Air tightness detection method and device for automobile sunroof sealing strip based on thermal imaging

Abstract

The present application provides a method and device for detecting the air tightness of a car sunroof sealing strip based on thermal imaging, and relates to the field of data processing technology. The method includes: performing all-round continuous detection of the sunroof sealing area in a preset time zone and building a space-time coordinate system, extracting airtightness measurement features and integrating space-time splicing, determining the space-time heat map distribution, establishing an airtightness detection model, performing self-distribution proofreading on the space-time heat map, and identifying the airtightness distribution for airtightness detection effective value identification. This application can solve the technical problem that the accuracy and reliability of airtightness detection in the prior art are poor due to the limitations and inefficiency of the detection method, which further affects the overall performance and user experience of the car sunroof, and realizes the identification of airtightness distribution, identification of airtightness detection effective value, determination of airtightness detection results, and visual display of terminal display devices. It helps users to intuitively understand the airtightness status of the sunroof sealing area.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and device for detecting air tightness of a car sunroof sealing strip based on thermal imaging. Background Art

[0002] Air tightness is a key factor affecting product quality and performance. If the air tightness is poor, it may cause abnormal product operation, reduced efficiency, and even environmental pollution and production safety hazards. Therefore, air tightness testing is regarded as an important technical indicator by many companies.

[0003] At present, the existing methods for testing the air tightness of sunroof sealing strips mostly rely on traditional testing methods, such as pressure change detection, water immersion method or bubble method. Although these methods can detect air tightness problems to a certain extent, they also have some inherent limitations. For example, pressure change detection may not be sensitive enough to slight leaks, water immersion method and bubble method may not be able to accurately locate the leak point, and the operation process is cumbersome and time-consuming.

[0004] In summary, due to the limitations and inefficiency of the detection methods in the prior art, the accuracy and reliability of airtightness detection are poor, which further affects the overall performance of the car sunroof and the technical problems of user experience. Summary of the invention

[0005] The purpose of this application is to provide a method and device for detecting the air tightness of a car sunroof sealing strip based on thermal imaging, so as to solve the technical problem that the accuracy and reliability of air tightness detection are poor due to the limitations and inefficiency of the detection method in the prior art, which further affects the overall performance of the car sunroof and the user experience.

[0006] In view of the above problems, the present application provides a method and device for detecting the air tightness of a car sunroof sealing strip based on thermal imaging.

[0007] In the first aspect, the present application provides a method for detecting the air tightness of a car sunroof sealing strip based on thermal imaging, the method being implemented by an air tightness detection device for a car sunroof sealing strip based on thermal imaging, wherein the method comprises: controlling a thermal infrared imager to perform all-round continuous detection of the sunroof sealing area in a preset time zone, and determining infrared imaging data. Set air tightness measurement features, build a spatiotemporal coordinate system, and supervise the training of a spatiotemporal conversion module, wherein the air tightness measurement feature is a temperature feature. Combined with the spatiotemporal conversion module, traverse the infrared imaging data to extract air tightness measurement features and spatiotemporal splicing and integration, and determine the spatiotemporal heat map distribution, wherein the distribution points are characterized as (x, y, z, t, T). Establish an air tightness detection model, the air tightness detection model includes a self-calibration unit and a mutual calibration unit in a dislocated twin network structure. Combined with the air tightness detection model, the spatiotemporal heat map distribution is self-distributed and calibrated with each other based on a standard airtight structure to determine the air tightness distribution. The airtightness distribution is identified, the effective value of the airtightness detection is marked, the airtightness detection result is determined and visualized on a terminal display device, wherein the effective value of the airtightness detection is a custom level label.

[0008] In the second aspect, the present application also provides an automobile sunroof sealing strip air tightness detection device based on thermal imaging, which is used to perform the automobile sunroof sealing strip air tightness detection method based on thermal imaging as described in the first aspect, wherein the device includes: an imaging data determination unit, the imaging data determination unit is used to control the thermal infrared imager, perform all-round continuous detection of the sunroof sealing area in a preset time zone, and determine the infrared imaging data. A spatiotemporal coordinate system construction unit, the spatiotemporal coordinate system construction unit is used to set the airtightness measurement feature, and build the spatiotemporal coordinate system, supervise the training of the spatiotemporal conversion module, wherein the airtightness measurement feature is a temperature feature. A spatiotemporal heat map distribution determination unit, the spatiotemporal heat map distribution determination unit is used to combine the spatiotemporal conversion module, traverse the infrared imaging data to extract the airtightness measurement feature and spatiotemporal splicing integration, and determine the spatiotemporal heat map distribution, wherein the distribution point is characterized as (x, y, z, t, T). A detection model construction unit, the detection model construction unit is used to establish an airtightness detection model, and the airtightness detection model includes a self-checking unit and a mutual verification unit in a dislocated twin network structure. An airtightness distribution determination unit is used to combine the airtightness detection model to perform self-distribution calibration on the spatiotemporal heat map distribution and mutual calibration based on the standard airtight structure to determine the airtightness distribution. An effective value identification unit is used to identify the airtightness distribution, perform airtightness detection effective value identification, determine the airtightness detection result and visualize it on the terminal display device, wherein the airtightness detection effective value is a custom level label.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] By controlling the thermal infrared imager, the skylight sealing area is subjected to all-round continuous detection in a preset time zone to determine the infrared imaging data. The airtightness measurement feature is set, and a space-time coordinate system is constructed to supervise the training of the space-time conversion module, wherein the airtightness measurement feature is a temperature feature. In combination with the space-time conversion module, the infrared imaging data is traversed to extract airtightness measurement features and integrate space-time splicing to determine the space-time heat map distribution, wherein the distribution points are represented as (x, y, z, t, T). An airtightness detection model is established, and the airtightness detection model includes a self-calibration unit and a mutual verification unit in a dislocated twin network structure. In combination with the airtightness detection model, the space-time heat map distribution is self-distributed and mutually verified based on a standard airtight structure to determine the airtightness distribution. Identify the airtightness distribution, mark the effective value of airtightness detection, determine the airtightness detection result and visualize it on the terminal display device, where the effective value of airtightness detection is a custom level label, which effectively solves the technical problem that the accuracy and reliability of airtightness detection are poor due to the limitations and inefficiency of the detection method in the prior art, further affecting the overall performance and user experience of the automobile sunroof, and realizes the identification of airtightness distribution, the marking of effective value of airtightness detection, the determination of airtightness detection result and the visualization display of the terminal display device. It helps users to intuitively understand the airtightness status of the sunroof sealing area.

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0013] Figure 1 This is a flow chart of a method for detecting air tightness of a car sunroof sealing strip based on thermal imaging in this application;

[0014] Figure 2 This is a schematic diagram of the structure of the automobile sunroof sealing strip air tightness detection device based on thermal imaging in this application.

[0015] Description of reference numerals:

[0016] An imaging data determining unit 11 , a space-time coordinate system building unit 12 , a space-time heat map distribution determining unit 13 , a detection model building unit 14 , an airtightness distribution determining unit 15 , and a valid value identifying unit 16 . DETAILED DESCRIPTION

[0017] This application provides a method and device for detecting air tightness of automobile sunroof sealing strips based on thermal imaging, which solves the technical problem that the accuracy and reliability of air tightness detection are poor due to the limitations and inefficiency of the detection method in the prior art, further affecting the overall performance and user experience of the automobile sunroof, and realizes the identification of air tightness distribution, the identification of air tightness detection effective value, the determination of air tightness detection results, and the visual display of terminal display devices. It helps users to intuitively understand the air tightness status of the sunroof sealing area.

[0018] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0019] Embodiment 1

[0020] Please refer to the attached Figure 1 The present application provides a method for detecting air tightness of a car sunroof sealing strip based on thermal imaging, wherein the method is applied to a device for detecting air tightness of a car sunroof sealing strip based on thermal imaging, and the method specifically comprises the following steps:

[0021] S1: Control the thermal infrared imager to conduct all-round continuous detection of the sunroof sealing area in a preset time zone to determine the infrared imaging data.

[0022] Specifically, select a thermal infrared imager suitable for skylight sealing area inspection, and ensure that it has high temperature resolution and spatial resolution. Install the thermal infrared imager in an appropriate location to ensure that the skylight sealing area can be fully covered. Debug the equipment according to the equipment manual to ensure that the thermal infrared imager works properly. Calibrate the equipment to ensure the accuracy of the measurement data. Set a specific time period as the inspection time zone, such as nighttime or early morning, to reduce external interference. Configure the thermal infrared imager for continuous inspection to ensure that thermal images of the skylight sealing area are continuously captured within the preset time zone. Use the thermal infrared imager to continuously capture infrared images of the skylight sealing area within the preset time zone.

[0023] S2: Set airtightness measurement features, build a space-time coordinate system, and supervise the training of the space-time conversion module, wherein the airtightness measurement features are temperature features.

[0024] Specifically, temperature features are extracted from infrared imaging data. This includes determining the temperature range, calculating the average temperature, identifying temperature anomalies, etc. A spatial coordinate system is established based on the actual size and shape of the skylight sealing area. This coordinate system can accurately locate every point in the sealing area. The infrared imaging data acquired during the continuous detection process are arranged in a time series to form a time dimension. In this way, the temperature data of each spatial point at different time points can be recorded and analyzed. The spatial coordinates and time series are combined to form a four-dimensional space-time coordinate system. In this coordinate system, each point corresponds to a specific spatial position and time point. A large amount of labeled infrared imaging data is collected, where the labels indicate the state of airtightness, such as normal, leakage, etc. These data will be used to train the space-time conversion module. A supervised learning algorithm, such as a deep learning network, is used to train the space-time conversion module. The task of this module is to learn the mapping relationship from the temperature features in the space-time coordinate system to the airtightness state. The performance of the model is evaluated and optimized by methods such as cross-validation, adjusting model parameters, and using regularization techniques. The goal is to improve the accuracy of the model's prediction of the airtightness state. After training, the spatiotemporal conversion module can be used to process new infrared imaging data and predict the airtightness status of the skylight sealing area based on the temperature characteristics.

[0025] S3: In combination with the space-time conversion module, the infrared imaging data is traversed to extract airtightness measurement features and integrate space-time splicing to determine the space-time heat map distribution, wherein the distribution points are represented as (x, y, z, t, T).

[0026] Specifically, ensure that the infrared imaging data of the skylight sealing area has been obtained by the thermal infrared imager. Perform necessary preprocessing on the infrared imaging data, such as denoising and distortion correction, to improve the data quality. Use image processing technology to extract the temperature value of each pixel from the infrared imaging data as the airtightness measurement feature. These temperature values ​​will be used for subsequent airtightness analysis. Arrange the infrared imaging data obtained during the continuous detection process in a time series to ensure that each time point corresponds to an infrared image. For each infrared image, the spatial coordinates (x, y, z) of each pixel point are associated with the corresponding temperature value T and timestamp t according to the spatial size and resolution of the image. Integrate all infrared image data in the time series to form a five-dimensional data set containing spatial coordinates, timestamps and temperature values, that is, each distribution point is represented as (x, y, z, t, T). Use specialized software or tools to visualize the integrated five-dimensional data set to form a spatiotemporal heat map distribution. This heat map can intuitively show the temperature distribution of the skylight sealing area at different time points. By observing the spatiotemporal heat map distribution, areas of abnormal temperature can be identified, which may indicate airtightness issues.

[0027] S4: Establishing an airtightness detection model, wherein the airtightness detection model includes a self-checking unit and a mutual verification unit in a dislocated twin network structure.

[0028] Specifically, a misaligned twin network structure, i.e. two parallel neural networks with different parameters, is used to extract and compare features. A deep learning network, such as a convolutional neural network (CNN), is constructed for feature extraction and preliminary airtightness judgment. The network receives preprocessed infrared imaging data as input. A self-attention mechanism or an adaptive calibration module is introduced inside the network to enable the network to self-correct the deviations in the feature extraction process. The network is self-supervised and adjusted by comparing the outputs of different layers of the network. A labeled dataset is used for supervised learning to train the network to identify the features of airtightness problems. Regularization, data augmentation and other techniques are introduced to prevent overfitting. A second neural network similar to but not identical to the self-correction unit is constructed. This network can be used as a control for the self-correction network to verify and confirm the airtightness test results. The same infrared imaging data is input into the two networks. The outputs of the two networks are compared, and if the results are consistent or highly similar, the confidence of the test results is enhanced. If the results are inconsistent, further inspection mechanisms are triggered, such as manual review or verification using more complex algorithms. Collect and annotate a large amount of infrared imaging data of the sunroof sealing area to form training sets, validation sets, and test sets. Train the networks of the self-verification unit and the mutual verification unit respectively. Use strategies such as cross-validation and learning rate adjustment to optimize the training process. Evaluate the performance of the model on the test set, including indicators such as accuracy, recall, and F1 score.

[0029] S5: In combination with the airtightness detection model, the spatiotemporal heat map distribution is subjected to self-distribution calibration and mutual calibration based on a standard airtight structure to determine the airtightness distribution.

[0030] Specifically, the spatiotemporal heat map distribution data is used as input, which contains the temperature distribution of the skylight sealing area at different time points. The spatiotemporal heat map is extracted using the self-checking unit in the airtightness detection model. This includes identifying temperature anomaly areas, temperature gradient changes, etc. The self-checking unit compares the extracted features with the normal airtightness features pre-learned by the model. The deviation in the feature extraction process is self-corrected through the self-attention mechanism or the adaptive calibration module to ensure the accuracy of the extracted features. After the self-checking is completed, the corrected spatiotemporal heat map distribution is output to highlight the possible airtightness problem areas. Data based on the standard airtight structure is obtained, which can be infrared imaging data collected from similar skylight sealing areas with known good airtightness performance. The spatiotemporal heat map distribution data is feature-matched with the standard airtight structure data. This includes comparing the temperature distribution pattern, the location and number of temperature anomalies, etc. The spatiotemporal heat map data is compared and verified with the standard airtight structure data using the mutual verification unit in the airtightness detection model. If the data of the two are consistent or highly similar, it means that the airtightness of the test area is good; if there is a significant difference, it means that there may be airtightness problems in the test area. After the mutual calibration is completed, the calibration results are output, and the areas that are inconsistent with the standard airtight structure are clearly marked as potential airtightness leakage points or weak areas. The results of the self-distribution calibration and the mutual calibration with the standard airtight structure are fused. This can be achieved through weighted voting, probability fusion and other methods. Based on the fusion results, the airtightness distribution is generated.

[0031] S6: Identify the airtightness distribution, identify the effective value of the airtightness detection, determine the airtightness detection result and visualize it on a terminal display device, wherein the effective value of the airtightness detection is a custom level label.

[0032] Specifically, the spatiotemporal heat map data after self-distribution calibration and mutual calibration with the standard airtight structure is deeply analyzed. Identify temperature abnormal areas, temperature gradient changes, and possible leakage points. According to the data analysis results, the skylight sealing area is divided into different airtightness areas, such as good areas, potential leakage areas, etc. Define a hierarchical labeling system as the effective value of airtightness detection. For example, numbers or letters can be used to represent different airtightness states, such as 1 for good airtightness, 2 for slight leakage, 3 for severe leakage, etc. According to the divided airtightness areas, the corresponding effective value of airtightness detection is assigned to each area. These effective values ​​will serve as the basis for subsequent visualization. The effective values ​​of airtightness detection in each area are summarized to form a comprehensive airtightness detection result. Different effective values ​​of airtightness detection can be represented by color coding, graphic icons, etc. The airtightness detection results are mapped to the visualization interface to ensure that the effective value of airtightness detection in each area can be accurately displayed.

[0033] Furthermore, step S2 of the present application also includes:

[0034] The time axis, space axis and feature axis are determined, and the space-time coordinate system is jointly constructed, wherein the space axis includes a three-dimensional axis.

[0035] Based on the spatiotemporal coordinate system, supervised training of the transformation layer is performed.

[0036] The front spatiotemporal splicing layer and the feature extraction layer are combined with the conversion layer to generate the spatiotemporal conversion module, wherein the spatiotemporal splicing layer uses the timestamp as the time splicing standard and the position distribution of the skylight sealing area as the space splicing standard.

[0037] Specifically, the time axis is used to represent the state changes of the skylight sealing area at different time points. The timestamp can be used as a mark on the time axis to record the state of the skylight sealing area corresponding to each time point. The spatial axis includes a three-dimensional axis, which is used to represent the position distribution of the skylight sealing area in three-dimensional space. These three axes usually correspond to the X, Y, and Z axes in the actual space, and can accurately locate the various parts of the skylight sealing area. The feature axis is used to represent different characteristic attributes of the skylight sealing area, such as temperature, pressure, deformation, etc. The time axis, space axis, and feature axis are combined to form a multi-dimensional space-time coordinate system. In this coordinate system, each point represents a certain characteristic value of the skylight sealing area at a specific time and a specific spatial position. Based on the established space-time coordinate system, a conversion layer is trained through supervised learning. The function of this conversion layer is to convert the original space-time data, such as infrared imaging data, deformation data, etc., into a higher-level feature representation. During the training process, a labeled data set is used to guide the conversion layer to learn how to extract useful features. The timestamp is used as the time splicing standard to ensure that the data at different time points can be correctly aligned. The location distribution of the skylight sealing area is used as the spatial stitching standard to ensure the accuracy of the spatial location information. Through the spatiotemporal stitching layer, the data of different time and space are integrated into a unified framework. The feature extraction layer extracts key features from the spatiotemporal data, which will be used for subsequent airtightness detection tasks. Deep learning techniques such as convolutional neural networks can be used to implement the functions of the feature extraction layer. The front spatiotemporal stitching layer, feature extraction layer and conversion layer are combined to form a complete spatiotemporal conversion module. This module can receive the original spatiotemporal data as input and output high-level feature representation after conversion and feature extraction.

[0038] Furthermore, step S3 of the present application also includes:

[0039] The infrared imaging data is traversed, and data clustering is performed based on acquisition timestamps to determine a plurality of data groups.

[0040] The plurality of data groups are traversed, and based on the position distribution of the skylight sealing area, each data group is spatially spliced ​​to determine the spatiotemporal infrared data, wherein the overlapped splicing positions are averaged.

[0041] Extract airtightness measurement features based on the spatiotemporal infrared data, perform conversion based on the spatiotemporal coordinate system, and determine the spatiotemporal heat map distribution.

[0042] Specifically, all infrared imaging data are obtained, which include infrared images of the skylight sealing area collected at different time points. Ensure that each piece of data has corresponding timestamp information. Traverse all infrared imaging data, cluster the data according to the acquisition timestamp, and classify the data with similar timestamps into the same data group. Clustering algorithms such as K-means, DBSCAN, etc. or simple time window partitioning methods can be used to achieve this step. For each data group formed, traversal processing begins. In each data group, spatially splice each infrared image based on the position distribution of the skylight sealing area. Ensure that the actual physical position of the skylight sealing area is taken into account during splicing so as to accurately align the same parts in different images. For the overlapping areas that appear during the splicing process, the mean strategy is adopted to obtain a smoother transition effect. This ensures that the data in the overlapping part will not generate outliers due to repeated calculations. Extract metric features related to airtightness from the spliced ​​spatiotemporal infrared data, such as temperature distribution, temperature gradient, abnormal hot spots, etc. These features will be used for subsequent airtightness analysis and evaluation. Using the constructed spatiotemporal coordinate system, the extracted airtightness measurement features are converted into this coordinate system. Through this conversion, the airtightness status of the skylight sealing area at different times and spaces can be more accurately located and analyzed. Based on the converted data and features, a spatiotemporal heat map distribution is generated in the spatiotemporal coordinate system. This heat map will intuitively show the changes in the airtightness status of the skylight sealing area at different time and space points.

[0043] Furthermore, step S4 of the present application also includes:

[0044] The self-checking unit includes a check analysis area and a data storage area, and the mutual verification unit includes a standardization area and an integrated output area.

[0045] The self-checking unit and the mutual checking unit are distributed in parallel, and a lateral interaction channel exists between the data storage area and the standardization area.

[0046] Specifically, the proofreading and analysis area is responsible for receiving the spatiotemporal heat map distribution data and performing self-distribution proofreading. The proofreading and analysis area will extract features from the data and compare them with the normal airtightness features pre-learned by the model to identify anomalies or deviations. Through the self-attention mechanism or adaptive calibration technology, the proofreading and analysis area can self-correct the deviations in the feature extraction process to ensure the accuracy of the data. The data storage area is used to store the data processed by the proofreading and analysis area, including the original data and the proofread data. This area is also responsible for lateral interaction with the standardization area of ​​the mutual verification unit to share necessary data and information for more comprehensive verification and analysis. The mutual verification unit standardization area receives data based on the standard airtight structure, which is used as a reference standard for mutual verification. This area will compare the received data with the data provided by the data storage area in the self-proofreading unit to verify the accuracy of the spatiotemporal heat map distribution. The standardization area is also responsible for standardizing the data for more accurate comparison and analysis. The integrated output area is responsible for integrating the data and comparison results of the standardization area. This area will generate a comprehensive verification report or output results, clearly indicating the areas that are inconsistent with the standard airtight structure and potential airtightness problems. There is a lateral interaction channel between the data storage area and the standardization area, allowing data and information to be shared between the two areas. Through this channel, the self-calibration unit and the mutual verification unit can work together to achieve more comprehensive and accurate airtightness testing and analysis.

[0047] Furthermore, the present application also includes:

[0048] Based on the same timestamp, a one-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the first calibration information.

[0049] Based on the time series transition, a two-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the second calibration information.

[0050] Based on the first proofreading information and the second proofreading information, a self-proofreading result is determined and stored in the data storage area.

[0051] Specifically, extract the temperature data of the skylight sealing area with the same timestamp from the spatiotemporal heat map data. Make sure that these data points are collected at the same time point. For each timestamp, calculate the temperature value of each point in the skylight sealing area. Extract temperature features, such as average temperature, maximum temperature, minimum temperature, temperature standard deviation, etc. Analyze the distribution and fluctuation range of the skylight sealing area temperature at the same timestamp. Identify temperature abnormal points or abnormal areas, which may be preliminary indications of airtightness problems. Record the temperature feature analysis results at each timestamp. The identified abnormal points and areas are used as part of the first proofreading information. Extract the skylight sealing area temperature data at a series of consecutive timestamps. These data will be used to analyze the temperature over time. For each time point, calculate the temperature features of the skylight sealing area, such as the trend of the average temperature change, the change of the temperature gradient, etc. Analyze the temperature change trend of the skylight sealing area over time. Identify patterns such as rapid temperature changes, periodic fluctuations, or abnormal mutations. Record the temperature feature analysis results under time series. The identified abnormal time series changes are used as part of the second proofreading information. Determination and storage of self-proofreading results. The first proofreading information and the second proofreading information are combined. The abnormal points and abnormal patterns in the two proofreading information are compared and analyzed to find the consistency and difference. Based on the comprehensive proofreading information, the airtightness status of the skylight sealing area is determined. Potential airtightness problem areas and abnormal changes in timing are identified. The self-proofreading results are stored in the data storage area. The self-proofreading results may include the location of the abnormal points, a chart of the timing changes, an evaluation of the airtightness status, etc.

[0052] Furthermore, the present application also includes:

[0053] In combination with the lateral interactive channel, the data storage area and the standardization area are interactively checked to determine the mutual check result, wherein the standardization area has a built-in standard airtight structure and is marked with airtight relaxation.

[0054] In conjunction with the integrated output region, the self-calibration result and the mutual calibration result are integrated to determine the air tightness distribution.

[0055] The airtight relaxation is set based on the allowable extreme value range of the standard airtight structure.

[0056] Specifically, the data storage area sends the self-calibration results to the standardization area through the lateral interaction channel. After receiving the self-calibration results, the standardization area compares them with the built-in standard airtight structure. The standardization area analyzes the abnormal points and areas in the self-calibration results based on the characteristics and parameters of the standard airtight structure. Special attention is paid to the areas that are significantly different from the standard airtight structure, which may indicate airtightness problems. Based on the comparison analysis, the standardization area generates mutual calibration results, including the degree of conformity with the standard airtight structure, the identification of abnormal areas, etc. The mutual calibration results also include the airtightness status evaluated according to the airtight relaxation of the standard airtight structure. The integrated output area receives the self-calibration results and the mutual calibration results. The self-calibration results and the mutual calibration results are integrated to compare the similarities and differences between the two. Special attention is paid to the areas marked as abnormal by both, which are likely to be the key to the airtightness problem. Based on the integrated results, the airtightness distribution of the skylight sealing area is determined. The areas with good airtightness and the areas with problems are identified. The relaxation of the areas with problems is evaluated with reference to the airtight relaxation of the standard airtight structure built-in in the standardization area. According to the allowable extreme value range, it is judged whether the airtight relaxation of these problem areas is within the acceptable range. The integrated output area outputs the final airtightness distribution results and airtight relaxation evaluation.

[0057] Furthermore, the present application also includes:

[0058] The airtightness level is set with the standard airtight structure as the boundary value, wherein the airtightness level includes a positive airtightness level and a negative airtightness level.

[0059] Based on the airtightness level, the airtightness distribution is labeled to determine the level identification distribution.

[0060] Based on the airtight relaxation, an off-limit mark is performed on the airtightness distribution to determine an off-limit mark distribution, wherein the off-limit mark is 1 or 0.

[0061] The air tightness detection result is determined by combining the grade identification distribution with the limit-crossing identification distribution.

[0062] Specifically, the airtightness level is set with the standard airtight structure as the boundary value. This boundary value is determined based on the performance and parameters of the standard airtight structure. The airtightness level is divided into positive airtightness level and negative airtightness level. A positive airtightness level indicates that the airtightness is good or better than the standard, while a negative airtightness level indicates that the airtightness is poor or lower than the standard. Based on the airtightness level, the airtightness distribution is labeled and each area is assigned a corresponding airtightness level mark. Through the labeling, you can clearly see the airtightness level distribution of the entire skylight sealing area, and understand which areas have good airtightness and which areas need improvement. Based on the airtightness relaxation, the airtightness distribution is judged to be out of limit. If the airtightness relaxation of a certain area exceeds the extreme value range allowed by the standard airtight structure, it is considered that the area is out of limit. The out-of-limit area is marked, and the over-limit mark is 1, indicating that the airtightness of the area does not meet the standard; the area that does not exceed the limit is marked as 0, indicating that the airtightness is within the acceptable range. Through the over-limit mark, you can intuitively see which areas have airtightness exceeding the standard range. Combine the distribution of grade marks with the distribution of limit violations to form a comprehensive airtightness test result. The comprehensive distribution of marks can provide a comprehensive understanding of the airtightness of the skylight sealing area. The area with a positive airtightness grade and no limit violations is the best airtight part; the area with a negative airtightness grade and a limit violation is the area that needs to be focused on for improvement.

[0063] In summary, the automotive sunroof sealing strip air tightness detection method based on thermal imaging provided by the present application has the following technical effects:

[0064] By controlling the thermal infrared imager, the skylight sealing area is subjected to all-round continuous detection in a preset time zone to determine the infrared imaging data. The airtightness measurement feature is set, and a space-time coordinate system is constructed to supervise the training of the space-time conversion module, wherein the airtightness measurement feature is a temperature feature. In combination with the space-time conversion module, the infrared imaging data is traversed to extract airtightness measurement features and integrate space-time splicing to determine the space-time heat map distribution, wherein the distribution points are represented as (x, y, z, t, T). An airtightness detection model is established, and the airtightness detection model includes a self-calibration unit and a mutual verification unit in a dislocated twin network structure. In combination with the airtightness detection model, the space-time heat map distribution is self-distributed and mutually verified based on a standard airtight structure to determine the airtightness distribution. Identify the airtightness distribution, mark the effective value of airtightness detection, determine the airtightness detection result and visualize it on the terminal display device, where the effective value of airtightness detection is a custom level label, which effectively solves the technical problem that the accuracy and reliability of airtightness detection are poor due to the limitations and inefficiency of the detection method in the prior art, further affecting the overall performance and user experience of the automobile sunroof, and realizes the identification of airtightness distribution, the marking of effective value of airtightness detection, the determination of airtightness detection result and the visualization display of the terminal display device. It helps users to intuitively understand the airtightness status of the sunroof sealing area.

[0065] Embodiment 2

[0066] Based on the same inventive concept as the automobile sunroof sealing strip air tightness detection method based on thermal imaging in the aforementioned embodiment, the present application also provides an automobile sunroof sealing strip air tightness detection device based on thermal imaging, please refer to the attached Figure 2 , the device comprises:

[0067] The imaging data determining unit 11 is used to control the thermal infrared imager to perform all-round continuous detection of the skylight sealing area in a preset time zone to determine infrared imaging data.

[0068] The space-time coordinate system building unit 12 is used to set airtightness measurement features, build a space-time coordinate system, and supervise the training of the space-time conversion module, wherein the airtightness measurement features are temperature features.

[0069] The spatiotemporal heat map distribution determination unit 13 is used to combine the spatiotemporal conversion module, traverse the infrared imaging data to extract airtightness measurement features and spatiotemporal splicing integration, and determine the spatiotemporal heat map distribution, wherein the distribution points are represented as (x, y, z, t, T).

[0070] The detection model building unit 14 is used to establish an airtightness detection model, and the airtightness detection model includes a self-checking unit and a mutual checking unit in a dislocated twin network structure.

[0071] The airtightness distribution determination unit 15 is used to combine the airtightness detection model, perform self-distribution calibration on the spatiotemporal heat map distribution, and perform mutual calibration based on a standard airtight structure to determine the airtightness distribution.

[0072] The effective value identification unit 16 is used to identify the airtightness distribution, perform airtightness detection effective value identification, determine the airtightness detection result and visualize it on the terminal display device, wherein the airtightness detection effective value is a custom level label.

[0073] Furthermore, the space-time coordinate system building unit 12 in the device is also used for:

[0074] The time axis, space axis and feature axis are determined, and the space-time coordinate system is jointly constructed, wherein the space axis includes a three-dimensional axis.

[0075] Based on the spatiotemporal coordinate system, supervised training of the transformation layer is performed.

[0076] The front spatiotemporal splicing layer and the feature extraction layer are combined with the conversion layer to generate the spatiotemporal conversion module, wherein the spatiotemporal splicing layer uses the timestamp as the time splicing standard and the position distribution of the skylight sealing area as the space splicing standard.

[0077] Furthermore, the spatiotemporal heat map distribution determination unit 13 in the device is also used for:

[0078] The infrared imaging data is traversed, and data clustering is performed based on acquisition timestamps to determine a plurality of data groups.

[0079] The plurality of data groups are traversed, and based on the position distribution of the skylight sealing area, each data group is spatially spliced ​​to determine the spatiotemporal infrared data, wherein the overlapped splicing positions are averaged.

[0080] Extract airtightness measurement features based on the spatiotemporal infrared data, perform conversion based on the spatiotemporal coordinate system, and determine the spatiotemporal heat map distribution.

[0081] Furthermore, the detection model building unit 14 in the device is also used for:

[0082] The self-checking unit includes a check analysis area and a data storage area, and the mutual verification unit includes a standardization area and an integrated output area.

[0083] The self-checking unit and the mutual checking unit are distributed in parallel, and a lateral interaction channel exists between the data storage area and the standardization area.

[0084] Furthermore, the device further comprises a two-dimensional fluctuation analysis unit, and the two-dimensional fluctuation analysis unit is used for:

[0085] Based on the same timestamp, a one-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the first calibration information.

[0086] Based on the time series transition, a two-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the second calibration information.

[0087] Based on the first proofreading information and the second proofreading information, a self-proofreading result is determined and stored in the data storage area.

[0088] Furthermore, the device further comprises a result integration unit, and the result integration unit is used for:

[0089] In combination with the lateral interactive channel, the data storage area and the standardization area are interactively checked to determine the mutual check result, wherein the standardization area has a built-in standard airtight structure and is marked with airtight relaxation.

[0090] In conjunction with the integrated output region, the self-calibration result and the mutual calibration result are integrated to determine the air tightness distribution.

[0091] The airtight relaxation is set based on the allowable extreme value range of the standard airtight structure.

[0092] Furthermore, the device further comprises a level identification distribution determining unit, wherein the level identification distribution determining unit is configured to:

[0093] The airtightness level is set with the standard airtight structure as the boundary value, wherein the airtightness level includes a positive airtightness level and a negative airtightness level.

[0094] Based on the airtightness level, the airtightness distribution is labeled to determine the level identification distribution.

[0095] Based on the airtight relaxation, an off-limit mark is performed on the airtightness distribution to determine an off-limit mark distribution, wherein the off-limit mark is 1 or 0.

[0096] The air tightness detection result is determined by combining the grade identification distribution with the limit-crossing identification distribution.

[0097] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The automobile sunroof sealing strip air tightness detection method based on thermal imaging and the specific examples in the first embodiment are also applicable to the automobile sunroof sealing strip air tightness detection device based on thermal imaging in this embodiment. Through the above detailed description of the automobile sunroof sealing strip air tightness detection method based on thermal imaging, those skilled in the art can clearly know the automobile sunroof sealing strip air tightness detection device based on thermal imaging in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0098] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0099] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for detecting air tightness of automobile sunroof sealing strips based on thermal imaging, characterized in that: The method comprises: Control the thermal infrared imager to conduct all-round continuous detection of the skylight sealing area in a preset time zone to determine the infrared imaging data; Setting airtightness measurement features, building a spatiotemporal coordinate system, and supervising the training of a spatiotemporal conversion module, wherein the airtightness measurement features are temperature features; Combined with the spatiotemporal conversion module, the infrared imaging data is traversed to extract airtightness measurement features and spatiotemporal splicing and integration to determine the spatiotemporal heat map distribution, wherein the distribution points are represented as (x, y, z, t, T); Establishing an airtightness detection model, wherein the airtightness detection model includes a self-checking unit and a mutual checking unit in a dislocated twin network structure; In combination with the airtightness detection model, the spatiotemporal heat map distribution is self-calibrated and mutually calibrated based on a standard airtight structure to determine the airtightness distribution; The airtightness distribution is identified, the effective value of the airtightness detection is marked, the airtightness detection result is determined and visualized on a terminal display device, wherein the effective value of the airtightness detection is a custom level label.

2. The method according to claim 1, characterized in that The supervised training spatiotemporal conversion module comprises: Determine the time axis, the space axis and the characteristic axis, and jointly build the space-time coordinate system, wherein the space axis includes a three-dimensional axis; Based on the spatiotemporal coordinate system, supervised training of the transformation layer; The front spatiotemporal splicing layer and the feature extraction layer are combined with the conversion layer to generate the spatiotemporal conversion module, wherein the spatiotemporal splicing layer uses the timestamp as the time splicing standard and the position distribution of the skylight sealing area as the space splicing standard.

3. The method according to claim 1, characterized in that Traversing the infrared imaging data to extract airtightness measurement features and spatiotemporal splicing and integration, including: Traversing the infrared imaging data, performing data clustering based on acquisition timestamps, and determining a plurality of data groups; Traversing the plurality of data groups, based on the position distribution of the skylight sealing area, spatially splicing each data group to determine the spatiotemporal infrared data, wherein the overlapped splicing positions are averaged; Extract airtightness measurement features based on the spatiotemporal infrared data, perform conversion based on the spatiotemporal coordinate system, and determine the spatiotemporal heat map distribution.

4. The method according to claim 1, characterized in that The dislocation twin network structure includes: The self-checking unit includes a check and analysis area and a data storage area, and the mutual verification unit includes a standardization area and an integrated output area; The self-checking unit and the mutual checking unit are distributed in parallel, and a lateral interaction channel exists between the data storage area and the standardization area.

5. The method according to claim 4, characterized in that The spatiotemporal heat map distribution is self-distributed proofread, including: Based on the same timestamp, a one-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the first calibration information; Based on the time series, a two-dimensional fluctuation analysis of the temperature characteristics of the skylight sealing area is performed to determine the second calibration information; Based on the first proofreading information and the second proofreading information, a self-proofreading result is determined and stored in the data storage area.

6. The method according to claim 5, characterized in that Conduct mutual calibration based on standard airtight structures, including: In combination with the lateral interactive channel, the data storage area and the standardization area are interactively checked to determine the mutual check result, wherein the standardization area has a built-in standard airtight structure and is marked with airtight relaxation; Integrate the self-calibration result and the mutual-calibration result in combination with the integrated output area to determine the air tightness distribution; The airtight relaxation is set based on the allowable extreme value range of the standard airtight structure.

7. The method according to claim 6, characterized in that The effective value of the airtightness detection is a custom level label, including: Taking the standard airtight structure as the boundary value, the airtightness level is set, wherein the airtightness level includes a positive airtightness level and a negative airtightness level; Based on the airtightness level, the airtightness distribution is labeled to determine the level identification distribution; Based on the airtight relaxation, an off-limit mark is performed on the airtight distribution to determine an off-limit mark distribution, wherein the off-limit mark is 1 or 0; The air tightness detection result is determined by combining the grade identification distribution with the limit-crossing identification distribution.

8. The air tightness detection device of automobile sunroof sealing strip based on thermal imaging is characterized in that: For implementing the steps of the method according to any one of claims 1 to 7, the device comprises: An imaging data determination unit, the imaging data determination unit is used to control the thermal infrared imager to perform all-round continuous detection of the skylight sealing area in a preset time zone to determine infrared imaging data; A space-time coordinate system building unit, the space-time coordinate system building unit is used to set airtightness measurement features, build a space-time coordinate system, and supervise the training of a space-time conversion module, wherein the airtightness measurement feature is a temperature feature; A spatiotemporal heat map distribution determination unit, which is used to combine the spatiotemporal conversion module, traverse the infrared imaging data to extract airtightness measurement features and spatiotemporal splicing integration, and determine the spatiotemporal heat map distribution, wherein the distribution points are represented as (x, y, z, t, T); A detection model building unit, the detection model building unit is used to establish an airtightness detection model, the airtightness detection model includes a self-checking unit and a mutual checking unit in a dislocated twin network structure; An airtightness distribution determination unit, the airtightness distribution determination unit is used to perform self-distribution calibration on the spatiotemporal heat map distribution in combination with the airtightness detection model, and mutual calibration based on a standard airtight structure to determine the airtightness distribution; The effective value identification unit is used to identify the airtightness distribution, perform effective value identification of airtightness detection, determine the airtightness detection result and visualize it on the terminal display device, wherein: The valid value of airtight detection is the custom level label.

Images