Method and system for detecting quality of outer package using polyurethane coating

Abstract

The invention discloses a method and a system for detecting the quality of an outer package using a polyurethane coating, and relates to the technical field of outer package material detection. Comprising the following steps: acquiring spectral data of a polyurethane coating outer package; based on the spectral data, extracting coating material information of the outer package; according to the coating material and the discrimination index, correcting the light source attribute and generating correction data; distributing defect grades for the outer package of the polyurethane coating according to the correction data and the discrimination indexes; and obtaining projection data of the outer package of the polyurethane coating at the packaging position. According to the method, the wavelength-reflectivity data matrix is constructed, principal component extraction and curvature feature analysis are performed on the wavelength-reflectivity data matrix, the spectral feature vector representing the coating material is constructed, and then the feature vector is matched with the preset material database, so that the coating material is identified; the problem that in the prior art, the detection accuracy of high-reflectivity coatings such as polyurethane coatings is insufficient is solved, and the reliability of extracting the intrinsic characteristics of materials is remarkably improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of outer packaging material detection, in particular to an outer packaging quality detection method and system using polyurethane coating. BACKGROUND

[0002] The outer packaging quality of polyurethane coating is prone to be affected by environmental factors such as temperature, humidity and mechanical stress during storage and use, resulting in defects such as cracking, peeling, bubbling or discoloration of the coating, and thus affecting the sealing property, weather resistance and overall protection performance of the outer packaging. Therefore, a kind of outer packaging quality detection method and system using polyurethane coating is needed to detect.

[0003] Through retrieval, the Chinese invention patent application with the publication number "CN115078365A" discloses a "soft package printing quality defect detection method". The application obtains the mask information of different regions in the standard image through analysis and design, can well obtain each different image region, facilitates different methods for different characteristic regions, and performs different processing. The application can detect text and icon deformation defects, and can also detect small bubbles generated in the production process. The application has better detection effect and high practical value.

[0004] In addition, the Chinese invention patent application with the publication number "CN117723518A" discloses a "method for rapidly detecting the source of oil stains on cigarette outer packaging paper". The application collects machine oil at different positions of a cigarette machine, acquires two-dimensional and three-dimensional fluorescence spectra of the collected machine oil, extracts characteristic fluorescence peaks of the machine oil, changes excitation wavelength and slit width test conditions, acquires two-dimensional and one-dimensional fluorescence spectra of the machine oil at different positions under different excitation wavelength and slit width test conditions, extracts characteristic fluorescence peaks of the machine oil at different positions, establishes a fingerprint spectrum library of two-dimensional and three-dimensional fluorescence characteristic peaks of the machine oil at different positions, collects and extracts yellow spot oil stains on the tobacco paper, measures the spectrum of the fluorescence characteristic peaks of the yellow spot oil stains, compares the obtained spectrum with the collected fingerprint spectrum library of the fluorescence characteristic peaks, and determines the type of the yellow spot oil stains.

[0005] However, in actual application, the above two disclosed schemes and similar technologies mainly aim at the identification of specific printed patterns or oil stains, and lack high sensitivity detection capability for the unique high-gloss surface characteristics, complex spectral response and intrinsic defects of polyurethane coating. Secondly, under the condition of change of light source or environmental interference, the existing spectral analysis method is prone to introduce significant errors, and it is difficult to accurately distinguish between changes in the optical properties of the material itself and abnormal responses caused by real defects. In addition, for the structural integrity detection of the outer packaging position, the above technical scheme usually relies on two-dimensional image analysis of a single perspective, and fails to effectively combine three-dimensional projection data to quantitatively evaluate the packaging state. SUMMARY

[0006] The present application aims to provide a polyurethane coating-based outer packaging quality detection method and system to solve the problems presented in the background.

[0007] In a first aspect, to achieve the above object, the present application provides the following technical solution: a polyurethane coating-based outer packaging quality detection method, comprising:

[0008] Obtaining spectral data of the polyurethane coating-based outer packaging;

[0009] Extracting coating material information of the outer packaging based on the spectral data;

[0010] Correcting light source properties and generating corrected data according to the coating material and the discrimination index;

[0011] Assigning a defect grade to the polyurethane coating-based outer packaging according to the corrected data and the discrimination index;

[0012] Obtaining projection data of the polyurethane coating-based outer packaging at the packaging position;

[0013] Combining the defect grade and the projection data set to determine a product that is complete and has a qualified polyurethane coating quality.

[0014] As a further preferred embodiment of the present technical solution, the method for obtaining spectral data comprises:

[0015] Using a spectral acquisition device to perform whole-image scanning on the surface of the outer packaging coated with polyurethane coating to obtain a sequence of reflectance spectral images at different wavebands;

[0016] Performing waveband registration and spectral fusion on the sequence of reflectance spectral images to generate a high-resolution spectral cube data set corresponding to the outer packaging;

[0017] Using a region segmentation algorithm in the prior art to divide the outer packaging region into multiple detection sub-regions, extracting average spectral curves of each sub-region, and finally constructing a spectral data set from the average spectral curves of each sub-region.

[0018] As a further preferred embodiment of the present technical solution, the method for adjusting the spectral acquisition device comprises:

[0019] Obtaining the acquisition range of the spectral acquisition device;

[0020] Based on the area of the acquisition range, obtaining the average acquisition time of the spectral acquisition device;

[0021] According to the average acquisition time, adjusting the scanning speed and exposure parameters of the spectral acquisition device to ensure uniformity and consistency in data acquisition at different area regions;

[0022] The adjusted acquisition parameters are verified in real time and stored as device configuration parameters for subsequent optimization of the spectral data acquisition process.

[0023] As the further preferred technical solution, the coating material information acquisition method comprises:

[0024] Based on the region segmentation algorithm, a polyurethane coating outer package surface detection region is extracted, a plurality of pixel points are selected in the detection region, corresponding wavelength-reflectivity data is extracted, and a reflectivity matrix is formed;

[0025] The reflectivity matrix is transposed according to the wavelength dimension to obtain a wavelength sample matrix, and each row represents the reflectivity of a wavelength at a plurality of sample points;

[0026] The wavelength sample matrix is subjected to principal component analysis to extract score data of each wavelength in the principal component direction;

[0027] A two-dimensional coordinate system is established corresponding to the wavelength data and the principal component score data, wherein the abscissa is the wavelength data and the ordinate is the principal component score, and a two-dimensional function curve of wavelength-principal component score is constructed;

[0028] Numerical derivative calculation is performed on the function curve to obtain the curvature value at each wavelength point;

[0029] Based on the curvature value, the extreme point position, number and intensity information are extracted to construct a curvature feature vector;

[0030] The curvature feature vector is compared with the feature vectors of each material in the preset material database to obtain the material type with the highest matching degree with the target sample as the coating material information of the outer package.

[0031] As the further preferred technical solution, the correction data acquisition method comprises:

[0032] Based on the coating material information, the standard spectral reflectance data corresponding to the material is called in the Internet;

[0033] Obtain the light source spectral distribution data of the current spectral acquisition device under actual detection conditions to form actual light source reflection data;

[0034] The actual light source reflection data and the standard spectral reflection data are compared point by point according to the wavelength to calculate the reflection deviation factor at each wavelength;

[0035] Using the reflection deviation factor, the outer package spectral data collected is corrected wave by wave to obtain a corrected reflectivity matrix;

[0036] The corrected reflectivity matrix is normalized to form the correction data under the standard light source condition.

[0037] As a further preferred technical solution, the defect level assignment method comprises:

[0038] According to the correction data, the spectral characteristic parameters of the coating are extracted, including reflectivity peak value, valley position and spectral slope;

[0039] According to the preset threshold in the discrimination index, the spectral characteristic parameters are compared with the standard value point by point, and the characteristic deviation value is calculated;

[0040] Based on the size range of the characteristic deviation value, the defect level interval is divided, wherein the defect level includes no defect, slight defect, moderate defect and serious defect;

[0041] Combined with the coating material information, the material-dependent correction is performed on the divided defect level to generate the final defect level judgment result.

[0042] As a further preferred technical solution, the projection data acquisition method at the packaging position comprises:

[0043] Based on image processing technology, the outer packaging of methoxylamine hydrochloride is scanned, and the edge profile of the outer packaging of methoxylamine hydrochloride is captured;

[0044] Based on the edge detection algorithm, the contour line of the polyurethane coating outer packaging is recognized;

[0045] Based on image recognition technology, the contour line is analyzed, and the packaging position of the polyurethane coating outer packaging is obtained;

[0046] Based on the packaging position of the polyurethane coating outer packaging, a horizontal axis is constructed;

[0047] Any one of a plurality of light source nodes is set on the vertical top surface of the polyurethane coating outer packaging;

[0048] The projection data of the remaining methoxylamine hydrochloride outer packaging sealing edge light source node in the horizontal axis after irradiation is collected.

[0049] Secondly, in order to improve the above technical solution, a polyurethane coating outer packaging quality detection system is also proposed, which uses the above polyurethane coating outer packaging quality detection method, and comprises:

[0050] The spectral data acquisition module scans the surface of the polyurethane coating outer packaging through a spectral acquisition device, acquires a sequence of reflectance spectral images under different wavebands, and performs waveband registration and fusion on the image sequence to generate a high-resolution spectral data cube;

[0051] The spectral data preprocessing and region segmentation module divides the outer packaging region into a plurality of detection sub-regions by a region segmentation algorithm, extracts the average spectral curve of each sub-region, and forms a complete spectral data set;

[0052] The coating material identification module processes the reflectivity matrix by using a principal component analysis method, extracts a wavelength-principal component score function curve, extracts a characteristic vector through curvature analysis, compares the characteristic vector with a material database, and identifies coating material information of the polyurethane coating;

[0053] The light source correction module is used for calling standard spectral reflection data, obtaining a current actual light source spectral distribution, calculating a reflection deviation factor, correcting spectral data collected wave by wave, obtaining a corrected reflectivity matrix, and performing normalization processing;

[0054] The defect grade determination module is used for extracting spectral characteristic parameters from the corrected data, calculating characteristic deviations based on a preset threshold, dividing defect grades, and performing material dependency correction in combination with coating material information, and outputting a final defect grade;

[0055] The packaging position and projection data acquisition module scans an outer package by using image processing and edge detection technology, identifies a packaging position, constructs a horizontal axis, and acquires projection data of a sealing edge irradiated by a light source node;

[0056] The determination module combines a defect grade determination result and packaging position projection data, determines whether the outer package is complete and whether the quality of the polyurethane coating meets a standard, and outputs a final determination result;

[0057] The device control and parameter management module is used for managing parameter adjustment, real-time parameter verification, and device configuration storage of a spectral acquisition device, and ensures consistency and uniformity of a spectral data acquisition process.

[0058] Compared with the prior art, the polyurethane coating quality detection method and system have the following beneficial effects:

[0059] The polyurethane coating quality detection method and system construct a wavelength-reflectivity data matrix, extract principal components and perform curvature characteristic analysis, construct a spectral characteristic vector representing coating material, match the characteristic vector with a preset material database, identify coating material, solve the problem of insufficient detection accuracy of high reflectivity coatings such as polyurethane coatings in the prior art, and significantly improve the reliability of material intrinsic characteristic extraction;

[0060] In addition, in view of detection errors that may be introduced by changes in ambient light sources, reflection deviation factors are calculated wave by wave by comparing standard spectral reflection data with light source distribution under actual collection environments, and spectral data is normalized and corrected, which reduces interference of environmental factors on spectral characteristics from the source, and significantly enhances consistency and stability of detection data;

[0061] In the aspect of defect detection, based on the corrected spectral data, multiple key parameters such as reflectivity peak value, trough position and spectral slope are extracted, multi-dimensional deviation analysis is carried out combined with preset discrimination threshold, fine division of defect grade is realized, at the same time, a correction model based on coating material dependence is introduced to modify the detection result, ensuring that the defect evaluation under different material conditions has pertinence and scientific nature, and the universality and adaptability of the detection system are significantly enhanced.

[0062] In the aspect of packaging structure integrity judgment, the packaging position contour is obtained through image processing and edge detection technology, and the three-dimensional projection data of the sealing area is collected by setting the light source array, the sealing integrity can be judged based on the construction of horizontal reference axis and edge projection curve analysis, and the quality evaluation of structure dimension is provided, which provides the basis for the overall qualification judgment of the outer package. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 The step flow chart for the disclosed application is illustrated;

[0064] Figure 2 The wavelength and main component score function curve graph of the polyurethane coating outer package surface disclosed by the application is illustrated;

[0065] Figure 3 The step flow chart for step S500 of the application is illustrated;

[0066] Figure 4 The system module composition diagram disclosed by the application is illustrated. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the application.

[0068] Before explaining the technical solutions of the application, it should be clear that the quality of polyurethane coating outer package is easily affected by temperature, humidity, mechanical stress and other environmental factors during storage and use. These factors can induce defects such as cracking, peeling, bubbling or discoloration of the coating, and thus damage the sealing, weather resistance and overall protection performance of the outer package. Therefore, a polyurethane coating outer package quality detection method and system are urgently needed.

[0069] REFERENCE Figure 1 It can be seen that a polyurethane coating outer package quality detection method comprises steps S100-S600.

[0070] Step S100: Obtain the spectral data of the polyurethane coating outer package.

[0071] It needs to be clear that the method of obtaining spectral data in step S100 includes steps S101-S103.

[0072] Specifically, step S101: through the spectral acquisition device, the surface of the polyurethane coating outer package is scanned for the whole image, and the reflection spectrum image sequence under different wave bands is obtained.

[0073] It is worth noting that the spectral acquisition device in step S101 is a snapshot hyperspectral camera in the prior art, which can capture reflection spectrum data in the 400-2500 nanometer wave band range with high resolution, and the reflection spectrum image sequence specifically refers to a series of image data continuously obtained by the high-resolution acquisition device in the 400-2500 nanometer wave band range, each image corresponds to the reflection intensity distribution of a specific wave band, and these sequence data completely record the spectral characteristic changes of the coating surface, which facilitates subsequent quantitative evaluation of coating uniformity, defect detection and thickness analysis.

[0074] Step S102: wave band registration and spectral fusion are performed on the reflection spectrum image sequence to generate a high-resolution spectral cube data set corresponding to the outer package.

[0075] It is worth noting that the wave band registration and spectral fusion in step S102 are realized by the image processing algorithm in the prior art, that is, the wave band registration method based on feature point matching and the spectral fusion technology of principal component analysis (PCA) are used to correct the spatial offset of the image sequence and integrate multi-wave band information, thereby generating a high-resolution spectral cube data set. Since this technical solution is a common technical solution known to those skilled in the art, the applicant does not make a specific description.

[0076] Step S103: the outer package area is divided into multiple detection sub-areas by using the region segmentation algorithm in the prior art, the average spectral curve of each sub-area is extracted, and finally the spectral data set is composed of the average spectral curves of each sub-area.

[0077] As one of the preferred embodiments, the present scheme can also be extended to adjust the parameters of the spectral acquisition device, specifically including: obtaining the acquisition range of the spectral acquisition device; based on the acquisition area, calculating the average acquisition time; according to the average acquisition time, adjusting the scanning rate and exposure parameters of the device to realize the uniformity and consistency of data acquisition in different area regions; real-time verification of the adjusted acquisition parameters, and storing them as device configuration parameters for subsequent optimization of the spectral data acquisition process. It needs to be further explained that the intelligent optimization of the acquisition parameters in this embodiment improves the spatial consistency and time sequence stability of the hyperspectral data, and guarantees the overall detection accuracy and efficiency.

[0078] Step S200: Extract the coating material information of the outer package based on the spectral data.

[0079] It should be noted that the method of obtaining the coating material information in step S200 includes steps S201-S207.

[0080] Step S201: Extract the polyurethane coating outer package surface detection area based on the region segmentation algorithm, select multiple pixel points in the detection area, extract the corresponding wavelength-reflectivity data, and form a reflectivity matrix.

[0081] Step S202: Transpose the reflectivity matrix by wavelength dimension to obtain a wavelength sample matrix, each row representing a reflection value of a wavelength at multiple sample points.

[0082] It should be noted that in step S202, transposing the reflectivity matrix by wavelength dimension specifically refers to implementing it through the matrix transposition operation in the prior art, that is, interchanging the rows (representing pixel points) and columns (representing wavelengths) of the original reflectivity matrix, thereby generating a wavelength sample matrix, where each row represents a reflection value sequence of a specific wavelength at multiple pixel points, facilitating subsequent feature extraction and statistical analysis based on wavelength dimension.

[0083] Step S203: Perform principal component analysis on the wavelength sample matrix to extract the score data of each wavelength in the principal component direction.

[0084] It should be noted that in step S203, the principal components in the wavelength sample matrix include brightness intensity, wavelength reflection number, spectral slope, and reflection peak value. These principal components are used to capture the main variation pattern in the spectral data, where brightness intensity reflects the overall reflection level, wavelength reflection number indicates the response density of a specific wavelength region, spectral slope describes the trend of reflectivity with wavelength, and reflection peak value identifies the key reflection position. Through principal component analysis, the dimension can be effectively reduced and the score data of each wavelength can be extracted, facilitating subsequent quantitative evaluation of the uniformity, defects, and aging degree of the coating material.

[0085] Step S204: Corresponding two-dimensional coordinate system is established between wavelength data and principal component score data, where the abscissa is the wavelength data and the ordinate is the principal component score, and a two-dimensional function curve of wavelength-principal component score is constructed.

[0086] Step S205: Numerical derivative calculation is performed on the function curve to obtain the curvature value at each wavelength point.

[0087] Step S206: Extract the extreme point position, number, and intensity information based on the curvature value, and construct a curvature feature vector.

[0088] Step S207: Comparing the curvature feature vector with the feature vectors of each material in the preset material database, obtaining the material type with the highest matching degree with the target sample as the coating material information of the outer package.

[0089] As a supplement to the content of steps S201 to S207, refer to Figure 2 It can be seen that, by matching the wavelength data with the corresponding principal component score data, a two-dimensional function curve of wavelength-principal component score is constructed to realize the intuitive expression and analysis of the spectral characteristics of the polyurethane coating outer package surface coating.

[0090] Specifically, first, the reflectance spectrum data of the coating outer package surface in the 400-2500 nanometer waveband is collected, and principal component analysis (PCA) is performed according to step S203 to extract the principal component scores corresponding to each wavelength point. Subsequently, a wavelength (unit: nanometer) is taken as the abscissa and the principal component score as the ordinate to draw the wavelength-principal component score function curve of the polyurethane coating outer package surface, which represents the spectral reflection characteristics of the coating at different wavelengths. It should be noted that, Figure 2 The curve intuitively presents the principal component scores corresponding to different wavelengths. The principal component scores comprehensively reflect the spectral brightness, slope, and reflection peak characteristics at the corresponding wavelengths. The peak region of the curve usually corresponds to the characteristic absorption peak or reflection peak of the coating material, which is an important basis for identifying material composition and detecting defects. It should be noted that, Figure 2 The curve in the figure uses a gradient color (color spectrum from blue to red) to represent the continuous change of wavelength from low to high. The gradient of the color helps to visually distinguish different wavebands, making it easier to observe the subtle differences and trends of spectral characteristics in a specific interval. In addition, the horizontal dashed line running through the curve in the figure represents the arithmetic mean of the principal component scores, which is calculated from the principal component scores of all wavelengths. This average line serves as a reference for evaluating the relative height of the scores in the local region of the curve, assisting in quickly locating reflection abnormalities or characteristic wavebands. Finally, refer to Figure 2 It can be seen that the peak (local maximum) of the curve corresponds to a wavelength with high spectral reflectance, which represents the characteristic of the key chemical composition in the coating or the change in the thickness of the coating.

[0091] Step S300: Correcting the light source properties and generating corrected data according to the coating material and the discrimination index.

[0092] It should be noted that the method for obtaining corrected data in step S300 includes steps S301-S305.

[0093] Step S301: Based on the coating material information, the standard spectral reflectance data corresponding to the material is called from the Internet.

[0094] Step S302: Obtain the light source spectral distribution data of the current spectral acquisition device under the actual detection condition, and form actual light source reflection data.

[0095] Step S303: Perform wavelength point-by-point comparison between the actual light source reflection data and the standard spectral reflection data, and calculate the reflection deviation factor at each wavelength.

[0096] It should be noted that the calculation method of the reflection deviation factor in step S303 is to calculate the difference between the reflectivity values of the actual light source reflection data and the standard spectral reflection data at the same wavelength point, and then take the ratio of the difference and the standard reflectivity value as the reflection deviation factor at the wavelength. This factor is used to quantify the difference between the actual light source and the standard light source.

[0097] Step S304: Use the reflection deviation factor to correct the collected outer packaging spectral data by wave band, and obtain the corrected reflectivity matrix.

[0098] Step S305: Normalize the corrected reflectivity matrix to form the correction data under the standard light source condition.

[0099] Step S400: Assign defect levels to the polyurethane coating outer packaging according to the correction data and the discrimination index. It should be noted that the discrimination index in steps S300 and S400 is a manually determined value.

[0100] In addition, it should be noted that the defect level assignment method in step S400 includes steps S401-S404.

[0101] Step S401: Extract the spectral feature parameters of the coating layer according to the correction data, including the reflectivity peak value, the valley position, and the spectral slope.

[0102] It should be noted that the correction data obtained in step S300 is the correction of the spectral data set in step S100, so the spectral feature parameters, including the reflectivity peak value, the valley position, and the spectral slope, can be extracted in step S401.

[0103] Step S402: According to the pre-set threshold in the discrimination index, perform point-by-point comparison between the spectral feature parameters and the standard value, and calculate the feature deviation value.

[0104] Step S403: Based on the size range of the feature deviation value, divide the defect level interval, wherein the defect level includes no defect, slight defect, moderate defect, and severe defect.

[0105] Step S404: Combine the coating material information to perform material-dependent correction on the divided defect level, and generate the final defect level judgment result.

[0106] It should be noted that steps S401 to S404 are mainly used to achieve comprehensive quantitative evaluation and quality control of polyurethane coating packaging defects in the present embodiment. Specifically, the spectral characteristic parameters (such as reflectance peak value, trough position and spectral slope) extracted in step S401 directly reflect the physical and chemical properties of the coating. After point-by-point comparison with the standard values in the preset discrimination index, the characteristic deviation value is generated in step S402. The deviation value, which is obtained by calculating the difference between the actual spectrum and the standard spectrum, provides an objective basis for defect classification. In step S403, according to the size range of the characteristic deviation value (such as no defect when the deviation is less than 5%, slight defect when the deviation is between 5% and 10%, moderate defect when the deviation is between 10% and 20%, and severe defect when the deviation is greater than 20%), the consistency and operability of the evaluation results are ensured. Step S404 optimizes and adjusts the preliminary classified defect levels through material-dependent correction combined with the coating material information obtained in step S200. For example, for polyurethane coatings on high-reflectivity materials such as metal substrates, the system will appropriately lower the deviation threshold of the reflection peak value to avoid misjudgment. For low-reflectivity materials such as plastic substrates, the weight of the spectral slope is increased to improve the sensitivity to subtle defects. Finally, accurate and reliable defect level judgment results are generated.

[0107] Step S500: Obtain projection data of the polyurethane coating packaging at the packaging position.

[0108] It should be noted that with reference to Figure 3 It can be seen that the method for obtaining projection data at the packaging position in step S500 includes the following detailed operations: steps S501 to S506.

[0109] Step S501: Based on image processing technology, use a high-resolution imaging device to scan the metaraminol bitartrate packaging, capture its surface features and accurately extract the edge profile, and generate preliminary profile data.

[0110] Step S502: Based on edge detection algorithms such as Canny operator, process the scanning data to identify the continuous contour line of the polyurethane coating packaging, filter noise and enhance edge information.

[0111] Step S503: Based on image recognition technology, perform geometric analysis on the contour line obtained in step S502, including angle and curvature calculation, and locate the accurate packaging position coordinates of the polyurethane coating packaging.

[0112] Step S504: Based on the packaging position coordinates of the polyurethane coating packaging, construct a horizontal axis in three-dimensional space as a reference datum for alignment and measurement of projection data.

[0113] Step S505: Obtain any one of the light source nodes in the polyurethane coating outer package vertical top surface, adjust the light source angle and intensity to ensure uniform irradiation of the packaging area.

[0114] Step S506: Collect the projection data of the remaining metaraminol bitartrate outer package seal edge after irradiation by the light source node in the horizontal axis, record the shadow distribution using an optical sensor, and convert it into a digital signal for storage and processing.

[0115] Step S600: Integrate the defect level and the projection data set to determine the product with complete outer packaging and polyurethane coating quality.

[0116] It should be noted that in step S600, the determination of the polyurethane coating quality product specifically includes the following detailed operation method: steps S601 to S605.

[0117] Step S601: Based on the defect level data generated in step S400, extract the weight coefficient corresponding to each defect level, and the high defect corresponds to the high coefficient weight.

[0118] Step S602: Obtain the projection data set collected in step S500, including the shadow distribution data at the packaging position, extract the projection length, shadow density and edge alignment degree and other parameters through image processing technology, and form the projection feature vector.

[0119] Step S603: Combine the defect level weight coefficient and the projection feature vector, and apply the linear weighting algorithm to calculate the comprehensive quality score, specifically multiply the defect weight and the projection parameter and sum to generate the normalized quality evaluation value.

[0120] Step S604: According to the preset quality threshold range, compare the comprehensive quality score with the standard value, wherein the threshold includes no defect (score ≥ 0.9), slight defect (0.7 ≤ score < 0.9), moderate defect (0.5 ≤ score < 0.7) and severe defect (score < 0.5), and determine whether the product meets the standard.

[0121] Step S605: If the comprehensive quality score reaches the no defect or slight defect level, it is determined that the outer packaging is complete and the polyurethane coating quality meets the standard, and the final qualified mark is generated, otherwise, it is marked as unqualified product, and the detailed detection report is output for subsequent quality traceability.

[0122] As shown in Figure 4 The present application also proposes and constructs an outer packaging quality detection system based on polyurethane coating, which is realized based on the polyurethane coating outer packaging quality detection method, and aims to comprehensively evaluate the integrity of the outer packaging and the coating quality. The system includes multiple functional modules, specifically as follows:

[0123] The spectral data acquisition module systematically scans the polyurethane coating outer packaging surface through a high-precision spectral acquisition device system, collects a sequence of reflectance spectral images under different wavebands, and uses image processing technology to accurately perform waveband registration and pixel-level fusion on the sequence, and finally generates a high-resolution spectral data cube, ensuring the comprehensiveness and accuracy of data acquisition.

[0124] The spectral data preprocessing and region segmentation module uses an intelligent region segmentation algorithm to finely divide the outer packaging region into multiple uniform detection sub-regions, and extracts the average spectral curve of each sub-region through calculation to form a complete standardized spectral data set, providing structured input for subsequent analysis.

[0125] The coating material identification module uses principal component analysis to process the reflectance matrix, extracts the wavelength-principal component score function curve, and identifies key feature points through curvature analysis algorithm, generates a feature vector, and performs efficient comparison with the built-in material database to accurately identify the coating material type and attribute information of the polyurethane coating.

[0126] The light source correction module calls the standard spectral reflectance database to obtain reference data, simultaneously monitors and obtains the spectral distribution parameters of the current actual light source, calculates the reflection deviation factor, dynamically corrects the collected spectral data by waveband, generates the corrected reflectance matrix, and performs normalization processing to eliminate environmental interference.

[0127] The defect grade determination module extracts key spectral feature parameters such as reflectance peak value and wavelength shift from the corrected data, calculates the feature deviation value based on the preset threshold model, divides the defect grade through classification algorithm, and performs material-dependent correction combined with the coating material information, and finally outputs the quantitative defect grade result.

[0128] The packaging position and projection data acquisition module uses image processing technology and edge detection algorithm to comprehensively scan the outer packaging, accurately identify the packaging position, construct a horizontal reference axis, and simulate the projection data of the sealing edge after the light source node is irradiated, generating geometric position information.

[0129] The determination module integrates the defect grade determination result and the packaging position projection data, performs comprehensive evaluation through a logic rule engine, determines whether the outer packaging structure is complete and whether the polyurethane coating quality meets the standards, and finally outputs an intuitive determination report.

[0130] The device control and parameter management module uniformly manages the parameter adjustment, real-time parameter verification mechanism and device configuration storage function of the spectral acquisition device, ensures the consistency and uniformity of the spectral data acquisition process, and supports efficient operation of the system.

[0131] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the embodiments disclosed except insofar as recited in the claims.

Claims

1. A method for detecting the quality of outer packaging using polyurethane coating, characterized in that: include: Obtain spectral data of polyurethane coating outer packaging; extracting coating material information of the outer packaging based on the spectral data; Correct the light source properties and generate correction data based on the coating material and discrimination index; Assign defect levels to polyurethane coating packaging based on correction data and discrimination indicators; Acquire projection data of the polyurethane coating outer packaging at the packaging position; Combining defect levels with projection datasets, we determine which products have intact outer packaging and meet polyurethane coating quality standards.

2. The method for detecting outer packaging quality using polyurethane coating according to claim 1, characterized in that: Methods for acquiring spectral data include: The entire surface of the outer packaging coated with polyurethane coating is scanned by spectral acquisition equipment to obtain a sequence of reflection spectrum images in different bands; Perform band registration and spectral fusion on the reflectance spectrum image sequence to generate a high-resolution spectral cube dataset corresponding to the outer packaging; The outer packaging area is divided into multiple detection sub-areas using the region segmentation algorithm in the prior art, and the average spectral curve of each sub-area is extracted. Finally, the spectral data set is formed by the average spectral curves of each sub-area.

3. The method for detecting the quality of outer packaging using polyurethane coating according to claim 2, characterized in that: The adjustment method of the spectrum acquisition equipment includes: Obtain the collection range of the spectrum collection device; Based on the acquisition range area, obtain the average acquisition time of the spectrum acquisition device; Adjust the scanning rate and exposure parameters of the spectral acquisition equipment according to the average acquisition time to ensure uniformity and consistency of data acquisition in different area areas; The adjusted acquisition parameters are verified in real time and stored as device configuration parameters for optimization of subsequent spectral data acquisition processes.

4. The method for detecting the quality of outer packaging using polyurethane coating according to claim 2, characterized in that: Methods for obtaining coating material information include: Extract the detection area on the surface of the polyurethane coating outer packaging based on the region segmentation algorithm, select multiple pixel points in the detection area, extract the corresponding wavelength-reflectivity data, and form a reflectivity matrix; Transposing the reflectivity matrix according to the band dimension to obtain a band sample matrix, where each row represents the reflectance value of a wavelength at multiple sample points; Perform principal component analysis on the band sample matrix and extract the score data of each wavelength in the principal component direction; A two-dimensional coordinate system is established by corresponding the wavelength data and the principal component score data, where the horizontal axis is the wavelength data and the vertical axis is the principal component score, and a two-dimensional function curve of wavelength-principal component score is constructed; Performing numerical derivative calculation on the function curve to obtain the curvature value at each wavelength point; Extracting the position, quantity and intensity information of the extreme points based on the curvature value, and constructing a curvature feature vector; The curvature feature vector is compared with the feature vectors of each material in a preset material database to obtain the material type with the highest matching degree with the target sample as the coating material information of the outer packaging.

5. The method for detecting the quality of outer packaging using polyurethane coating according to claim 1, characterized in that: Methods for obtaining correction data include: Based on the coating material information, calling the standard spectral reflectance data corresponding to the material on the Internet; Obtain the light source spectrum distribution data of the current spectrum acquisition device under actual detection conditions to form actual light source reflection data; Comparing the actual light source reflection data with the standard spectrum reflection data point by point in wavelength, and calculating the reflection deviation factor at each wavelength; Using the reflection deviation factor, the collected outer packaging spectral data is corrected band by band to obtain a corrected reflectance matrix; The corrected reflectivity matrix is ​​normalized to form correction data under standard light source conditions.

6. The method for detecting the quality of outer packaging using polyurethane coating according to claim 1, characterized in that: Methods for assigning defect levels include: Extracting spectral characteristic parameters of the coating based on the corrected data, including reflectivity peak, trough position and spectral slope; According to the preset threshold value in the discrimination index, the spectral characteristic parameters are compared with the standard value point by point to calculate the characteristic deviation value; Based on the size range of the characteristic deviation value, the defect level range is divided into no defect, slight defect, moderate defect and severe defect; Combined with the coating material information, the material dependency correction is performed on the divided defect level to generate the final defect level determination result.

7. The method for detecting the quality of outer packaging using polyurethane coating according to claim 1, characterized in that: Method for obtaining projection data at the package position: Scan the outer packaging of meta-hydroxylamine bitartrate based on image processing technology and capture the edge contour of the outer packaging of meta-hydroxylamine bitartrate; Identify the contour lines of polyurethane coating packaging based on edge detection algorithms; Analyze the contours based on image recognition technology and obtain the packaging position of the polyurethane coating outer packaging; The horizontal axis is constructed based on the packaging position of the polyurethane coating outer packaging; Obtain any one of a plurality of light source nodes that is set on the vertical top surface of the polyurethane coating outer packaging; Collect projection data on the horizontal axis of the sealing edge of the outer packaging of the remaining meta-hydroxylamine bitartrate after being irradiated by the light source node.

8. A system for inspecting the quality of outer packaging using polyurethane coating, using the method for inspecting the quality of outer packaging using polyurethane coating according to any one of claims 1 to 7, characterized in that: include: The spectral data acquisition module scans the surface of the polyurethane coating outer packaging through a spectral acquisition device, collects reflection spectrum image sequences in different bands, and performs band registration and fusion on the image sequences to generate a high-resolution spectral data cube; The spectral data preprocessing and region segmentation module divides the outer packaging area into multiple detection sub-regions through the region segmentation algorithm, extracts the average spectral curve of each sub-region, and forms a complete spectral data set; The coating material identification module uses principal component analysis to process the reflectance matrix, extract the wavelength-principal component score function curve, extract the feature vector through curvature analysis, compare the feature vector with the material database, and identify the coating material information of the polyurethane coating; The light source correction module is used to call the standard spectral reflectance data and obtain the current actual light source spectral distribution, calculate the reflection deviation factor, correct the collected spectral data band by band, obtain the corrected reflectance matrix and perform normalization processing; The defect level determination module is used to extract spectral characteristic parameters from the corrected data, calculate the characteristic deviation based on the preset threshold, classify the defect level, perform material dependency correction based on the coating material information, and output the final defect level; The packaging position and projection data acquisition module uses image processing and edge detection technology to scan the outer packaging, identify the packaging position, construct the horizontal axis, and obtain the projection data of the sealing edge after the light source node is illuminated; The judgment module combines the defect level judgment results with the package position projection data to determine whether the outer packaging is intact and whether the polyurethane coating quality meets the standards, and outputs the final judgment result; The device control and parameter management module is used to manage the parameter adjustment, real-time parameter verification, and device configuration storage of the spectral acquisition equipment to ensure the consistency and uniformity of the spectral data acquisition process.

Images