprayed coating include transmittance peak value, transmittance change rate at different angles, color deviation distribution, and thickness abnormal area; the quality assessment data of the sprayed coating specifically includes defect area location, defect area density, thickness non-uniformity analysis results, color deviation analysis results, and factors affecting sprayed coating quality; the defect location and classification data of the sprayed coating includes color deviation exceeding limit category, thickness abnormality category, defect type classification, detection position deviation range, deviation ratio, thickness offset, and thickness uniformity analysis results.

[0013] As a further aspect of the present invention, the specific steps for obtaining incident light of a target wavelength with a set polarization angle, irradiating the glass bottle coating layer, acquiring the transmission spectrum signal, calculating the parallel and perpendicular polarization transmittance of the coating layer, calculating the ratio between the two, and generating the polarization transmittance ratio are as follows:

[0014] S101: Acquire the target wavelength incident light with set horizontal and vertical polarization angles, irradiate the coating layer on the glass bottle surface, adjust the incident and receiving angles of the spectral detector, record the polarization transmission spectrum signal of the coating layer on the glass bottle surface, format the spectral data, and group the spectral data according to the polarization direction to obtain polarization transmission spectrum signal data.

[0015] S102: Based on the polarization transmission spectrum signal data, extract the light intensity values ​​of the horizontal polarization component and the vertical polarization component for the differentiated wavelength points, calculate the transmittance of the multi-wavelength points, and calculate the ratio of the transmitted light intensity value in the differentiated polarization direction to the corresponding incident light intensity value to obtain the polarization direction transmittance data.

[0016] S103: Call the polarization direction transmittance data, calculate the ratio of horizontal polarization component transmittance to vertical polarization component transmittance for all wavelength points, sort the comparison data by wavelength points, and generate polarization transmittance ratio.

[0017] As a further aspect of the present invention, the specific formula for calculating the transmittance at multiple wavelength points is as follows:

[0018]

[0019] Calculate the multi-wavelength point transmittance T λ The ratio of the transmitted light intensity value in the differentiated polarization direction to the corresponding incident light intensity value is calculated to obtain the polarization direction transmittance data.

[0020] Among them, T λI represents the transmittance at wavelength λ. ⊥,λ I represents the intensity of the vertically polarized component light at wavelength λ. ||,λ This represents the intensity of the horizontal polarization component of the light at wavelength λ. This represents the incident light intensity at wavelength λ3. This represents the cumulative sum of the products of the vertically and horizontally polarized light intensities from wavelength λ1 to λ2. This represents the sum of incident light intensity values ​​from wavelength λ3 to λ4.

[0021] As a further aspect of the present invention, the specific steps for obtaining the vertically incident narrowband laser scattering signal based on the polarization transmittance, calculating the coherent attenuation rate, setting a normal region threshold to determine whether the attenuation rate exceeds the range, and generating a record of local abnormal areas in the sprayed coating are as follows:

[0022] S201: Obtain the polarization transmittance of the sprayed coating, detect the scattering signal generated after the narrowband laser is perpendicularly incident on the sprayed coating on the surface of the glass bottle, collect the coherence length data of the reflected and scattered light recorded by the detector, calculate the coherence attenuation rate of the scattered light based on the coherence length data, and generate a coherence attenuation rate distribution record of the sprayed coating.

[0023] S202: Based on the coherent attenuation rate distribution record of the sprayed coating, a coherent attenuation rate threshold for the normal area of ​​the sprayed coating is set. The coherent attenuation rate data at the detection location is compared with the set threshold to determine whether the attenuation rate exceeds the threshold range. The locations that exceed the threshold range are recorded, and the areas that are higher or lower than the threshold are classified and marked to generate a labeling result for the abnormal attenuation rate area of ​​the sprayed coating.

[0024] S203: Call the abnormal decay rate region marking results of the sprayed coating, filter out the actual locations of the weak areas, particle defects, sprayed coating accumulation and over-spraying areas of the sprayed coating, and mark the abnormal areas in combination with the detection location coordinate information to obtain the local abnormal area marking data of the sprayed coating.

[0025] As a further aspect of the present invention, the specific steps for calculating the rate of change of transmittance at different angles, analyzing color deviation and thickness anomaly areas, and obtaining the thickness distribution and color deviation data of the sprayed coating are as follows: (Based on the identification record of local abnormal areas in the sprayed coating,)

[0026] S301: Based on the polarization transmittance and the identification data of the local abnormal area of ​​the sprayed coating, obtain the transmission signal of the multi-angle light source during the rotation process, calculate the multi-angle transmittance curve, and record the spectral distribution of the corresponding wavelength range to obtain the multi-angle transmittance peak value and spectral distribution data.

[0027] S302: Based on the multi-angle transmittance peak and spectral distribution data, calculate the transmittance gradient change between adjacent angles, analyze the rate of transmittance change at multiple angles during rotation, collect the transmittance gradients at all angles, and obtain multi-angle transmittance gradient change information.

[0028] S303: Based on the multi-angle transmittance gradient change information, compare the multi-angle transmittance gradient with the overall trend, filter areas where the transmittance gradient change exceeds the benchmark threshold and mark them as color deviation areas, and filter areas where the transmittance gradient change is below the benchmark threshold and mark them as abnormal coating thickness areas, thereby obtaining coating thickness distribution and color deviation data.

[0029] As a further aspect of the present invention, the formula for calculating the multi-angle transmittance curve is specifically as follows:

[0030]

[0031] The transmission signals of the light source from multiple angles during the rotation process are acquired, and the spectral distribution of the corresponding wavelength range is recorded to obtain the peak transmittance and spectral distribution data of the multiple angles.

[0032] Among them, T θ This represents the transmittance at different angles during rotation. Representing wavelength λ i The corresponding light source power, α i The representative coating at wavelength λ i absorption coefficient at d i The representative coating at wavelength λ i The local thickness at Δd i The representative coating at wavelength λ i The local thickness change at a given point, where θ represents the rotation angle, cos(θ) represents the influence factor of the rotation angle on the transmittance, and Δθ j Representing angle θ j The rotational error at the point is where B represents the number of angles measured and n represents the number of wavelengths measured.

[0033] As a further aspect of the present invention, based on the recorded identification of local abnormal areas in the sprayed coating and the data on the thickness distribution and color deviation of the sprayed coating, the specific steps for determining the location and density of surface defect areas, analyzing thickness non-uniformity and color deviation, calculating correlation, and generating sprayed coating quality assessment data are as follows:

[0034] S401: Obtain the identification data of the local abnormal area of ​​the sprayed coating and the thickness and color deviation data of the sprayed coating, extract the coordinate information, area and morphological features of the abnormal area, calculate the distribution mean, range and thickness gradient change value of the thickness, and generate statistical values ​​of the abnormal area of ​​the sprayed coating by combining the mean, maximum deviation value and deviation range of the color deviation.

[0035] S402: Based on the statistical values ​​of the abnormal areas of the sprayed coating, the distribution density of the coordinate information of the abnormal areas is calculated. The thickness non-uniformity is classified by using the area and thickness gradient change value. The proportion of abnormal areas under the differential thickness category is calculated. The maximum deviation value and deviation range of the color deviation are called to classify the distribution of color deviation in the differential thickness category and generate the defect distribution area of ​​the sprayed coating.

[0036] S403: Call the defect distribution area of ​​the sprayed coating, analyze the correlation between the coordinate information, thickness non-uniformity classification and color deviation classification data of the abnormal area, calculate the degree of influence on the overall quality of the sprayed coating, and generate sprayed coating quality assessment data.

[0037] As a further aspect of the present invention, based on the coating quality assessment data, the specific steps for distinguishing color deviation, thickness anomalies, and defect types, marking the corresponding out-of-limit categories, calculating the deviation range and deviation ratio, analyzing thickness offset and uniformity, and generating coating defect location classification data are as follows:

[0038] S501: Based on the coating quality assessment data, extract the color deviation value, coating thickness value and defect feature value of the sprayed area, perform deviation calculation on the data, filter out areas that exceed the average threshold, mark the corresponding out-of-limit category, and generate out-of-limit area classification label data.

[0039] S502: Based on the classification and marking data of the out-of-limit areas, extract the out-of-limit area data according to color deviation, thickness anomaly and defect type, match the detection position of the spray coating according to the out-of-limit category, calculate the deviation range and deviation ratio of the detection position, and generate the coordinate data of the abnormal detection of the spray coating.

[0040] S503: Call the coordinate data of the abnormal detection of the sprayed coating, calculate the color offset data for the color deviation area, analyze the thickness offset and thickness uniformity for the abnormal thickness area of ​​the sprayed coating, and generate the defect location classification data of the sprayed coating for the defect area.

[0041] As a further aspect of the present invention, the specific formula for matching the detection position of the sprayed coating according to the over-limit category is as follows:

[0042]

[0043] Calculate the detection position deviation range D p Detection position deviation ratio R d The specific formula is:

[0044]

[0045] Calculate the detection position deviation ratio R d Based on the category of exceeding limits, match the detection position of the sprayed coating, calculate the deviation range and deviation ratio of the detection position, and generate coordinate data for abnormal detection of the sprayed coating.

[0046] Among them, D p X represents the range of detection position deviation. i X represents the x-coordinate of the i-th over-limit region. c The x-coordinate, Y, represents the detection position of the matched coating layer. j Y represents the ordinate of the j-th overlimit region. c The vertical coordinate represents the location of the matched coating layer detection, S represents the number of horizontal coordinate data points in the out-of-limit area, m represents the number of vertical coordinate data points in the out-of-limit area, and T represents the number of vertical coordinate data points in the out-of-limit area. k R represents the thickness value at the k-th detection location, p represents the total number of thickness detection points, and R d L represents the detection position deviation ratio. r This represents the length of the detection reference.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0048] In this invention, the polarization component of the transmitted light is obtained by irradiating the sprayed coating with light of a polarization angle, enabling precise analysis of thickness uniformity and internal structure. This allows for the identification of weak areas, particle defects, and overspray in the sprayed coating, improving defect detection accuracy. Multi-angle transmittance data acquisition and rate of change calculation accurately locate color deviations and thickness anomalies. Cross-analysis of multiple optical parameters further refines the evaluation of adhesion, curing quality, and color consistency, accurately classifying spraying anomalies. The synergistic effect of multiple detection methods reduces interference from ambient light and material reflectivity, improving the accuracy of spraying quality detection. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the steps of the present invention;

[0050] Figure 2 This is a flowchart of steps S1 of the present invention;

[0051] Figure 3 This is a flowchart of steps S2 of the present invention;

[0052] Figure 4 This is a flowchart of steps S3 of the present invention;

[0053] Figure 5 This is a flowchart of step S4 of the present invention;

[0054] Figure 6 This is a flowchart of step S5 of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0056] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0057] Please see Figure 1 A method for inspecting the quality of coating on the surface of glass bottles, comprising the following steps:

[0058] S1: At a specified wavelength, the target incident light is obtained by adjusting the polarization angle, which is then used to irradiate the coating layer on the surface of the glass bottle. The polarization direction transmission spectrum signal recorded by the spectral detector is collected, and the transmittance of the parallel polarization component and the vertical polarization component of the coating layer at the wavelength point is extracted. The ratio of the two components is calculated to generate the polarization transmittance ratio.

[0059] S2: Based on polarization transmittance, obtain the scattering signal of narrowband laser perpendicularly incident on the coating layer on the surface of the glass bottle, collect the coherence length data of the reflected scattered light recorded by the detector, calculate the coherence attenuation rate of the scattered light at the detection position, set the coherence attenuation rate threshold of the normal area of ​​the coating layer, and determine whether the attenuation rate exceeds the threshold range. If the attenuation rate is higher than the threshold, it is marked as a weak area or particle defect of the coating layer. If the attenuation rate is lower than the threshold, it is marked as a coating layer accumulation or over-spraying area, and a record of local abnormal area identification of the coating layer is generated.

[0060] S3: Based on the record of local abnormal areas of the sprayed coating, and based on the glass bottle rotation detection process, obtain the peak transmittance data at multiple angles, calculate the transmittance change rate at multiple angles, and mark the area as a color deviation area if the transmittance change rate is higher than the overall trend, and mark the area as an abnormal area of ​​the sprayed coating thickness if the transmittance change rate is lower than the overall trend, thus obtaining the sprayed coating thickness and color deviation data.

[0061] S4: Using the identification and recording of local abnormal areas of the spray coating and the data on the thickness and color deviation of the spray coating, determine the location and density of surface defect areas, analyze the thickness unevenness and capture color deviation during the spraying process, calculate the correlation between the three, analyze the factors affecting the spray coating quality, and generate spray coating quality assessment data.

[0062] S5: Based on the coating quality assessment data, the deviation areas exceeding the set threshold are classified, color deviation, coating thickness abnormality and defect type are distinguished, the corresponding excess category is marked, the deviation range and deviation ratio of the detection position are calculated, the thickness offset and thickness uniformity are analyzed, and coating defect location classification data are generated.

[0063] The polarization transmittance includes parallel polarization transmittance, perpendicular polarization transmittance, and transmittance ratio. The local abnormal area identification record of the sprayed coating includes the coordinates of the abnormal area, the area of ​​the abnormal area, and the abnormal value of the coherence attenuation rate. The thickness distribution and color deviation data of the sprayed coating include the peak transmittance, the rate of change of transmittance at different angles, the color deviation distribution, and the thickness abnormal area. The quality assessment data of the sprayed coating includes the location of the defect area, the density of the defect area, the results of the thickness non-uniformity analysis, the results of the color deviation analysis, and the factors affecting the sprayed coating quality. The defect location and classification data of the sprayed coating includes the color deviation exceeding the limit category, the thickness abnormality category, the defect type classification, the detection position deviation range, the deviation ratio, the thickness offset, and the thickness uniformity analysis results.

[0064] Please see Figure 2 The specific steps of S1 are as follows:

[0065] S101: At a specified wavelength, target incident light is obtained by adjusting the polarization angle, which is then used to irradiate the coating layer on the glass bottle surface. The incident angle and receiving angle of the spectral detector are adjusted to record the polarization transmission spectral signal of the coating layer on the glass bottle surface. The spectral data is formatted and grouped according to the polarization direction to obtain polarization transmission spectral signal data.

[0066] The system acquires target wavelength incident light with set horizontal and vertical polarization angles. A light source system generates a beam with a specific polarization angle, which can be controlled by a polarizer or a liquid crystal tunable polarizer to ensure the incident light has a strictly defined polarization direction and wavelength distribution. This beam illuminates the coating layer on the glass bottle surface, causing it to interact with the outer coating. Depending on the glass surface material and coating thickness, part of the incident light is transmitted while the other part is reflected. By adjusting the incident and receiving angles of the spectral detector, the detector can receive the horizontal and vertical polarization components of the light signal transmitted through the coating layer. To ensure data accuracy, spectral data needs to be collected within multiple angle ranges, for example, from an incident angle of 0° to 60°, with measurements taken every 5°. The detector's receiving angle is also synchronized to achieve complete acquisition of the polarization transmission spectrum. After the spectral detector acquires the spectral data, the data is formatted, including noise filtering and intensity value normalization. The data is then grouped according to the polarization direction, and the horizontal and vertical polarization transmission spectrum data are stored separately to form complete polarization transmission spectrum signal data.

[0067] S102: Based on polarization transmission spectrum signal data, for different wavelength points, extract the light intensity values ​​of the horizontal polarization component and the vertical polarization component, calculate the transmittance of multiple wavelength points, and calculate the ratio of the transmitted light intensity value in the different polarization direction to the corresponding incident light intensity value to obtain the polarization direction transmittance data.

[0068] Based on the acquired polarization transmission spectral signal data, the light intensity values ​​of the horizontal and vertical polarization components are extracted for different wavelengths. A high-precision spectrometer is used to measure the transmitted light intensity of these components. For example, for laser sources with wavelengths of 532 nm, 632.8 nm, and 1064 nm, the transmission intensities Ih in the horizontal polarization direction are measured to be 0.45, 0.62, and 0.38, respectively, and the transmission intensities Iv in the vertical polarization direction are measured to be 0.30, 0.55, and 0.28, respectively. To calculate the transmittance at multiple wavelengths, the incident light intensity I0 needs to be compared. Setting I0 to 1.00, then... The transmittance calculation formula is T = I / I0, where I represents the transmitted light intensity. At 532 nm, the horizontal polarization transmittance Th = 0.45 / 1.00 = 0.45, and the vertical polarization transmittance Tv = 0.30 / 1.00 = 0.30. To obtain the ratio of transmitted light intensity in different polarization directions, the transmittance data is calculated. By comparing the ratio of horizontal to vertical polarization transmittance, R = Th / Tv is defined. Then, at 532 nm, R = 0.45 / 0.30 = 1.5. Similarly, the ratio at different wavelengths is calculated to obtain polarization transmittance data at different wavelengths, thus completing the basic calculation for spectral analysis.

[0069] S103: Based on polarization transmission spectrum signal data, extract the light intensity values ​​of the horizontal polarization component and the vertical polarization component for different wavelength points, calculate the transmittance of multiple wavelength points, and calculate the ratio of the transmitted light intensity value in the different polarization direction to the corresponding incident light intensity value to obtain the polarization direction transmittance data.

[0070] The specific formula for calculating the transmittance at multiple wavelengths is as follows:

[0071]

[0072] Calculate the multi-wavelength point transmittance T λ The ratio of the transmitted light intensity value in the differentiated polarization direction to the corresponding incident light intensity value is calculated to obtain the polarization direction transmittance data.

[0073] Among them, T λ I represents the transmittance at wavelength λ. ⊥,λ I represents the intensity of the vertically polarized component light at wavelength λ. ||,λ This represents the intensity of the horizontal polarization component of the light at wavelength λ. This represents the incident light intensity at wavelength λ3. This represents the cumulative sum of the products of the vertically and horizontally polarized light intensities from wavelength λ1 to λ2. This represents the sum of incident light intensity values ​​from wavelength λ3 to λ4;

[0074] Parameter definition and data acquisition:

[0075] I ⊥,λ The intensity of the vertically polarized component of light at wavelength λ. This is obtained by measuring light at a specific wavelength using a polarization spectrometer under experimental conditions.

[0076] I ||,λ The intensity of the horizontal polarization component at wavelength λ is also obtained by measuring with a polarization spectrometer.

[0077] I 0,λ The intensity of the incident light at wavelength λ is obtained from data provided by a standard light source or from a pre-calibrated spectrometer.

[0078] Specific numerical settings and calculation examples:

[0079] Assuming I is measured at a wavelength of 500 nm ⊥,500nm =20 units, I ||,500nm =30 units, I 0,500nm = 100 units.

[0080] Suppose we need to calculate data in the wavelength range from 400nm to 700nm, we set λ1 = 400nm, λ2 = 700nm, λ3 = 400nm, and λ4 = 700nm.

[0081] calculate Value:

[0082] If I 0,λ The average value in the 400nm to 700nm range is approximately 150 units. There are 301 data points within this range. unit.

[0083] calculate

[0084] Assuming the range is 400nm to 700nm, I ⊥,λ and I ||,λ The average of the product is 600 units. For the same 301 data points within this interval, then... thereby unit.

[0085] Calculate T λ :

[0086] Substituting the value of λ = 500nm, the calculation is as follows:

[0087]

[0088] This result indicates that at a wavelength of 500 nm, the transmittance T in the polarization direction... 500 The transmittance is 0.96%. This means that the transmittance at the measured wavelength is 0.96% of the incident light intensity, a value that can be used for further analysis of the polarization properties of the material.

[0089] Please see Figure 3 The specific steps of S2 are as follows:

[0090] S201: Obtain the polarization transmittance, detect the scattering signal generated after the narrowband laser is perpendicularly incident on the coating layer on the surface of the glass bottle, collect the coherence length data of the reflected and scattered light recorded by the detector, calculate the coherence attenuation rate of the scattered light based on the coherence length data, and generate a record of the coherence attenuation rate distribution of the coating layer.

[0091] To obtain the polarization transmittance, a polarization measurement device is needed to perform multi-angle optical measurements on the coating layer on the glass bottle surface, recording the transmittance parameters when a specific wavelength of laser light passes through the coating layer. The transmittance is calculated as follows: Among them I t I represents the intensity of transmitted light. i The transmittance data, representing the incident light intensity, measured at different angles, needs to be normalized to eliminate the influence of light source power fluctuations in the experimental environment. The normalization formula is: Where T refThe transmittance of the uncoated substrate is represented by the average polarization transmittance of the coated layer obtained through multiple sets of measurement data. Then, a narrow-band laser is incident perpendicularly onto the surface of the coated glass bottle, and a detector records the intensity distribution of the scattered signal. The scattered signal can be acquired using a photomultiplier tube (PMT) or a photon counter, and the photon arrival time is recorded to form a time distribution curve. Based on the changes in the scattered signal, the coherence length of the reflected and scattered light is calculated. The specific calculation method is as follows: Where L c Let be the coherence length, c be the speed of light, n be the refractive index of the coating, and Δf be the spectral width of the light source. By collecting coherence length data at different locations, the coherent attenuation rate of the scattered light is calculated. The attenuation rate is calculated as follows: Among them I s The intensity of the scattered light is represented by x, which is the detection position. The coherent attenuation rate distribution record of the sprayed coating is constructed by using the attenuation rate calculation results of multiple detection points.

[0092] S202: Based on the coherent attenuation rate distribution record of the sprayed coating, a coherent attenuation rate threshold for the normal area of ​​the sprayed coating is set. The coherent attenuation rate data at the detection location is compared with the set threshold to determine whether the attenuation rate exceeds the threshold range. The locations that exceed the threshold range are recorded, and the areas that are higher or lower than the threshold are classified and marked to generate the labeling result of the abnormal attenuation rate area of ​​the sprayed coating.

[0093] Based on the coherent attenuation rate distribution records of the sprayed coating, a threshold for the coherent attenuation rate in the normal region of the sprayed coating is set. This requires selecting data from a uniform region of the glass bottle sprayed coating as a benchmark to calculate the average coherent attenuation rate R at each point within that region. avg and standard deviation σ R The threshold is set to R. th =R avg ±kσ R Where k is an empirical coefficient, typically ranging from 2 ≤ k ≤ 3 to cover most of the normal region's data. The coherent attenuation rate data at the detection location is compared with a set threshold. If the attenuation rate R at a certain location exceeds R... avg +kσ R If the area has a thin coating or particle defects, then the area may have these defects. If R is lower than R... avg -kσ R If R > R, then the area may have coating buildup or over-coating. The specific classification method is as follows: If R > R th,upper Then mark it as a weak area of ​​the spray coating or a particle defect point, if R <R th,lower The area is marked as a region of coating accumulation or over-coating. The data after classification and marking form the regional distribution of abnormal coating decay rate, and the regional marking results of abnormal coating decay rate are obtained.

[0094] S203: Call the abnormal decay rate area marking results of the sprayed coating, filter out the actual location of the weak area of ​​the sprayed coating, particle defects, sprayed coating accumulation and over-spraying area, and mark the abnormal area in combination with the detection location coordinate information to obtain the local abnormal area marking data of the sprayed coating.

[0095] The system retrieves the results of marking areas with abnormal attenuation rates in the sprayed coating, filters out the actual locations of weak areas, particle defects, coating buildup, and over-coating areas, and performs location matching for all abnormal areas based on the spatial coordinate information in the detection data. Weak areas are defined as regions with attenuation rates exceeding the normal range, while particle defects typically appear as locally prominent high-attenuation-rate points. The system then sets the standard deviation σ of the attenuation rate within a local window w. w If the decay rate at a certain point exceeds 3σ of the window mean w Then this point can be defined as a particle defect point, and the area of ​​coating accumulation will show a low coherent decay rate region. If multiple points around the detection point satisfy R <R th,lower If it is defined as an area of ​​coating accumulation, after the abnormal area is classified by the above method, the identification data of local abnormal areas of the coating is obtained by combining the detection location coordinates.

[0096] Please see Figure 4 The specific steps of S3 are as follows:

[0097] S301: Based on the polarization transmittance and the identification data of local abnormal areas of the sprayed coating, obtain the transmission signals of the light source from multiple angles during the rotation process, calculate the multi-angle transmittance curves, and record the spectral distribution of the corresponding wavelength range to obtain the multi-angle transmittance peak value and spectral distribution data.

[0098] The formula for calculating multi-angle transmittance curves is as follows:

[0099]

[0100] The transmission signals of the light source from multiple angles during the rotation process are acquired, and the spectral distribution of the corresponding wavelength range is recorded to obtain the peak transmittance and spectral distribution data of the multiple angles.

[0101] Among them, T θ This represents the transmittance at different angles during rotation. Representing wavelength λ i The corresponding light source power, α i The representative coating at wavelength λ i absorption coefficient at d i The representative coating at wavelength λ i The local thickness at Δd i The representative coating at wavelength λ iThe local thickness change at a given point, where θ represents the rotation angle, cos(θ) represents the influence factor of the rotation angle on the transmittance, and Δθ j Representing angle θ j The rotational error at the point, where B represents the number of angles measured and n represents the number of wavelengths measured;

[0102] Formula calculation derivation:

[0103] The molecular part of the transmittance calculation:

[0104] First, calculate the transmission contribution at each wavelength:

[0105]

[0106] Taking a wavelength of 500nm as an example:

[0107]

[0108] The calculation yields:

[0109]

[0110] The denominator for calculating transmittance:

[0111] Sum the power of the light source for all wavelengths. Assume five wavelengths were measured, and the power for each wavelength is 1.5 W / m. 2 :

[0112]

[0113] The first part of calculating transmittance:

[0114] Substitute the above results into:

[0115]

[0116] Calculate the normalized sum of squares for thickness variation and angular error:

[0117] For thickness variations:

[0118]

[0119] Regarding angular error, assuming three angles were measured, and the error for each angle is 0.0087 radians:

[0120]

[0121] Calculate the normalization factor:

[0122] Add the above results and take the square root:

[0123]

[0124] Take the reciprocal:

[0125]

[0126] Final transmittance calculation:

[0127] Multiply all parts:

[0128] T θ =0.866 × 4.436 ≈ 3.844;

[0129] The results show that, under the set measurement conditions, the normalized transmittance of the sprayed coating at a 30° angle is [data missing].

[0130] S302: Based on the peak transmittance and spectral distribution data of multiple angles, calculate the transmittance gradient change of adjacent angles, analyze the rate of transmittance change of multiple angles during rotation, collect the transmittance gradients of all angles, and obtain the information on the transmittance gradient change of multiple angles.

[0131] After obtaining the transmittance peak values ​​and spectral distribution data from multiple angles, it is necessary to calculate the gradient change of transmittance between adjacent angles and analyze the rate of transmittance change at different angles during rotation. First, the transmittance gradient change is defined as the rate of transmittance change between adjacent angles. Specifically, in the calculation, the transmittance difference between adjacent angles is calculated in 5° increments and divided by the angle increment to obtain the transmittance gradient per unit angle change. The calculation of the transmittance gradient change needs to cover all angles and summarize the overall gradient change trend. If the transmittance gradient change within a certain angle range is significantly higher than the average of adjacent angles, there may be non-uniformity in the coating layer or abnormal coating thickness during the spraying process. For further analysis, the standard deviation σ can be introduced. ΔT To measure the overall fluctuation of the transmittance gradient change, the mean of the transmittance gradient change at all angles is calculated. and standard deviation σ ΔT It allows defining the abnormal range of different gradient changes. For example, if the gradient change at a certain angle exceeds the mean... This angle region may exhibit abnormal variations in coating thickness, and if the gradient change in a certain region is lower than... The coating in this area may be too uniform and lack normal optical gradient characteristics. After the calculation is completed, the transmittance gradient change information at all angles is compiled for subsequent analysis.

[0132] S303: Based on the multi-angle transmittance gradient change information, compare the multi-angle transmittance gradient with the overall trend, filter the transmittance gradient change that exceeds the benchmark threshold and mark it as a color deviation area, and filter the transmittance gradient change that is lower than the benchmark threshold and mark it as an abnormal area of ​​coating thickness, thus obtaining the coating thickness distribution and color deviation data.

[0133] Based on the transmittance gradient change information, regions where the transmittance gradient change exceeds the baseline threshold are further filtered and marked as color deviation regions. Simultaneously, regions where the transmittance gradient change is below the baseline threshold are filtered and marked as areas with abnormal coating thickness. In practice, the baseline threshold ΔT for the gradient change must first be set. thresh This threshold can be based on the average gradient change across all measurement points. With standard deviation σ ΔT Configure, for example, configure If the transmittance gradient change value at a certain angle exceeds the threshold, it indicates that the transmittance in that area is changing too rapidly, possibly due to uneven color distribution of the coating layer. Therefore, this area is marked as a color deviation area. On the other hand, a baseline threshold ΔT for abnormal coating thickness is set. low , usually ΔT low It can be set to If the transmittance gradient change value at a certain angle is lower than the threshold, it indicates that the thickness of the coating in that area may be abnormal, such as being too large or too small. After the calculation is completed, the detection results of color deviation and thickness abnormality are visualized to identify abnormal areas at different angles, and finally the coating thickness distribution and color deviation data are obtained.

[0134] Please see Figure 5 The specific steps of S4 are as follows:

[0135] S401: Obtain the identification data of local abnormal areas of the sprayed coating and the thickness and color deviation data of the sprayed coating, extract the coordinate information, area and morphological features of the abnormal areas, calculate the mean, range and thickness gradient change value of the thickness distribution, and generate statistical values ​​of abnormal areas of the sprayed coating by combining the mean, maximum deviation value and deviation range of the color deviation.

[0136] Data on local anomaly areas in the sprayed coating, as well as coating thickness and color deviation data, are obtained. The coordinates, area, and morphological features of these anomaly areas are extracted. Based on the coordinates of the anomaly areas, coating thickness data is extracted and its distribution is statistically analyzed using a mean formula. Calculate the mean thickness, where x i For the thickness data at the detection points, where n is the total number of detection points, the range of the coating thickness is obtained, using the formula: range = x max -x min The difference between the maximum and minimum thickness is calculated to assess the thickness distribution characteristics. Combined with coating thickness data, the variation trend along the detection path is calculated. Calculate the thickness change rate between different detection points, where Δh is the thickness difference between adjacent detection points and Δd is the distance between detection points. Obtain the thickness gradient change value of the sprayed coating, call the color deviation data, and use... Calculate the mean color deviation, where c i For each detection point, the maximum deviation value and deviation range of the color deviation are obtained. Combined with the thickness data and color deviation data, the characteristic data of the abnormal area of ​​the sprayed coating is constructed, and finally the statistical value of the abnormal area of ​​the sprayed coating is obtained.

[0137] S402: Based on the statistical values ​​of abnormal areas in the sprayed coating, the distribution density of the coordinate information of the abnormal areas is calculated. The thickness non-uniformity is classified by using the area and thickness gradient change values. The proportion of abnormal areas under the differential thickness category is calculated. The maximum deviation value and deviation range of the color deviation are called to classify the distribution of color deviation in the differential thickness category and generate the distribution area of ​​the sprayed coating defect.

[0138] Based on statistical values ​​of abnormal areas in the sprayed coating, the coordinate information of these abnormal areas is extracted. A two-dimensional density calculation method is then employed. Calculate the distribution density, where N is the number of marked points within the abnormal region and A is the total area of ​​the abnormal region, to obtain the distribution density data of the abnormal region. Then, retrieve the coating thickness gradient change value and classify the thickness non-uniformity using a threshold judgment method, setting a thickness gradient threshold θ. h When the thickness gradient > θ h When the thickness change is abrupt, the region is marked as a region of abrupt thickness change; otherwise, it is marked as a region of uniform thickness. The proportion of abnormal regions under the differential thickness category is calculated, and the following method is used: Calculate the proportion of each thickness category region in the overall detection area, where A c Let A be the total area of ​​a certain type of abnormal region. t To determine the overall detection area, the maximum color deviation value and range are retrieved to classify the distribution of color deviation within the thickness difference category, and a color deviation threshold θ is set. c When color deviation > θ c If the color deviation is severe, it is marked as a region with a serious color deviation; otherwise, it is marked as a region with a slight color deviation. By combining the thickness and color deviation classification results, the distribution area of ​​the coating defect is obtained.

[0139] S403: Call the defect distribution area of ​​the sprayed coating, analyze the correlation between the coordinate information, thickness non-uniformity classification and color deviation classification data of the abnormal area, calculate the degree of impact on the overall quality of the sprayed coating, and generate sprayed coating quality assessment data.

[0140] By accessing the defect distribution area of ​​the sprayed coating, and analyzing the correlation between the coordinate information of the abnormal areas, the classification of thickness non-uniformity, and the classification of color deviation, the Pearson correlation coefficient was used. Calculate the linear correlation between thickness non-uniformity and color deviation, where X i For thickness gradient data, Y i For color deviation data, and The mean values ​​are used to obtain the correlation coefficient between thickness and color deviation. The changing trends of thickness non-uniformity and color deviation are analyzed. Regression analysis is used to fit the relationship between the two, and the fitting curve equation between thickness change Δh and color deviation Δc is calculated. Combined with the coating quality standard, a quality influence threshold θ is set. q When ρ>θ q This indicates that thickness unevenness has a significant impact on color deviation. Further calculations were made to determine the degree of influence of different defect areas on the overall quality of the sprayed coating. An influence ratio of % was used to calculate the percentage of influence of abnormal areas within the overall sprayed area, where A... d A represents the total area of ​​the defective region. t The total area to be sprayed is used to obtain the final coating quality assessment data.

[0141] Please see Figure 6 The specific steps of S5 are as follows:

[0142] S501: Based on the coating quality assessment data, extract the color deviation value, coating thickness value and defect characteristic value of the sprayed area, calculate the deviation of the data, filter the areas that exceed the average threshold, mark the corresponding out-of-limit categories, and generate out-of-limit area classification label data.

[0143] Based on the coating quality assessment data, the color deviation value, coating thickness value, and defect feature value of the sprayed area are first extracted. The color deviation value can be obtained by extracting the color difference information of the RGB channels through image processing algorithms. The calculation method can use the ΔE color difference formula, that is, ΔE=√[(L2-L1)]. 2 +(a2-a1) 2 +(b2-b1) 2L, a, and b represent the brightness, red-green deviation, and blue-yellow deviation values ​​in the color space, respectively. The coating thickness can be obtained using an ultrasonic or laser thickness gauge. Defect feature values ​​can be extracted based on machine vision to identify surface texture features, edge contour irregularities, and other parameters. Then, data deviation calculations are performed on each parameter, calculated as the deviation ratio relative to the average or standard value. For example, the deviation rate R = (Vi - Vavg) / Vavg × 100%, where Vi is the single-point measurement value and Vavg is the area average. A threshold Th is set to determine if the deviation exceeds the standard range. If |R| > Th, the area is considered defective. For areas exceeding the limit, screening is performed on areas that exceed the threshold, and they are classified based on the type of deviation. Color deviation can be classified according to the ΔE threshold, such as ΔE<2 being within the normal range, 2≤ΔE<5 being a slight deviation, and ΔE≥5 being a severe deviation. Thickness deviation can be determined based on the standard deviation σ of the thickness measurement data and the set deviation range Th, such as |T_measurement - T_standard|>Th. Defects are classified according to defect size, depth, and density. Combining the measurement results, areas exceeding the average threshold are classified and marked, for example, using a three-color marking method (red for severe deviation, yellow for slight deviation, and green for normal). Finally, the classification and marking data of areas exceeding the limit are generated.

[0144] S502: Based on the classification and marking data of the out-of-limit areas, extract the out-of-limit area data according to color deviation, thickness anomaly and defect type, match the detection position of the spray coating according to the out-of-limit category, calculate the deviation range and deviation ratio of the detection position, and generate the coordinate data of the spray coating anomaly detection.

[0145] The specific formula for matching the detection location of the sprayed coating according to the category of exceeding the limit is as follows:

[0146]

[0147] Calculate the detection position deviation range D p Detection position deviation ratio R d The specific formula is:

[0148]

[0149] Calculate the detection position deviation ratio R d Based on the category of exceeding limits, match the detection position of the sprayed coating, calculate the deviation range and deviation ratio of the detection position, and generate coordinate data for abnormal detection of the sprayed coating.

[0150] Among them, D p X represents the range of detection position deviation. i X represents the x-coordinate of the i-th over-limit region. c The x-coordinate, Y, represents the detection position of the matched coating layer. j Y represents the ordinate of the j-th overlimit region.c The vertical coordinate represents the location of the matched coating layer detection, S represents the number of horizontal coordinate data points in the out-of-limit area, m represents the number of vertical coordinate data points in the out-of-limit area, and T represents the number of vertical coordinate data points in the out-of-limit area. k R represents the thickness value at the k-th detection location, p represents the total number of thickness detection points, and R d L represents the detection position deviation ratio. r Represents the detection reference length.

[0151] Detection position deviation range D p Calculation:

[0152] Step 1: Calculate the average deviation between the x-axis and y-axis.

[0153] Average deviation of the horizontal axis:

[0154] Set the x-coordinate of the over-limit region. i The x-coordinate of the matched coating detection position c .

[0155] Calculate the deviation |X| between each point in the out-of-limit area and the detection location. i -X c |

[0156] Calculate the sum of these deviations and divide by the number of data points n to obtain the average deviation of the x-axis.

[0157] Average deviation of the ordinate:

[0158] Similarly, set the ordinate Y of the over-limit region. j The ordinate Y of the matched spray coating detection position c .

[0159] Calculate the deviation between each out-of-limit point and the detection location |Y j -Y c |

[0160] Calculate the sum of these deviations and divide by the number of data points m to obtain the average deviation of the ordinate.

[0161] Step 2: Calculate the root mean square of the thickness value.

[0162] Set the thickness value T at the detection location k .

[0163] Calculate the square of each thickness value

[0164] Calculate the sum of these squared values, divide by the total number of thickness detection points p, and then take the square root to obtain the root mean square of the thickness value.

[0165] Step 3: Calculate the detection position deviation range Dp .

[0166] The total average deviation is obtained by adding the average deviations of the horizontal and vertical axes.

[0167] Multiplying the total average deviation by the root mean square of the thickness value yields the detection position deviation range D. p .

[0168] Substitute specific numerical values ​​into the calculation:

[0169] Setting parameters:

[0170] The number of x-coordinate data points in the out-of-limit region is S = 5.

[0171] The number of ordinate data points in the out-of-limit region is m = 5.

[0172] The total number of thickness detection points at the detection location is p = 5.

[0173] The x-coordinate value of the over-limit region X i They are: 100, 105, 98, 102, 101.

[0174] The x-coordinate of the matched coating detection position c =100.

[0175] The ordinate value Y of the over-limit region j The values ​​are: 200, 198, 202, 199, 201. The ordinate (Y) of the matched coating detection position is... c =200.

[0176] Thickness value T at the detection location k They are 50, 52, 51, 49, and 50 micrometers, respectively.

[0177] Calculation process:

[0178] Average deviation of the horizontal axis:

[0179]

[0180] Average deviation of the ordinate:

[0181]

[0182] Total average deviation:

[0183]

[0184] Root mean square of thickness value:

[0185]

[0186] Detection position deviation range D p:

[0187] 1.6 × 50.41 ≈ 80.66 micrometers;

[0188] Detection position deviation ratio R d Calculation:

[0189] Step 1: Determine the detection reference length L r .

[0190] Assuming the detection reference length L r = 500 micrometers.

[0191] Step 2: Calculate the detection position deviation ratio R d .

[0192] Use the formula:

[0193]

[0194] Substitute the values:

[0195]

[0196] Results analysis:

[0197] Detection position deviation range D p It is approximately 80.66 micrometers.

[0198] This result represents the average deviation between the detection location and the matching coating detection location, taking into account the influence of thickness values.

[0199] Detection position deviation ratio R d Approximately 16.13%.

[0200] This result represents the percentage of the detection position deviation range relative to the detection reference length, and is used to assess the degree of deviation in the detection position.

[0201] S503: Call the coordinate data of the coating abnormality detection, calculate the color offset data for the color deviation area, analyze the thickness offset and thickness uniformity for the coating thickness abnormal area, and generate coating defect location classification data for the defect area.

[0202] The system retrieves coordinate data from coating anomaly detection. For areas with color deviation, it calculates color offset data and uses the ΔE color difference formula to calculate the color offset at different detection points. Furthermore, it analyzes the direction of color offset, for example, by calculating the changing trend of color components through the RGB channels. For areas with abnormal coating thickness, it analyzes the thickness offset and thickness uniformity. The thickness offset can be calculated using the deviation ratio R = (T_measured - T_standard) / T_standard × 100%. Thickness uniformity can be assessed using the variance σ. 2 =∑(Ti-Tavg)2 / (N-1) is used to measure and generate defect location and classification data for the sprayed coating area. Classification can be based on defect size, shape, depth, and distribution density. For example, a defect size d < 0.5mm is considered a minor defect, 0.5mm ≤ d < 2mm is a medium defect, and d ≥ 2mm is a severe defect. The classification of defect size is based on the surface quality standards and functional requirements of the sprayed coating. For instance, in automotive coating inspection, if the defect size d is less than 0.5mm, it is difficult to detect with the naked eye and has little impact on the overall appearance, thus it can be classified as a minor defect. When 0.5mm ≤ d < 2mm, such defects are visible under certain lighting conditions, affecting local aesthetics, and can be classified as... Medium defects; when d ≥ 2 mm, defects can significantly affect the appearance integrity and even the protective performance of the coating, and are therefore classified as severe defects. This classification can be dynamically adjusted by combining coating thickness T, surface gloss G, and visual visibility V. For example, if T is thin (< 50 μm), even a 0.4 mm defect may significantly affect the surface quality, so the threshold for minor defects can be appropriately lowered to d < 0.4 mm. If G is high (> 85 GU, gloss units), the visibility of defects increases, and the upper limit for medium defects can be lowered to 1.5 mm. In addition, if the defect density D exceeds a certain threshold (e.g., D > 0.01 defects / mm), the defect can be classified as severe defects. 2 If small defects accumulate into medium defects, they need to be reclassified. Therefore, the setting of the defect size threshold needs to be dynamically adjusted in combination with the coating material characteristics, spraying process parameters and final application environment to ensure the rationality of classification. The defect distribution density D can be calculated as D = (total number of defects / total inspection area), which finally forms the coating defect classification data.

[0203] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for inspecting the quality of coating on the surface of glass bottles, characterized in that, Includes the following steps: S1: At a specified wavelength, the target incident light is obtained by adjusting the polarization angle, which is then used to irradiate the glass bottle coating. The transmission spectrum signal is collected to calculate the parallel and perpendicular polarization transmittance and ratio of the coating, and to generate the polarization transmittance ratio. S2: Based on the polarization transmittance, obtain the vertically incident narrowband laser scattering signal, calculate the coherent attenuation rate, set a normal region threshold to determine whether the attenuation rate exceeds the range, and generate a record of the local abnormal area of ​​the spray coating. S3: Based on the record of the local abnormal area identification of the sprayed coating, calculate the rate of change of transmittance at the different angle, analyze the color deviation and thickness abnormal area, and obtain the thickness distribution and color deviation data of the sprayed coating. S4: Based on the local abnormal area identification record of the sprayed coating and the thickness distribution and color deviation data of the sprayed coating, determine the location and density of the surface defect area, analyze the thickness non-uniformity and color deviation, calculate the correlation, and generate the sprayed coating quality assessment data. S5: Based on the coating quality assessment data, distinguish between color deviation, thickness anomaly and defect type, mark the corresponding out-of-limit categories, calculate the deviation range and deviation ratio, analyze the thickness offset and uniformity, and generate coating defect location classification data. Based on the polarization transmittance, the following steps are taken to obtain the vertically incident narrowband laser scattering signal, calculate the coherent attenuation rate, set a normal region threshold to determine whether the attenuation rate exceeds the range, and generate a record of local abnormal areas in the sprayed coating: S201: Obtain the polarization transmittance of the sprayed coating, detect the scattering signal generated after the narrowband laser is perpendicularly incident on the sprayed coating on the surface of the glass bottle, collect the coherence length data of the reflected and scattered light recorded by the detector, calculate the coherence attenuation rate of the scattered light based on the coherence length data, and generate a coherence attenuation rate distribution record of the sprayed coating. S202: Based on the coherent attenuation rate distribution record of the sprayed coating, a coherent attenuation rate threshold for the normal area of ​​the sprayed coating is set. The coherent attenuation rate data at the detection location is compared with the set threshold to determine whether the attenuation rate exceeds the threshold range. The locations that exceed the threshold range are recorded, and the areas that are higher or lower than the threshold are classified and marked to generate a labeling result for the abnormal attenuation rate area of ​​the sprayed coating. S203: Call the abnormal decay rate area marking results of the sprayed coating, filter out the actual locations of the weak areas, particle defects, sprayed coating accumulation and over-spraying areas of the sprayed coating, and mark the abnormal areas in combination with the detection location coordinate information to obtain the local abnormal area marking data of the sprayed coating. The specific steps for calculating the rate of change of transmittance at different angles, analyzing color deviation and thickness anomaly areas, and obtaining the coating thickness distribution and color deviation data are as follows, based on the identified records of local anomaly areas in the sprayed coating. S301: Based on the polarization transmittance and the identification data of the local abnormal area of ​​the sprayed coating, obtain the transmission signal of the multi-angle light source during the rotation process, calculate the multi-angle transmittance curve, and record the spectral distribution of the corresponding wavelength range to obtain the multi-angle transmittance peak value and spectral distribution data. S302: Based on the multi-angle transmittance peak and spectral distribution data, calculate the transmittance gradient change between adjacent angles, analyze the rate of transmittance change at multiple angles during rotation, collect the transmittance gradients at all angles, and obtain multi-angle transmittance gradient change information. S303: Based on the multi-angle transmittance gradient change information, compare the multi-angle transmittance gradient with the overall trend, filter the transmittance gradient change that exceeds the benchmark threshold and mark it as a color deviation area, and filter the transmittance gradient change that is lower than the benchmark threshold and mark it as an abnormal coating thickness area, so as to obtain the coating thickness distribution and color deviation data. The specific formula for calculating the multi-angle transmittance curve is as follows: ； The transmission signals of the light source from multiple angles during the rotation process are acquired, and the spectral distribution of the corresponding wavelength range is recorded to obtain the peak transmittance and spectral distribution data of the multiple angles. in, This represents the transmittance at different angles during rotation. Representative wavelength The corresponding light source power, Representative coating at wavelength The absorption coefficient at that location, Representative coating at wavelength Local thickness at that location Representative coating at wavelength The amount of local thickness change at that location. Represents the rotation angle. The factor representing the effect of rotation angle on transmittance. Representative angle The rotational error at point B represents the number of angles measured. This represents the number of wavelengths measured.

2. The method for detecting the quality of glass bottle surface coating according to claim 1, characterized in that, The polarization transmittance includes parallel polarization transmittance, perpendicular polarization transmittance, and transmittance ratio. The local abnormal area identification record of the sprayed coating specifically includes the coordinates of the abnormal area, the area of ​​the abnormal area, and the abnormal value of the coherence attenuation rate. The thickness distribution and color deviation data of the sprayed coating include transmittance peak value, transmittance change rate at different angles, color deviation distribution, and thickness abnormal area. The quality assessment data of the sprayed coating specifically includes defect area location, defect area density, thickness non-uniformity analysis results, color deviation analysis results, and factors affecting spraying quality. The defect location and classification data of the sprayed coating includes color deviation exceeding limit category, thickness abnormality category, defect type classification, detection position deviation range, deviation ratio, thickness offset, and thickness uniformity analysis results.

3. The method for detecting the quality of glass bottle surface coating according to claim 1, characterized in that, At a specified wavelength, the target incident light is obtained by adjusting the polarization angle, irradiating the glass bottle coating layer. The transmission spectrum signal is collected, and the parallel and perpendicular polarization transmittance and ratio of the coating layer are calculated. The specific steps for generating the polarization transmittance ratio are as follows: S101: Acquire the target wavelength incident light with set horizontal and vertical polarization angles, irradiate the coating layer on the glass bottle surface, adjust the incident and receiving angles of the spectral detector, record the polarization transmission spectrum signal of the coating layer on the glass bottle surface, format the spectral data, and group the spectral data according to the polarization direction to obtain polarization transmission spectrum signal data. S102: Based on the polarization transmission spectrum signal data, extract the light intensity values ​​of the horizontal polarization component and the vertical polarization component for the differentiated wavelength points, calculate the transmittance of the multi-wavelength points, and calculate the ratio of the transmitted light intensity value in the differentiated polarization direction to the corresponding incident light intensity value to obtain the polarization direction transmittance data. S103: Call the polarization direction transmittance data, calculate the ratio of horizontal polarization component transmittance to vertical polarization component transmittance for all wavelength points, sort the comparison data by wavelength points, and generate polarization transmittance ratio.

4. The method for detecting the quality of glass bottle surface coating according to claim 3, characterized in that, The specific formula for calculating the transmittance at multiple wavelength points is as follows: ； Calculate the transmittance at multiple wavelengths The ratio of the transmitted light intensity value in the differentiated polarization direction to the corresponding incident light intensity value is calculated to obtain the polarization direction transmittance data. in, Representative wavelength Transmittance at that location Representative wavelength The intensity value of the vertical polarization component light at that location. Representative wavelength The horizontal polarization component light intensity value at that location, Representative wavelength The incident light intensity at that point, Represents from wavelength arrive The cumulative sum of the products of the vertically polarized and horizontally polarized light intensities. Represents from wavelength arrive The sum of the incident light intensities between them.

5. The method for detecting the quality of glass bottle surface coating according to claim 1, characterized in that, Based on the records of local anomaly areas in the sprayed coating and the data on coating thickness distribution and color deviation, the specific steps for determining the location and density of surface defect areas, analyzing thickness non-uniformity and color deviation, calculating correlation, and generating sprayed coating quality assessment data are as follows: S401: Obtain the identification data of the local abnormal area of ​​the sprayed coating and the thickness and color deviation data of the sprayed coating, extract the coordinate information, area and morphological features of the abnormal area, calculate the distribution mean, range and thickness gradient change value of the thickness, and generate statistical values ​​of the abnormal area of ​​the sprayed coating by combining the mean, maximum deviation value and deviation range of the color deviation. S402: Based on the statistical values ​​of the abnormal areas of the sprayed coating, the distribution density of the coordinate information of the abnormal areas is calculated. The thickness non-uniformity is classified by using the area and thickness gradient change value. The proportion of abnormal areas under the differential thickness category is calculated. The maximum deviation value and deviation range of the color deviation are called to classify the distribution of color deviation in the differential thickness category and generate the defect distribution area of ​​the sprayed coating. S403: Call the defect distribution area of ​​the sprayed coating, analyze the correlation between the coordinate information, thickness non-uniformity classification and color deviation classification data of the abnormal area, calculate the degree of influence on the overall quality of the sprayed coating, and generate sprayed coating quality assessment data.

6. The method for detecting the quality of glass bottle surface coating according to claim 1, characterized in that, Based on the coating quality assessment data, the specific steps for generating coating defect location and classification data are as follows: distinguishing color deviation, thickness anomalies, and defect types, marking the corresponding out-of-limit categories, calculating the deviation range and deviation ratio, analyzing thickness offset and uniformity. S501: Based on the coating quality assessment data, extract the color deviation value, coating thickness value and defect feature value of the sprayed area, calculate the deviation of the data, filter out areas that exceed the average threshold, mark the corresponding out-of-limit category, and generate out-of-limit area classification label data. S502: Based on the classification and marking data of the out-of-limit areas, extract the out-of-limit area data according to color deviation, thickness anomaly and defect type, match the detection position of the spray coating according to the out-of-limit category, calculate the detection position deviation range and deviation ratio, and generate the spray coating anomaly detection coordinate data. S503: Call the coordinate data of the coating abnormality detection, calculate the color offset data for the color deviation area, analyze the thickness offset and thickness uniformity for the coating thickness abnormal area, and generate coating defect location classification data for the defect area.

7. The method for detecting the quality of glass bottle surface coating according to claim 6, characterized in that, The specific formula for matching the detection position of the sprayed coating based on the over-limit category is as follows: ； Calculate the detection position deviation range Detection position deviation ratio The specific formula is: ； Calculate the detection position deviation ratio Based on the category of exceeding limits, match the detection position of the sprayed coating, calculate the deviation range and deviation ratio of the detection position, and generate coordinate data for abnormal detection of the sprayed coating. in, This represents the range of detection position deviation. Representing the The x-coordinate of each out-of-limit region The x-coordinate represents the detection position of the matched coating layer. Representing the The ordinate of each out-of-limit region The ordinate represents the detection position of the matched coating layer. This represents the number of x-axis data points in the out-of-limit region. This represents the number of data points on the ordinate of the region exceeding the limit. Representing the The thickness value at the detection location. This represents the total number of thickness detection points. This represents the ratio of detection position deviation. This represents the length of the detection reference.

Images