Automatic sealing detection method and system for plastic packaging barrel

Abstract

The invention provides an automatic sealing detection method and system for a plastic packaging barrel. The method comprises the following steps: acquiring infrared thermal imaging data and beat pulse signals of a sealing area of a plastic packaging barrel, and generating a synchronous infrared thermal sequence by using time sequence calibration; fusing long and short wave energy to eliminate steam interference to form a thermal field distribution body, and performing axial gradient sampling to obtain a dynamic thermal flow vector group; synchronously obtaining a visible light image, calculating a curvature compensation coefficient based on the barrel-shaped contour, reconstructing a three-dimensional model, and extracting wrinkle features; and registering heat flow and wrinkles under a curvature coordinate system, analyzing and identifying an annular cold region and a wrinkle fracture zone through a thermal field, and judging incomplete fusion or stress cracking defects according to position correlation and thermal diffusion parameter deviation. Through multi-source data fusion and space registration technologies, automatic detection of the sealing defects of the plastic packaging barrel is realized, and the problems of incomplete fusion and stress cracking are accurately identified.

Description

Technical Field

[0001] This application relates to the field of industrial automation nondestructive testing technology, and in particular to an automatic sealing detection method and system for plastic packaging barrels. Background Technology

[0002] In automated filling production lines, the heat-sealed plastic packaging barrels must meet the sealing performance testing requirements under high-speed continuous operation. Due to interference from residual vapor after heat sealing, deformation of irregular curved surfaces of the barrel, and thermal stress concentration effects, the testing method must have anti-interference capabilities and be able to simultaneously identify incomplete fusion defects and stress cracks under dynamic conditions.

[0003] The current mainstream in the industry adopts a dynamic infrared thermal imaging online detection system. Its specific implementation process includes three steps: thermal image sequence acquisition, planar temperature mapping, and binarized morphology determination. A mid-wave infrared camera is deployed at the conveyor belt station to capture the transient temperature field changes in the sealing area; the curved surface of the barrel is projected onto a two-dimensional plane to generate a normalized thermal difference map; the contour of the temperature anomaly area is extracted through adaptive threshold segmentation, and defects are determined according to preset area and location rules.

[0004] The scheme has three inherent limitations: the vapor scattering effect causes the thermal image boundary to diffuse, resulting in the positioning deviation of the unfused area; the two-dimensional plane mapping compresses the axial curvature of the barrel, causing the thermal flow gradient direction of the stress crack to be distorted; and the single temperature criterion is divorced from the thermal stress wrinkling evolution mechanism, frequently misjudging the normal thermal relaxation phenomenon in the cooling process. Summary of the Invention

[0005] This application provides an automatic sealing detection method and system for plastic packaging barrels to solve the problems in the prior art.

[0006] In a first aspect, this application provides an automatic sealing detection method for plastic packaging drums, including: Infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device are collected. The beat pulse signal is used as the time reference to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations to generate a timestamp synchronized infrared thermal sequence. By fusing the long and short wave energy information in the infrared thermal sequence, the image blurring interference caused by residual heat-sealing vapor is eliminated, and a thermal field distribution with spatial coordinates and time dimension is generated. The thermal field distribution is then sampled along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector group. Simultaneously acquire visible light images of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel outline in the visible light image, reconstruct a three-dimensional model of the sealing surface based on the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. In the curvature compensation coordinate system established according to the barrel curvature compensation coefficient, the dynamic heat flow vector group and the fold distribution characteristics are spatially registered, and the annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the fold fracture zone that is mismatched with the stress direction of the barrel wall are identified by thermal field analysis. Based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group, it is determined that the seal has incomplete fusion defects or stress cracking defects.

[0007] Optionally, in a curvature compensation coordinate system established based on the barrel curvature compensation coefficient, the dynamic heat flow vector set and the fold distribution characteristics are spatially registered, and annular cold zones deviating from the reference thermal diffusion parameters of the ellipse major axis and fold fracture zones mismatched with the barrel wall stress direction are identified through thermal field analysis, including: A three-dimensional curvature compensation coordinate system is established based on the curvature compensation coefficient of the plastic packaging barrel. The heat flow direction component in the dynamic heat flow vector group is mapped to the corresponding position in the curvature compensation coordinate system, and the concave and convex region distribution contour map in the fold distribution shape feature is aligned to the curvature compensation coordinate system; A standard thermal diffusion channel is selected as a reference along the major axis of the ellipse, and the average thermal diffusion intensity value at each point in the diffusion channel is calculated as the baseline. Identify regions in the curvature compensation coordinate system where the heat diffusion intensity is consistently below the baseline threshold and exhibits a closed-loop distribution, and mark them as annular cooling zones. The fractured uneven area in the outline of the uneven area distribution map that deviates from the main shrinkage direction of the plastic packaging barrel wall by more than a preset angle is marked as a wrinkle fracture area.

[0008] Optionally, based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature-compensated coordinate system and the deviation of anisotropic thermal diffusion parameters in the dynamic heat flow vector set, it is determined that the seal has incomplete fusion defects or stress cracking defects, including: Calculate the minimum interval distance from the geometric center point of the annular cold zone to the boundary of the fold fracture zone in the curvature compensation coordinate system; The distribution of the degree of deviation between each point of the dynamic heat flow vector within the annular cold zone and the reference heat diffusion value along the major axis of the ellipse is statistically analyzed. Extract the angular distribution characteristics of the heat diffusion direction and the principal stress direction of the barrel wall at each point in the area covered by the folded fracture zone; When the minimum interval distance is lower than the preset proximity threshold and the proportion of high deviation points in the deviation distribution within the annular cold zone exceeds the dominant proportion, and the direction of most points in the included angle distribution characteristics is basically consistent with the direction of the principal stress of the barrel wall, the non-fusion defect is determined to be established. The stress cracking defect is determined when the minimum interval distance exceeds the preset separation relationship threshold, the low deviation value points occupy the main proportion in the deviation distribution of the annular cold zone, and the points that significantly deviate from the principal stress direction of the barrel wall in the included angle distribution characteristics form a concentrated distribution group.

[0009] Optionally, by fusing long-wave and short-wave energy information from the infrared thermal sequence to eliminate image blurring interference caused by residual heat-sealing vapor, a thermal field distribution with spatial coordinates and time dimensions is generated, including: A set of key points with location identifiers is extracted from the long-wave and short-wave thermal signature quantitative maps in the fused infrared thermal sequence, and a spatiotemporal correlation model containing the location of key points and their corresponding time scale points is constructed. The energy values ​​of long and short wave key points marked at the same location are cross-compared. When the long wave energy value exceeds the preset ratio threshold of the short wave energy value, it is determined that there is heat sealing steam interference at the location and the corresponding key point is excluded. The key points that were not excluded were sorted in order of time scale, and the energy values ​​of each time point were associated with the location identifier to form a three-dimensional energy value matrix; The time interval is divided into layers based on the beat pulse of the transmission device. Within each interval layer, the three-dimensional energy value matrix is ​​mapped to a three-dimensional spatial coordinate grid to form a multi-layer thermodynamic spatial unit. The multiple thermal spatial units are superimposed along the time axis to form a thermal field distribution with spatial coordinates and time scale.

[0010] Optionally, gradient sampling is performed on the heat field distribution body along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector set, including: The axial direction of the plastic packaging barrel is determined as the main reference axis on the heat distribution body, and measurement point lines are set along the main reference axis at fixed intervals. Each measurement point line contains heat measurement points arranged at equal intervals. For each thermal measurement point on the measurement point line, calculate the heat change ratio of the point at the same axial position as the previous measurement point line, and continuously record the heat change ratio of the same axial position along the time axis to form a change intensity curve with time scale. The variation intensity curves of all axial positions on each measurement point line are integrated into the heat flow direction components of the corresponding time series points, and the heat flow direction components of all time series points are collected according to the beat pulse time series to form a dynamic heat flow vector group.

[0011] Optionally, a visible light image of the sealing area is acquired synchronously, and a barrel curvature compensation coefficient is calculated based on the elliptical barrel contour in the visible light image, including: In a visible light image, mark multiple equally spaced positioning points on the outer edge of the plastic packaging barrel opening, and connect all positioning points to form a closed outer edge connection loop; Measure the maximum spacing value in the major axis direction and the minimum spacing value in the minor axis direction of the connecting rings of the outer edge, and calculate the geometric difference between the maximum spacing value in the major axis and the minimum spacing value in the minor axis as a reference value for the degree of bending; Using the minimum spacing of the minor axis as the reference unit, the ratio of the curvature reference value to the reference unit is calculated, and the difference between the ratio value and the standard circular reference value is used to calculate the barrel curvature compensation coefficient, which represents the curvature of the barrel surface.

[0012] Optionally, a three-dimensional model of the sealing surface is reconstructed based on the barrel curvature compensation coefficient, and the wrinkle distribution characteristics under thermal stress are extracted, including: Input the barrel curvature compensation coefficient, which represents the degree of curvature of the barrel surface, into the preset surface reconstruction model to generate the compensated barrel covering surface structure model. On the covering surface structure model, dense convex and concave unit blocks are divided along the boundary of the sealing area, and the vertical position difference between the center point of each convex and concave unit block and its four adjacent boundary points is calculated. When the vertical position difference is greater than the preset smoothness threshold, the corresponding unit block is determined to be a surface bump. The boundary positions of the concentrated distribution areas of the surface bumps are statistically analyzed to form a bump distribution profile map. The distance change rate curve between adjacent bumps in the bump distribution profile map is calculated as the shape feature of the fold distribution.

[0013] Secondly, this application provides an automatic sealing and detection system for plastic packaging drums, comprising: The generation module is used to collect infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device, and to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations using the beat pulse signal as a time reference, so as to generate an infrared thermal sequence with timestamp synchronization. The generation module is also used to fuse the long and short wave energy information in the infrared thermal sequence, eliminate image blurring interference caused by residual heat seal vapor, generate a three-dimensional thermal field distribution with spatial coordinates and time dimension, perform gradient sampling on the three-dimensional thermal field distribution along the axial direction of the plastic packaging barrel, and generate a dynamic heat flow vector group. The calculation module is used to synchronously acquire the visible light image of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel contour in the visible light image, reconstruct the three-dimensional model of the sealing surface according to the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. The registration module is used to spatially register the dynamic heat flow vector group with the wrinkle distribution characteristics in a curvature compensation coordinate system established according to the barrel curvature compensation coefficient. It identifies the annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the wrinkle fracture zone that is mismatched with the stress direction of the barrel wall through thermal field analysis. The judgment module is used to determine whether there is a lack of fusion or stress cracking defect in the seal based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group.

[0014] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement an automatic sealing detection method for plastic packaging barrels as described in the first aspect above.

[0015] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements an automatic sealing detection method for plastic packaging drums as described in the first aspect.

[0016] This application uses the acquisition of beat pulse signals as a time reference for multi-station infrared thermal imaging timing calibration, which can eliminate the temporal misalignment caused by conveyor belt jitter and establish continuous observation of thermal evolution across stations; by fusing long and short wave energy to generate a spatiotemporal thermal field distribution and axial gradient sampling, it can suppress vapor scattering noise and extract three-dimensional heat conduction anisotropy features; by reconstructing a three-dimensional model based on the barrel contour of visible light images to extract wrinkle features, it can compensate for elliptical curvature deformation and quantify the microscopic deformation distribution caused by thermal stress; by registering heat flow vectors and wrinkle features in a curvature compensation coordinate system, it can couple heat conduction and mechanical stress fields and establish a dual-modal correlation mapping of defects; by analyzing the spatial correlation and thermal diffusion deviation between the annular cold zone and the wrinkle fracture zone, it can distinguish the differences in the physical mechanisms of non-fusion (thermal blockage) and stress cracking (thermal mismatch).

[0017] Furthermore, based on the curvature-compensated coordinate system, the dynamic heat flow vector is mapped to the corresponding position in three-dimensional space, and the folded contour map is simultaneously aligned. A baseline is established by selecting a standard heat diffusion channel along the major axis of an ellipse, and a closed-loop distribution of a continuously low-temperature zone is identified as an annular cooling zone. Fractured folded zones with excessive angular deviation from the main contraction direction of the barrel wall are detected as folded fracture zones. The annular cold zone, abnormal thermal blockage caused by non-fusion, and structural fracture caused by stress mismatch in folded fracture zones are accurately separated in three-dimensional space. By using two physical quantities—temperature gradient and stress direction deviation—for joint judgment, misjudgment by a single parameter is avoided, and the accuracy of identifying micron-level internal defects is improved.

[0018] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of an automatic sealing detection method for plastic packaging drums provided in this application is shown; Figure 2 A schematic diagram of a scenario for an automatic sealing detection method for plastic packaging drums provided in this application is shown; Figure 3 This application provides a schematic diagram of the structure of an automatic sealing and detection system for plastic packaging drums. Figure 4 A schematic diagram of the structure of a computing device provided in this application is shown. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0022] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0023] Current automatic sealing and inspection technologies for plastic packaging drums rely on inherent limitations in mainstream infrared thermal imaging schemes: steam interference causes dispersion at the thermal image boundary, leading to positioning errors in unfused areas; two-dimensional planar mapping compresses the axial curvature of the drum, distorting the direction of the thermal flow gradient for stress cracks; and the reliance on a single temperature criterion deviates from the thermal stress wrinkling evolution mechanism, frequently misjudging normal thermal relaxation phenomena during the cooling process. These defects stem from spatial characterization distortions, such as uncompensated surface deformation and thermodynamic field decoupling, neglect of mechanical stress correlation and insufficient suppression of dynamic interference, and failure to eliminate steam noise, resulting in persistently high rates of missed and false alarms for micron-level defects.

[0024] To address the shortcomings of existing technologies, this invention proposes a multi-source spatiotemporal fusion-based method for detecting sealing defects in plastic packaging barrels. This method synchronously calibrates multi-station infrared thermal imaging sequences using transmission pulses as the time reference, fusing long and short wave energy to construct a spatiotemporal thermal field distribution resistant to steam interference, and generates dynamic heat flow vector sets through gradient sampling along the barrel's axial direction. Simultaneously, it extracts the elliptical barrel contour from visible light images and reconstructs a three-dimensional surface model based on curvature compensation coefficients to quantify wrinkle features. Spatially registering the heat flow vectors and wrinkle features in a curvature compensation coordinate system, it couples and analyzes the correlation between thermal diffusion deviation in the annular cold zone and stress mismatch in the wrinkle fracture zone, achieving accurate joint determination of incomplete fusion defects and stress cracking defects. This method overcomes the limitations of traditional single-temperature field analysis. Through the synergistic analysis of heat conduction and mechanical stress fields, it significantly improves the accuracy and reliability of identifying micron-level internal defects while eliminating steam interference and compensating for surface distortion.

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] Figure 1 This application provides a flowchart of an automatic sealing detection method for plastic packaging drums, as shown in the embodiments below. Figure 1 As shown, the method includes: 101. Collect infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device, and use the beat pulse signal as the time reference to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations to generate an infrared thermal sequence with timestamp synchronization. In the above scheme, the clock pulse signal refers to the electrical signal triggered when the transmission device moves a fixed distance, used to mark the time reference point. Timing calibration is the process of adjusting data from different acquisition devices to a unified time axis through time compensation. Infrared thermal sequence refers to a set of infrared thermal images arranged sequentially after time alignment.

[0027] In this embodiment, the encoder of the conveyor first generates a beat pulse signal when it detects a fixed distance of movement on the conveyor belt, such as the distance between the plastic buckets. This signal is immediately sent to three consecutively arranged infrared cameras as a synchronization trigger command. When the infrared cameras receive the pulse signal, they synchronously capture thermal images of the sealing area. However, due to equipment response delays and transmission path differences, the actual shooting time of each camera deviates from the pulse reference time by milliseconds. To eliminate this time misalignment, the system automatically records the actual shooting timestamp of each camera and calculates the difference between it and the pulse reference time using a time interpolation algorithm: for cameras that shoot later than the reference time, the timestamp of their thermal image data is adjusted forward by the corresponding delay time; for cameras that shoot earlier than the reference time, the timestamp is adjusted backward. For example, when the reference pulse time is 10.000 seconds, if camera at station 2 actually shoots at 10.008 seconds, the system corrects its thermal image timestamp to 10.000 seconds; if camera at station 3 shoots at 9.999 seconds, it is corrected to 10.000 seconds. Finally, an infrared thermal sequence with strict alignment and uniform time axis is generated to ensure that the thermal conduction state captured at different workstations is at the same evolution stage, thus establishing an accurate time reference for subsequent dynamic analysis.

[0028] In practical applications, on production line A, the conveyor belt runs at a constant speed. When the encoder detects that the conveyor belt has moved 15 centimeters, it sends a pulse signal to Camera-1, Camera-2, and Camera-3. Camera-1 takes a picture 2 milliseconds after the pulse, Camera-2 takes a picture 8 milliseconds after the pulse due to transmission delay, and Camera-3 mistakenly takes a picture 1 millisecond before the pulse. Time compensation is used: the data from Camera-1 is delayed by 2 milliseconds, the data from Camera-2 is advanced by 8 milliseconds for correction, and the data from Camera-3 is delayed by 1 millisecond for adjustment. Finally, the three sets of thermal images are aligned to the pulse trigger time, forming a synchronized sequence.

[0029] This solution overcomes the time skew problem of multi-station data caused by conveyor belt vibration, establishing a precisely synchronized heat conduction observation sequence. The consistency of the time axis ensures the accuracy of subsequent dynamic analysis of the thermal field, avoiding misjudgments of the thermal evolution process due to time sequence discrepancies.

[0030] 102. By fusing the long and short wave energy information in the infrared thermal sequence, the image blurring interference caused by residual heat-sealing vapor is eliminated, and a thermal field distribution with spatial coordinates and time dimension is generated. The thermal field distribution is then sampled along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector group. Optionally, step 102 may specifically include the following steps: 1021. Extract a set of key points with location identifiers from the long-wave and short-wave thermal signature quantitative maps in the fused infrared thermal sequence, and construct a spatiotemporal association model that includes the location of key points and the corresponding time scale points; 1022. Cross-compare the energy values ​​of long and short wave key points marked at the same location. When the long wave energy value exceeds the preset ratio threshold of the short wave energy value, it is determined that there is heat sealing steam interference at the location and the corresponding key point is excluded. 1023. Sort the key points that have not been excluded in order of time scale, and form a three-dimensional energy value matrix by associating the energy values ​​of each time point with the location identifier; 1024. Divide the time interval layer based on the beat pulse of the transmission device, and map the three-dimensional energy value matrix into the three-dimensional spatial coordinate grid in each interval layer to form a multi-layer thermodynamic space unit. 1025. The multi-layered thermal spatial units are superimposed along the time axis to form a thermal field distribution with spatial coordinates and time scale.

[0031] 1026. On the heat distribution body, determine the direction of the plastic packaging barrel axis as the main reference axis, and set measurement point lines along the main reference axis at fixed intervals. Each measurement point line contains heat measurement points arranged at equal intervals. 1027. For each thermal measurement point on the measurement point line, calculate the heat change ratio of the point at the same axial position as the previous measurement point line, and continuously record the heat change ratio of the same axial position along the time axis to form a change intensity curve with time scale. 1028. Integrate the variation intensity curves of all axial positions on each measurement point line into the heat flow direction component of the corresponding time series point, and collect the heat flow direction components of all time series points according to the beat pulse time series to form a dynamic heat flow vector group.

[0032] In the above scheme, long and short wave energy information refers to thermal radiation data captured in different bands (long wave 5-14μm / short wave 3-5μm) in infrared thermal images; the thermal field distribution volume is a four-dimensional matrix of thermal values ​​composed of spatial coordinates (X,Y,Z) and time dimension (T); the dynamic heat flow vector group is a time-series dataset of heat conduction direction and intensity calculated point by point along the axis of the plastic bucket.

[0033] In this embodiment of the application, the timestamp-synchronized infrared thermal sequence generated in step 101 is first processed in step 1021. The system automatically identifies the high-temperature core region in each frame of thermal image, extracts key points with unique location identifiers, and records their long-wave and short-wave energy values, for example, in Detect points near the sealing line of plastic packaging barrels at all times The long-wave energy value is 185 W / m², and the short-wave energy value is 120 W / m². An initial dataset containing location ID, timestamp, and dual-band energy is established. Next, step 1022 cross-validates the long-wave and short-wave energy values ​​of key points at the same location. When the long-wave energy of a point exceeds a preset threshold for the short-wave energy (e.g., 1.5 times), it is identified as a steam interference point and excluded. For example, point... exist Points with a long-wavelength intensity of 200 W / m² and a short-wavelength intensity of 110 W / m² were discarded because the ratio 1.82 > 1.5. Next, in step 1023, the remaining keypoints were arranged in chronological order, and their location IDs were mapped to 1 mm³ voxel elements in the 3D coordinate system using spatial mesh mapping. For example, point... (Coordinates x=15, y=32, z=10) in to The energy value at each moment is stored in the timing queue corresponding to grid G15_32_10.

[0034] Subsequently, in step 1024, the time layer is divided according to the pulse interval of the transmission device (e.g., 0.5 seconds), and the three-dimensional energy matrix within each pulse cycle is filled into the spatial grid to form a thermal unit layer, for example, a pulse. to The energy value sequence of grid G20_40_30 within the period is [142W / m², 135W / m², 129W / m²]. Then, in step 1025, all thermodynamic spatial unit layers are superimposed along the time axis to construct a four-dimensional thermal field distribution body with XYZ three-dimensional spatial coordinates superimposed on the T-time dimension. For example, this distribution body stores the complete time axis energy sequence [156, 149, 143, ..., 112] W / m² at coordinates (25, 18, 12).

[0035] Then, in step 1026, measurement lines are set at fixed intervals along the axial direction (Z-axis) of the plastic bucket. Measurement points are evenly distributed along each line according to the spatial grid. For example, measurement line L5 is set at Z=50mm, and 9 measurement points are evenly distributed along the circumference of the bucket. to Then, in step 1027, the rate of change of heat in adjacent time layers is calculated for each thermal measurement point on the measurement point line, for example, point... In pulse to The calorific value decreased from 126 W / m² to 118 W / m² during the period. Calculate the rate of change. And recorded to Time nodes. Finally, step 1028 integrates the rate of change values ​​of all measurement points at the same time node into a heat flow component vector, and arranges them according to the pulse sequence to form a dynamic heat flow vector group, for example, pulse. The corresponding vector group is .

[0036] In practical applications, within the detection system of Packaging Line A, the system processes synchronized thermal sequence data: 320 key points are extracted from the sealing area, of which 48 points are excluded because the long-wave / short-wave ratio exceeds 1.5 times the threshold (e.g., point S7 has a long-wave ratio of 205 W / m² and a short-wave ratio of 128 W / m², with a ratio of 1.60); the remaining key points are mapped to a 0.8 mm³ spatial grid, and a four-dimensional thermal field is constructed with a pulse interval of 0.4 seconds; 12 measurement lines (1.5 mm spacing) are set along the barrel axis, and the rate of change of heat at point R3 during pulse T2→T3 from 153 W / m² to 142 W / m² is calculated. The rate of change of all measurement points at the pulse moment is integrated to form a heat flow vector [-0.072,+0.005,-0.068,...,+0.012].

[0037] This scheme effectively filters out steam interference signals through dual-band energy cross-validation, transforming the complex heat conduction process into an accurate four-dimensional thermal field model; based on dynamic heat flow analysis using axial gradient sampling, it quantifies the microscopic heat transfer laws into a time-trackable vector dataset, providing a physical mechanism-level basis for identifying sealing defects.

[0038] 103. Simultaneously acquire visible light images of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel outline in the visible light image, reconstruct a three-dimensional model of the sealing surface based on the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. Optionally, step 103 may specifically include the following steps: 1031. Mark multiple equally spaced positioning points on the outer edge of the plastic packaging barrel opening in the visible light image, and connect all positioning points to form a closed outer edge connection loop; 1032. Measure the maximum spacing value in the major axis direction and the minimum spacing value in the minor axis direction of the connecting ring of the outer edge line, and calculate the geometric difference between the maximum spacing value in the major axis and the minimum spacing value in the minor axis as a reference value for the degree of bending; 1033. Using the minimum spacing of the minor axis as the reference unit, calculate the ratio of the curvature reference value to the reference unit, and calculate the barrel curvature compensation coefficient, which represents the curvature of the barrel surface, by taking the difference between the ratio value and the standard circular reference value.

[0039] 1034. Input the barrel curvature compensation coefficient, which represents the degree of curvature of the barrel surface, into the preset surface reconstruction model to generate the compensated barrel covering surface structure model. 1035. Divide the covered surface structure model into dense convex and concave unit blocks along the boundary of the sealing area, and calculate the vertical position difference between the center point of each convex and concave unit block and its four adjacent boundary points. 1036. When the vertical position difference is greater than a preset smoothness threshold, the corresponding unit block is determined to be a surface bump. 1037. Statistically determine the boundary positions of the concentrated distribution areas of the surface concave and convex points to form a concave and convex area distribution contour map, and calculate the distance change rate curve between adjacent concave and convex areas in the concave and convex area distribution contour map as the fold distribution shape feature.

[0040] In the above scheme, the elliptical barrel outline refers to the elliptical boundary line formed by the outer edge of the opening of the plastic packaging barrel in the visible light image; the barrel curvature compensation coefficient is a proportional value that quantifies the degree of curvature of the barrel surface; the wrinkle distribution characteristics refer to the distribution law of the position and intensity of the concave and convex changes in the sealing area caused by thermal stress.

[0041] In this embodiment, firstly, step 1031 marks the equally spaced positioning points on the outer edge of the plastic packaging barrel opening in the visible light image. For example, the system automatically identifies 12 evenly distributed points on the barrel opening edge and connects them to form a closed elliptical outer edge connection loop. The coordinate data of all points are recorded as the boundary ring dataset corresponding to the outer edge connection loop. Secondly, step 1032 measures the maximum spacing value in the major axis direction and the minimum spacing value in the minor axis direction of the outer edge connection loop. For example, the distance from point P1 to P7 is measured to be 215 pixels (maximum value in the major axis), and the distance from point P3 to P9 is measured to be 198 pixels (minimum value in the minor axis). The geometric difference 215-198=17 pixels is calculated as a reference value for the degree of curvature. Next, in step 1033, using the minimum minor axis value of 198 pixels as the reference unit, the ratio of the curvature reference value to the reference unit is calculated as 17 / 198≈0.086. This value is compared with the standard circular reference value of 0 (an ideal circle with no difference) to obtain the barrel curvature compensation coefficient 0.086-0=0.086. This coefficient directly reflects the actual ellipticity of the barrel. Subsequently, in step 1034, the curvature compensation coefficient is input into the 3D reconstruction algorithm. For example, a coefficient of 0.086 drives the standard cylindrical model to be stretched by 8.6% along the major axis to generate a covering surface structure model that conforms to the actual barrel shape, solving the measurement error problem caused by barrel deformation. Next, in step 1035, 0.5mm×0.5mm dense grid units are divided on the surface of the sealing area, and the vertical height difference between the center point of each unit and the four corner points is calculated. For example, when the center height of unit M is 2.85mm and the average height of the four corner points is 2.70mm, the height difference is 0.15mm. Then, in step 1036, the height difference of each unit is compared with a preset smoothness threshold of 0.12mm. When the height difference exceeds the threshold, it is determined to be a concave or convex point. For example, the height difference of unit M is 0.15mm > 0.12mm and is marked as a convex point. Finally, in step 1037, the boundaries of the concentrated concave and convex point areas are statistically analyzed. For example, two main concave and convex areas, region R1 (diameter 3.2mm) and region R2 (diameter 2.5mm), are identified. The fluctuation of the distance between the center points of adjacent concave and convex areas is calculated to generate a curve of the wrinkle distribution shape characteristics. The curve quantitatively reflects the distribution law of thermal stress wrinkles in the sealing area.

[0042] In practical applications, on the automated inspection line for plastic packaging drums at Packaging Plant A, after acquiring a visible light image of the sealing area of ​​the plastic packaging drum, the system first identifies 10 equally spaced positioning points on the edge of the drum opening and connects them to form a closed loop. The maximum spacing of the major axis is measured to be 228 pixels and the minimum spacing of the minor axis is 210 pixels. Then, the geometric difference of 18 pixels is calculated, and the curvature ratio of 0.0857 is obtained as the drum curvature compensation coefficient based on the minor axis. This coefficient is input into the 3D reconstruction algorithm to generate a drum covering surface model that adapts to the actual ellipticity. Subsequently, the sealing area is divided into 0.5mm×0.5mm grid units. When detecting the surface height difference, it is found that the center height of unit U37 is 2.98mm and the corner height is 3.15mm, forming a height difference of -0.17mm (exceeding the 0.12mm threshold is marked as a concave point). Finally, two main wrinkled areas (with diameters of approximately 2.8mm and 3.1mm respectively) are identified and the spacing fluctuation feature curve is extracted.

[0043] This solution eliminates projection distortion in visual inspection by accurately quantifying the elliptical curvature of the barrel; and accurately captures the surface deformation distribution caused by thermal stress by extracting wrinkle features based on microscopic height differences, providing a structural deformation data foundation for thermo-mechanical coupling analysis.

[0044] 104. In the curvature compensation coordinate system established according to the barrel curvature compensation coefficient, the dynamic heat flow vector group and the fold distribution characteristics are spatially registered, and the annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the fold fracture zone that is mismatched with the stress direction of the barrel wall are identified by thermal field analysis. Optionally, step 104 may specifically include the following steps: 1041. Establish a three-dimensional curvature compensation coordinate system based on the curvature compensation coefficient of the plastic packaging barrel; 1042. Map the heat flow direction component in the dynamic heat flow vector group to the corresponding position in the curvature compensation coordinate system, and align the concave and convex region distribution contour map in the fold distribution shape feature to the curvature compensation coordinate system; 1043. Select a standard heat diffusion channel as a reference along the major axis of the ellipse, and calculate the average heat diffusion intensity value at each point in the diffusion channel as the baseline. 1044. Identify the region in the curvature compensation coordinate system where the heat diffusion intensity is continuously lower than the baseline threshold and exhibits a closed-loop distribution, and mark it as an annular cooling zone; 1045. Detect fracture-type uneven zones in the outline of the uneven region distribution map that deviate from the main shrinkage direction of the plastic packaging barrel wall by more than a preset angle, and mark them as wrinkle fracture zones.

[0045] In the above scheme, the curvature compensation coordinate system refers to a three-dimensional spatial coordinate system constructed based on the actual elliptical curvature of the barrel, which is used to eliminate measurement errors caused by barrel deformation; the annular cold zone refers to a thermal anomaly area that is continuously at low temperature and distributed in a closed loop; the fold fracture zone refers to a concave-convex zone that deviates significantly from the direction of the principal stress of the barrel wall, reflecting the internal cracking phenomenon of the material.

[0046] In this embodiment, firstly, a three-dimensional curvature compensation coordinate system is constructed based on the barrel curvature compensation coefficient in step 1041. For example, when the compensation coefficient is 0.12, all coordinate values ​​in the X-axis direction of the coordinate system are multiplied by 1.12, so that point [5.0,0,0] becomes [5.6,0.0] and point [10.0,0,0] becomes [11.2,0.0], achieving geometric matching with the elliptical barrel. Secondly, in step 1042, the heat flow component in the dynamic heat flow vector group is mapped to the corresponding position in the curvature compensation coordinate system. For example, the heat flow value of measurement point P at pulse T1 of -0.15 is mapped to the new coordinate system position [10.2,15.3,8.7]. At the same time, the key points of the fold contour (such as the contour boundary point [12.5,16.8,9.0]) in the concave-convex region distribution contour map of the fold distribution shape feature in step 103 are aligned according to the same spatial transformation rules.

[0047] Next, in step 1043, a 2mm wide reference channel is selected in the direction of the major axis of the ellipse, such as the coordinate range Y = -1.0 to +1.0mm. The average heat flux of 8 measurement points in the area is calculated. For example, the 8 values ​​of pulse T1 are: -0.08, -0.12, +0.03, -0.05, -0.10, +0.01, -0.06, and -0.09. After calculation, (total -0.46) / 8 = -0.0575, which is approximately equal to -0.058, and is used as the reference value. Then, in step 1044, the coordinate system grid is scanned. When the heat flux value of a region for three consecutive pulse cycles is lower than 70% of the reference value, i.e., -0.058 × 0.7 = -0.0406, and a closed-loop spatial distribution is formed, if the T1 heat flux value of 8 points in the region [12.8-15.6, 0.0, 7.3-9.1] is detected to be -0.38 (< -0.0406), it is marked as a 3.8mm diameter annular cold zone. Finally, in step 1045, fracture-type uneven zones that deviate from the main shrinkage direction of the plastic packaging barrel wall by more than a preset angle in the uneven zone distribution contour map are detected and marked as wrinkle fracture zones. For example, the wrinkle zone G starts from [8.3, 17.1, 8.5] and ends at [6.7, 22.6, 8.5], with a lateral change. Longitudinal changes Direction angle The angle deviates from the long axis of the barrel by more than 45°, and is marked as a 4.2mm long fold fracture zone.

[0048] In practical applications, in the inspection system of Packaging Plant A, a three-dimensional coordinate system is first constructed based on the barrel curvature compensation coefficient of 0.12. The original standard coordinate system is stretched by 12% along the major axis, for example, coordinate point 5,0,0 becomes 5,6,0,0. Then, the measurement point P in the dynamic heat flux vector group at the heat flux value of -0,15 in pulse T1 is mapped to the positions 10,2,15,3,8,7 in the new coordinate system. At the same time, the boundary points of the wrinkled region R from 12,5,16,8,9,0 to 14,2,18,3,9,0 are aligned to the coordinate system. Then, a 2 mm wide reference channel was selected along the long axis, and the heat flux values ​​at 8 measurement points within this channel during pulse T1 were calculated: these values ​​are -0.08, -0.12, +0.03, -0.05, -0.10, +0.01, -0.06, and -0.09. These values ​​were summed to obtain a total of -0.46, which, divided by 8, yielded a reference value of -0.0575, approximately equal to -0.058. Subsequently, within the coordinate range of 12.8 to 15.6, 0 to 0, and 7.3 to 9.1, a heat flux value of -0.38 was detected during pulse T1. This value is lower than the threshold of -0.058 multiplied by 0.7, which equals -0.0406, and remained below this threshold for three consecutive pulse cycles. Furthermore, these 8 points formed a closed loop, marked as a 3.8 mm diameter annular cold zone. Finally, the direction of the folds was analyzed. The folds G started at 8.3, 17.1, and 8.5 and ended at 6.7, 22.6, and 8.5. The lateral change was calculated as 6.7 minus 8.3, which equals -1.6. The longitudinal change was calculated as 22.6 minus 17.1, which equals 5.5. The direction angle was calculated using the arctangent function to be approximately 106.6 degrees. This angle deviates from the main contraction direction of the barrel wall by more than 45 degrees, and is marked as a fold fracture zone with a length of 4.2 mm.

[0049] This solution achieves seamless integration of thermal conduction data and structural deformation characteristics through a unified spatial benchmark, simultaneously identifying closed-loop cold zones formed by thermal blockage and folded fracture zones caused by stress mismatch, significantly improving the spatial positioning accuracy of defects such as incomplete fusion and stress cracking.

[0050] 105. Based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group, it is determined that the seal has a non-fusion defect or a stress cracking defect.

[0051] Optionally, step 105 may specifically include the following steps: 1051. Calculate the minimum distance between the geometric center point of the annular cold zone and the boundary of the fold fracture zone in the curvature compensation coordinate system; 1052. Statistically analyze the distribution of the deviation of each point of the dynamic heat flow vector within the annular cold zone from the reference heat diffusion value along the major axis of the ellipse; 1053. Extract the angular distribution characteristics of the heat diffusion direction and the principal stress direction of the barrel wall at each point in the area covered by the folded fracture zone; 1054. When the minimum interval distance is lower than the preset proximity relationship threshold and the proportion of high deviation points in the deviation distribution within the annular cold zone exceeds the dominant proportion, and the direction of most points in the included angle distribution characteristics is basically consistent with the direction of the principal stress of the barrel wall, the non-fusion defect is determined to be established. 1055. The stress cracking defect is determined when the minimum interval distance exceeds the preset separation relationship threshold, the low deviation value points occupy the main proportion in the deviation distribution of the annular cold zone, and the points that significantly deviate from the principal stress direction of the barrel wall in the included angle distribution characteristics form a concentrated distribution group.

[0052] In the above scheme, positional correlation refers to the distance relationship between the annular cold zone and the folded fracture zone in the spatial coordinate system; the anisotropic thermal diffusion parameter deviation refers to the degree of deviation between the heat conduction direction and the reference direction; the non-fusion defect is a thermal blockage formed by insufficient bonding of materials; and the stress cracking defect is a structural cracking caused by thermal stress.

[0053] In this embodiment, step 1051 first calculates the shortest spatial distance from the geometric center of the annular cold zone to the boundary of the fold fracture zone. The coordinate data is processed using the Euclidean distance formula. For example, the distance between the center coordinates of the cold zone [20.5, 15.3, 10.0] and the boundary point of the fracture zone [18.7, 16.2, 10.0] is calculated as √[(20.5-18.7)^2 + (15.3-16.2)^2] ≈√[3.24 + 0.81] ≈ 2.01 mm. Next, step 1052 calculates the deviation ratio of the thermal diffusion values ​​of all measurement points from the reference value within the annular cold zone. For example, among the 36 points in the 4 mm diameter cold zone, 28 points have a deviation exceeding 40% of the reference value. The calculated percentage of high deviation points is 28 / 36 × 100 ≈ 77.8%. Next, step 1053 is used to extract the heat diffusion direction angle of each point in the folded fracture zone area and calculate the angle distribution with the principal stress direction of the barrel wall (0 degree direction). For example, among the 50 points in the fracture zone, 35 points have an angle difference greater than 60 degrees, and the proportion of significantly deviated points is 35 / 50×100=70%.

[0054] Then, step 1054 is used to determine the lack of fusion defect: the determination is made when three conditions are met simultaneously: 1. The minimum interval distance is less than the 2 mm threshold (e.g., 1.8 mm in the implementation case); 2. The proportion of high deviation points in the cold zone exceeds 75% (e.g., 77.8% in the implementation case); 3. The majority of points have an angle of less than 30 degrees in the fracture zone (e.g., 66% of the points have an angle of less than 30 degrees in the implementation case); for example, if the distance between a cold zone and the fracture zone is 1.8 mm, which is less than 2 mm, the proportion of high deviation points in the cold zone is 77.8%, which is greater than 75%, and the angle of 66% of the points in the fracture zone is less than 30 degrees, then the system determines that there is a lack of fusion defect.

[0055] Finally, stress cracking defects are determined in step 1055: the determination is valid when three conditions are met simultaneously: 1. The minimum interval distance exceeds the 5 mm threshold (e.g., 6.3 mm in the implementation case); 2. Low deviation points (deviation value <20%) account for more than 70% of the main proportion in the cold zone (e.g., 75% of the points in the implementation case are low deviation points); 3. Significant deviation points of the fracture zone (angle difference >45 degrees) form clusters and account for more than 60% (e.g., 68% of the points in the implementation case have an angle difference >45 degrees and are spatially concentrated). For example, if the cold zone-fracture zone distance of 6.3 mm is greater than 5 mm, 75% of the points in the cold zone are in a low deviation state, and 68% of the points in the fracture zone have an angle difference >45 degrees and are densely distributed, the system determines that stress cracking defects exist.

[0056] In practical applications, in the A-package inspection system, the case for determining non-fusion defects is as follows: Calculate the distance from the center of the annular cold zone (15.2, 18.3, 9.5) to the boundary of the fold fracture zone (13.6, 19.1, 9.5) to 1.79 mm (formula: √[(15.2-13.6)²+(18.3-19.1)²)]); Statistically, 21 out of 25 points in the cold zone have thermal diffusion deviation values ​​exceeding 40% (84%); 27 out of 40 points in the fracture zone have an angle less than 30 degrees with the principal stress direction of the barrel wall (67.5%); The system determines that if the following conditions are met simultaneously: distance 1.79 mm < 2 mm threshold, high deviation points 84% ​​> 75%, and small angle points 67.5% > 60%, then a non-fusion defect is confirmed. Case study of stress cracking defect assessment: The distance from the center of the cold zone (22.7, 14.8, 10.2) to the boundary of the fracture zone (16.3, 20.6, 10.2) is 8.64 mm; 25 out of 32 points in the cold zone have a deviation value of <20% (accounting for 78.1%); 32 out of 45 points in the fracture zone have an angle deviation >45 degrees (of which 28 are spatially clustered); the system determines that the stress cracking defect is confirmed because the distance of 8.64 mm is greater than the 5 mm threshold, the low deviation rate of 78.1% is greater than 70%, and the large angle point cluster is met simultaneously.

[0057] This solution uses a triple combination analysis of spatial distance correlation, thermal diffusion distribution characteristics, and angular deviation patterns to accurately distinguish between incomplete fusion defects caused by insufficient material fusion and structural cracking defects caused by thermal stress, significantly improving the accuracy of defect identification.

[0058] Figure 2 This application provides a scenario diagram illustrating an automatic sealing detection method for plastic packaging drums, as shown in the embodiments below. Figure 2 As shown, a complete embodiment of steps 101-105 includes: On the automatic inspection line of Packaging Plant A for plastic packaging drums, the system first collects the conveyor belt's cycle pulse signal as a time reference. This pulse triggers three infrared cameras at different workstations to simultaneously capture sealing thermal images. Data with deviations in actual shooting time is interpolated for compensation (e.g., if camera 2 is delayed by 8ms, the timestamp is moved forward by 8ms), generating a time-stamped synchronized infrared thermal sequence. Subsequently, the long-wave and short-wave energy information in this sequence is fused. By eliminating interference points where the long-wave / short-wave ratio exceeds 1.5, a four-dimensional thermal field distribution resistant to steam interference is constructed. Measurement lines are set every 1mm along the drum axis, and the heat change rate at each point is calculated to form a dynamic heat flow vector group. Simultaneously, visible light images are acquired, and the major axis (228 pixels) and minor axis (210 pixels) are calculated based on the elliptical contour of the drum opening. The geometric difference of the pixels yielded a curvature compensation coefficient of 0.0857. After reconstructing the 3D barrel surface model, 0.5mm mesh units were divided. Concave and convex points with a height difference exceeding 0.12mm between the center point and corner point of the unit were detected. Finally, two main wrinkled regions with diameters of 2.8mm and 3.1mm were extracted. The heat flow vector was registered with the wrinkle feature space in the curvature compensation coordinate system, and a 3.8mm diameter annular cold zone with a continuous heat flow intensity lower than the baseline value of 40% and a 4.2mm long wrinkle fracture zone with a directional deviation of 106.6° were identified. Finally, by judging the 1.79mm distance between the cold zone and the fracture zone, the 8.64mm stress cracking, and the corresponding thermal diffusion deviation and angular distribution characteristics, the non-fusion defects and stress cracking defects were accurately classified.

[0059] Figure 3 This application provides a schematic diagram of the structure of an automatic sealing and detection system for plastic packaging drums, as shown in the embodiment. Figure 3 As shown, the system includes: The generation module 31 is used to collect infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device, and to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations using the beat pulse signal as a time reference, so as to generate an infrared thermal sequence with timestamp synchronization. The generation module 31 is also used to fuse the long and short wave energy information in the infrared thermal sequence, eliminate the image blurring interference caused by residual heat seal vapor, generate a three-dimensional thermal field distribution with spatial coordinates and time dimension, perform gradient sampling on the three-dimensional thermal field distribution along the axial direction of the plastic packaging barrel, and generate a dynamic heat flow vector group. The calculation module 32 is used to synchronously acquire the visible light image of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel contour in the visible light image, reconstruct the three-dimensional model of the sealing surface according to the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. The registration module 33 is used to spatially register the dynamic heat flow vector group with the wrinkle distribution characteristics in the curvature compensation coordinate system established according to the barrel curvature compensation coefficient, and to identify the annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the wrinkle fracture zone that is mismatched with the stress direction of the barrel wall through thermal field analysis. The judgment module 34 is used to determine whether there is a non-fusion defect or a stress cracking defect in the seal based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group.

[0060] Figure 3 The aforementioned automatic sealing and detection system for plastic packaging drums can perform... Figure 1 The implementation principle and technical effects of the automatic sealing and detection method for plastic packaging drums described in the illustrated embodiment will not be repeated here. The specific operation methods of each module and unit in the automatic sealing and detection system for plastic packaging drums in the above embodiments have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0061] In one possible design, Figure 3 The automatic sealing and detection system for plastic packaging drums shown in the embodiment can be implemented as a computing device, such as... Figure 4 As shown, the computing device may include a storage component 41 and a processing component 42; The storage component 41 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 42.

[0062] The processing component 42 is used for the above Figure 1 The embodiment describes an automatic sealing detection method for plastic packaging barrels.

[0063] The processing component 42 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0064] Storage component 41 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0065] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0066] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0067] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0068] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0069] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 An automatic sealing detection method for plastic packaging barrels is shown in the embodiment.

[0070] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0071] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An automatic sealing detection method for plastic packaging barrels, characterized in that, include: Infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device are collected. The beat pulse signal is used as the time reference to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations to generate a timestamp synchronized infrared thermal sequence. By fusing the long and short wave energy information in the infrared thermal sequence, the image blurring interference caused by residual heat-sealing vapor is eliminated, and a thermal field distribution with spatial coordinates and time dimension is generated. The thermal field distribution is then sampled along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector group. Simultaneously acquire visible light images of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel outline in the visible light image, reconstruct a three-dimensional model of the sealing surface based on the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. In the curvature compensation coordinate system established according to the barrel curvature compensation coefficient, the dynamic heat flow vector group and the fold distribution characteristics are spatially registered, and the annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the fold fracture zone that is mismatched with the stress direction of the barrel wall are identified by thermal field analysis. Based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group, it is determined that the seal has incomplete fusion defects or stress cracking defects.

2. The method according to claim 1, characterized in that, In a curvature compensation coordinate system established based on the barrel curvature compensation coefficient, the dynamic heat flow vector set and the wrinkle distribution characteristics are spatially registered, and annular cold zones deviating from the reference thermal diffusion parameters of the ellipse major axis and wrinkle fracture zones mismatched with the barrel wall stress direction are identified through thermal field analysis, including: A three-dimensional curvature compensation coordinate system is established based on the curvature compensation coefficient of the plastic packaging barrel. The heat flow direction component in the dynamic heat flow vector group is mapped to the corresponding position in the curvature compensation coordinate system, and the concave and convex region distribution contour map in the fold distribution shape feature is aligned to the curvature compensation coordinate system; A standard thermal diffusion channel is selected as a reference along the major axis of the ellipse, and the average thermal diffusion intensity value at each point in the diffusion channel is calculated as the baseline. Identify regions in the curvature compensation coordinate system where the heat diffusion intensity is consistently below the baseline threshold and exhibits a closed-loop distribution, and mark them as annular cooling zones. The fractured uneven area in the outline of the uneven area distribution map that deviates from the main shrinkage direction of the plastic packaging barrel wall by more than a preset angle is marked as a wrinkle fracture area.

3. The method according to claim 1, characterized in that, Based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature-compensated coordinate system and the anisotropic thermal diffusion parameter deviation in the dynamic heat flow vector set, it is determined that the seal has incomplete fusion defects or stress cracking defects, including: Calculate the minimum interval distance from the geometric center point of the annular cold zone to the boundary of the fold fracture zone in the curvature compensation coordinate system; The distribution of the degree of deviation between each point of the dynamic heat flow vector within the annular cold zone and the reference heat diffusion value along the major axis of the ellipse is statistically analyzed. Extract the angular distribution characteristics of the heat diffusion direction and the principal stress direction of the barrel wall at each point in the area covered by the folded fracture zone; When the minimum interval distance is lower than the preset proximity threshold and the proportion of high deviation points in the deviation distribution within the annular cold zone exceeds the dominant proportion, and the direction of most points in the included angle distribution characteristics is basically consistent with the direction of the principal stress of the barrel wall, the non-fusion defect is determined to be established. The stress cracking defect is determined when the minimum interval distance exceeds the preset separation relationship threshold, the low deviation value points occupy the main proportion in the deviation distribution of the annular cold zone, and the points that significantly deviate from the principal stress direction of the barrel wall in the included angle distribution characteristics form a concentrated distribution group.

4. The method according to claim 1, characterized in that, By fusing long-wave and short-wave energy information from the infrared thermal sequence, and eliminating image blurring interference caused by residual vapor from heat sealing, a thermal field distribution with spatial coordinates and time dimensions is generated, including: A set of key points with location identifiers is extracted from the long-wave and short-wave thermal signature quantitative maps in the fused infrared thermal sequence, and a spatiotemporal correlation model containing the location of key points and their corresponding time scale points is constructed. The energy values ​​of long and short wave key points marked at the same location are cross-compared. When the long wave energy value exceeds the preset ratio threshold of the short wave energy value, it is determined that there is heat sealing steam interference at the location and the corresponding key point is excluded. The key points that were not excluded were sorted in order of time scale, and the energy values ​​of each time point were associated with the location identifier to form a three-dimensional energy value matrix; The time interval is divided into layers based on the beat pulse of the transmission device. Within each interval layer, the three-dimensional energy value matrix is ​​mapped to a three-dimensional spatial coordinate grid to form a multi-layer thermodynamic spatial unit. The multiple thermal spatial units are superimposed along the time axis to form a thermal field distribution with spatial coordinates and time scale.

5. The method according to claim 1, characterized in that, Gradient sampling is performed on the heat field distribution body along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector set, including: The axial direction of the plastic packaging barrel is determined as the main reference axis on the heat distribution body, and measurement point lines are set along the main reference axis at fixed intervals. Each measurement point line contains heat measurement points arranged at equal intervals. For each thermal measurement point on the measurement point line, calculate the heat change ratio of the point at the same axial position as the previous measurement point line, and continuously record the heat change ratio of the same axial position along the time axis to form a change intensity curve with time scale. The variation intensity curves of all axial positions on each measurement point line are integrated into the heat flow direction components of the corresponding time series points, and the heat flow direction components of all time series points are collected according to the beat pulse time series to form a dynamic heat flow vector group.

6. The method according to claim 1, characterized in that, Simultaneously acquire visible light images of the sealed area, and calculate the barrel curvature compensation coefficient based on the elliptical barrel contour in the visible light image, including: In a visible light image, mark multiple equally spaced positioning points on the outer edge of the plastic packaging barrel opening, and connect all positioning points to form a closed outer edge connection loop; Measure the maximum spacing value in the major axis direction and the minimum spacing value in the minor axis direction of the connecting rings of the outer edge, and calculate the geometric difference between the maximum spacing value in the major axis and the minimum spacing value in the minor axis as a reference value for the degree of bending; Using the minimum spacing of the minor axis as the reference unit, the ratio of the curvature reference value to the reference unit is calculated, and the difference between the ratio value and the standard circular reference value is used to calculate the barrel curvature compensation coefficient, which represents the curvature of the barrel surface.

7. The method according to claim 1, characterized in that, Based on the barrel curvature compensation coefficient, a three-dimensional model of the sealing surface is reconstructed, and the wrinkle distribution characteristics under thermal stress are extracted, including: Input the barrel curvature compensation coefficient, which represents the degree of curvature of the barrel surface, into the preset surface reconstruction model to generate the compensated barrel covering surface structure model. On the covering surface structure model, dense convex and concave unit blocks are divided along the boundary of the sealing area, and the vertical position difference between the center point of each convex and concave unit block and its four adjacent boundary points is calculated. When the vertical position difference is greater than the preset smoothness threshold, the corresponding unit block is determined to be a surface bump. The boundary positions of the concentrated distribution areas of the surface bumps are statistically analyzed to form a bump distribution profile map. The distance change rate curve between adjacent bumps in the bump distribution profile map is calculated as the shape feature of the fold distribution.

8. An automatic sealing and detection system for plastic packaging drums, characterized in that, include: Infrared thermal imaging data of the sealing area of ​​the plastic packaging barrel and the beat pulse signal generated by the transmission device are collected. The beat pulse signal is used as the time reference to perform time-series calibration on the infrared thermal imaging data of three consecutive workstations to generate a timestamp synchronized infrared thermal sequence. By fusing the long and short wave energy information in the infrared thermal sequence, the image blurring interference caused by residual steam from heat sealing is eliminated, and a three-dimensional thermal field distribution with spatial coordinates and time dimension is generated. The three-dimensional thermal field distribution is then sampled along the axial direction of the plastic packaging barrel to generate a dynamic heat flow vector set. Simultaneously acquire visible light images of the sealing area, calculate the barrel curvature compensation coefficient based on the elliptical barrel outline in the visible light image, reconstruct a three-dimensional model of the sealing surface based on the barrel curvature compensation coefficient, and extract the wrinkle distribution characteristics under thermal stress. In the curvature compensation coordinate system established according to the barrel curvature compensation coefficient, the dynamic heat flow vector group and the fold distribution characteristics are spatially registered. The annular cold zone that deviates from the reference thermal diffusion parameters of the ellipse major axis and the fold fracture zone that is mismatched with the stress direction of the barrel wall are identified through thermal field analysis. Based on the positional correlation between the annular cold zone and the folded fracture zone in the curvature compensation coordinate system and the deviation of the anisotropic thermal diffusion parameters in the dynamic heat flow vector group, it is determined that the seal has incomplete fusion defects or stress cracking defects.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement an automatic sealing detection method for plastic packaging drums as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The device contains a computer program that, when executed by a computer, implements an automatic sealing detection method for plastic packaging drums as described in any one of claims 1 to 7.

Images