An AI-based seal integrity detection method and system

Abstract

An embodiment of the present invention relates to the technical field of seal detection, and discloses an airtightness detection method based on artificial intelligence, including: determining corresponding control parameter information according to the volume of the object under test and the deformation parameter of the object under test in the information of the object under test; controlling the working states of the gas source assembly and the inflation ratio valve according to the control parameter information to perform an inflation operation on the object under test; if the pressure change information is less than the first set threshold, it is determined that the airtightness of the corresponding object under test meets the requirements, and if the pressure change information is greater than the first set threshold, the next step is executed; obtaining an infrared detection image of the object under test captured by an infrared camera, and inputting the infrared detection image into an airtightness detection model for recognition to determine the airtightness state of the corresponding object under test, and outputting the corresponding airtightness state. The solution of the embodiment of the present invention can improve the detection efficiency while ensuring the detection accuracy through precise inflation control and an intelligent detection process.

Description

Technical Field

[0001] The present invention relates to the technical field of sealing detection, and particularly relates to a sealing detection method and system based on artificial intelligence. Background Art

[0002] Currently, the following problems exist when existing equipment conducts sealing detection: First, accurate volume estimation of the object to be measured cannot be performed, so rapid and accurate sealing detection cannot be achieved; Second, accurate leakage measurement cannot be achieved during measurement. Therefore, designing a solution that can efficiently perform sealing tests has become a technical problem that needs to be urgently solved by those skilled in the art. Summary of the Invention

[0003] In view of the above defects, an embodiment of the present invention discloses a sealing detection method based on artificial intelligence, which can achieve rapid and accurate sealing detection.

[0004] A first aspect of an embodiment of the present invention discloses a sealing detection method based on artificial intelligence, including:

[0005] Obtain information of the object to be measured, and determine corresponding control parameter information according to the volume of the object to be measured and the deformation parameter of the object to be measured in the information of the object to be measured;

[0006] Control the working states of the gas source assembly and the inflation ratio valve according to the control parameter information to inflate the object to be measured according to a set inflation standard. Among them, the operation of inflating the object to be measured includes inflating the cavity where the object to be measured is placed or inflating the cavity of the object to be measured;

[0007] After the inflation operation is completed, obtain the pressure change information of the object to be measured within a period of time through a pressure sensor. If the pressure change information is less than the first set threshold, determine that the airtightness of the corresponding object to be measured meets the requirements and determine it as a qualified state. If the pressure change information is greater than the first set threshold, perform the next step;

[0008] Obtain the infrared detection image of the object to be measured captured by an infrared camera, and input the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and output the corresponding airtightness state.

[0009] As an optional implementation manner, in the first aspect of an embodiment of the present invention, the step of obtaining the infrared detection image of the object to be measured captured by an infrared camera, inputting the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and outputting the corresponding airtightness state includes:

[0010] Obtain the set of detection image information of the object under test captured by the infrared camera within a period of time, and input each detection image information in the set of detection image information into the temperature detection model for detection to obtain the temperature state change at each time point;

[0011] Input the temperature state change into the airtightness timing detection model for identification to determine whether the object under test meets the airtightness requirements. If so, determine that the airtightness of the corresponding object under test meets the requirements and determine it as a qualified state. If not, determine that the corresponding object under test does not meet the airtightness requirements and determine it as an unqualified state, and output the airtightness detection result of the corresponding leakage point. The airtightness detection result includes the leakage point position information and the leakage point size information.

[0012] As an alternative implementation manner, in the first aspect of the embodiments of the present invention, before obtaining the set of detection image information of the object under test captured by the infrared camera within a period of time, it further includes:

[0013] Obtain the image information of the object under test captured by the camera, and input the image information of the object under test into the detection and recognition model to determine the position information of each detection point in the image information of the object under test; or, transmit the image information of the object under test captured by the camera to the user terminal, so that the user circles the corresponding detection area at the user terminal and transmits the corresponding detection area to the detection terminal;

[0014] The step of inputting each detection image information in the set of detection image information into the airtightness detection model for detection to obtain the leakage state change at each time point includes:

[0015] Determine the corresponding infrared detection area in each detection image information in the set of detection image information according to the position information of each detection point or the detection area;

[0016] Input the infrared detection area into the airtightness detection model for detection to determine the corresponding temperature change information.

[0017] As an alternative implementation manner, in the first aspect of the embodiments of the present invention, the step of obtaining the infrared detection image of the object under test captured by the infrared camera and inputting the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object under test further includes:

[0018] Obtain the infrared detection image of the object under test captured by the infrared camera, and process the infrared detection image to obtain the preprocessed infrared detection data;

[0019] Determine the temperature parameter of each pixel point in the infrared detection data, and determine the gray difference parameter and the gradient difference parameter of adjacent pixel points according to the temperature parameter of each pixel point;

[0020] Determine the gray-scale difference parameter and gradient difference parameter between the corresponding data point and adjacent pixel points according to the temperature parameters of each pixel point to determine the corresponding change node information;

[0021] Determine the change parameter information between each pixel point and other pixel points according to the change node information of each pixel point in the obtained infrared detection data, and determine the association information between each pixel point and other pixel point regions according to the transformation parameter information;

[0022] Perform clustering operation on the infrared detection data according to the association information between each pixel point and other pixel point regions to obtain the clustered image result, perform feature extraction on the image result to obtain the corresponding recognition features, and input the recognition features into the airtightness detection model for recognition to determine the airtightness state of the corresponding object under test.

[0023] As an optional implementation manner, in the first aspect of the embodiments of the present invention, the obtaining the information of the object under test and determining the corresponding control parameter information according to the volume of the object under test and the deformation parameter of the object under test includes:

[0024] Obtain the point cloud basic data of the object under test through a three-dimensional scanner, identify the voxels occupied by the object under test through image processing technology, and calculate the number of voxels;

[0025] Judge whether the object under test is a known detection object according to the point cloud basic data. If so, directly retrieve the volume parameter and deformation state parameter associated with the corresponding object under test. If not, identify the number of voxels occupied by the object under test through image processing technology, determine its volume parameter according to the number of voxels, and determine the corresponding deformation state parameter according to the volume parameter and the point cloud basic data. The deformation state parameter refers to the influence of the air pressure parameter on the deformation of the object under test;

[0026] Determine the corresponding control parameter information according to the volume parameter and the deformation state parameter.

[0027] As an optional implementation manner, in the first aspect of the embodiments of the present invention, the airtightness detection method further includes:

[0028] During the inflation process, obtain the point cloud dynamic data collected by the three-dimensional scanner, and construct the voxel association relationship between the point cloud basic data and the point cloud dynamic data;

[0029] Calculate the similarity parameter between the point cloud basic data and the point cloud dynamic data according to the voxel association, and input the similarity parameter into the pre-constructed deformation detection model to determine the corresponding deformation result and output the corresponding deformation result.

[0030] As an alternative embodiment, in the first aspect of the embodiments of the present invention, controlling the working states of the gas source assembly and the inflation ratio valve according to the control parameter information to inflate the object to be measured according to the set inflation standard includes:

[0031] Determining the target value of the inflation pressure according to the control parameter information, and measuring the actual output pressure of the inflation system in real time through a pressure sensor;

[0032] Comparing the target value of the inflation pressure with the actual output pressure detected by the pressure sensor to calculate the corresponding error parameter;

[0033] Calculating the corresponding control quantity according to the error parameter and the parameter set of the PID controller, the parameter combination including the proportional coefficient, the integral coefficient and the differential coefficient;

[0034] Adjusting the inflation parameters of the inflation system with the calculated control quantity so that the actual measured value gradually approaches the target value and inflating the object to be measured.

[0035] The second aspect of the embodiments of the present invention discloses an airtightness detection system based on artificial intelligence, including:

[0036] An acquisition module: used to acquire information of the object to be measured, and determine the corresponding control parameter information according to the volume of the object to be measured and the deformation parameter of the object to be measured in the information of the object to be measured;

[0037] An inflation module: used to control the working states of the gas source assembly and the inflation ratio valve according to the control parameter information to inflate the object to be measured according to the set inflation standard, wherein, the operation of inflating the object to be measured includes inflating the cavity where the object to be measured is placed or inflating the cavity of the object to be measured;

[0038] A pressure detection module: used to acquire the pressure change information of the object to be measured within a period of time through a pressure sensor after the inflation operation is completed. If the pressure change information is less than the first set threshold, it is determined that the airtightness of the corresponding object to be measured meets the requirements and is determined to be in a qualified state. If the pressure change information is greater than the first set threshold, the next step is executed;

[0039] An infrared detection module: used to acquire the infrared detection image of the object to be measured captured by an infrared camera, and input the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and output the corresponding airtightness state.

[0040] A third aspect of an embodiment of the present invention discloses an electronic device, including: a memory storing executable program code; a processor coupled to the memory; the processor invoking the executable program code stored in the memory for executing the airtightness detection method based on artificial intelligence disclosed in the first aspect of an embodiment of the present invention.

[0041] A fourth aspect of an embodiment of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program causes a computer to execute the airtightness detection method based on artificial intelligence disclosed in the first aspect of an embodiment of the present invention.

[0042] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0043] In the embodiment of the present invention, the airtightness detection method based on artificial intelligence obtains information of the object to be measured and automatically determines control parameters according to this information, thereby controlling the working states of the air source assembly and the inflation ratio valve. This process is highly automated, reducing the dependence on manual operations and improving the detection efficiency. At the same time, the airtightness detection model is used to identify the infrared detection image, further enhancing the intelligent level of the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 is a flowchart of the airtightness detection method based on artificial intelligence disclosed in an embodiment of the present invention;

[0046] Figure 2 is a specific airtightness detection process disclosed in an embodiment of the present invention;

[0047] Figure 3 is another specific airtightness detection process disclosed in an embodiment of the present invention;

[0048] Figure 4 is a flowchart of the deformation detection disclosed in an embodiment of the present invention;

[0049] Figure 5 is a structural diagram of an airtightness detection system based on artificial intelligence provided by an embodiment of the present invention;

[0050] Figure 6 is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] It should be noted that the terms "first", "second", "third", "fourth", etc. in the specification and claims of the present invention are used to distinguish different objects, rather than to describe a specific order. The terms "comprising" and "having" in the embodiments of the present invention and any variations thereof are intended to cover non-exclusive inclusion. Exemplarily, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0053] The existing devices have the following problems when performing leak tightness detection: First, due to the inability to accurately estimate the volume of the object to be measured, rapid and accurate leak tightness detection cannot be achieved; Second, accurate leak measurement cannot be achieved during measurement. The embodiments of the present invention disclose an artificial intelligence-based leak tightness detection method, system, electronic device and storage medium, which obtain information about the object to be measured and automatically determine control parameters based on this information, thereby controlling the working states of the gas source assembly and the inflation ratio valve. This process is highly automated, reducing the dependence on manual operations and improving the detection efficiency. At the same time, an airtightness detection model is used to identify the infrared detection image, further enhancing the intelligence level of the detection.

[0054] Embodiment 1

[0055] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of an artificial intelligence-based leak tightness detection method disclosed in the embodiments of the present invention. Among them, the execution subject of the method described in the embodiments of the present invention is an execution subject composed of software or / and hardware, which can receive relevant information through wired or / and wireless means and can send certain instructions. Of course, it can also have certain processing functions and storage functions. The execution subject can control multiple devices, such as remote physical servers or cloud servers and related software, or it can also be a local host or server and related software that performs related operations on devices placed somewhere. In some scenarios, multiple storage devices can also be controlled, and the storage devices can be placed in the same place or different places as the devices. As Figure 1 shown, the artificial intelligence-based leak tightness detection method includes the following steps:

[0056] S101: Obtain the information of the object under test, and determine the corresponding control parameter information according to the volume of the object under test and the deformation parameters of the object under test in the information of the object under test;

[0057] S102: Control the working states of the gas source assembly and the inflation ratio valve according to the control parameter information to perform an inflation operation on the object under test according to the set inflation standard, where the inflation operation on the object under test includes inflating the cavity where the object under test is placed or inflating the cavity of the object under test;

[0058] S103: After the inflation operation is completed, obtain the pressure change information of the object under test within a period of time through a pressure sensor. If the pressure change information is less than the first set threshold, determine that the airtightness of the corresponding object under test meets the requirements and determine it as a qualified state. If the pressure change information is greater than the first set threshold, perform the next step;

[0059] S104: Obtain the infrared detection image of the object under test captured by an infrared camera, and input the infrared detection image into the airtightness detection model for recognition to determine the airtightness state of the corresponding object under test, and output the corresponding airtightness state.

[0060] In the embodiment of the present invention, the control parameter information is determined by obtaining the volume and deformation parameters of the object under test, and then the working states of the gas source assembly and the inflation ratio valve are accurately controlled, so as to be able to perform an inflation operation on the object under test according to the set inflation standard. This precise inflation method can ensure that during the detection process, conditions such as the pressure borne by the object under test meet the detection requirements, avoid affecting the detection results due to insufficient or excessive inflation, and improve the accuracy and reliability of the detection.

[0061] After inflation, the pressure change information of the object under test within a period of time is obtained by a pressure sensor and compared with the first set threshold. The airtightness of the object under test can be quickly and preliminarily judged. For the object under test with small pressure change and meeting the requirements, it can be quickly determined as a qualified state, quickly screening out the obviously qualified products, improving the detection efficiency, and reducing unnecessary subsequent detection processes.

[0062] When the pressure change information is greater than the first set threshold, an infrared camera is introduced to capture the infrared detection image of the object under test, and the airtightness detection model is used for recognition. Combining the two methods of pressure detection and infrared image detection realizes multiple detections of the airtightness of the object under test, makes up for the limitations of a single detection method, and makes the detection results more comprehensive and accurate.

[0063] In the embodiments of the present invention, the infrared detection image is input into the airtightness detection model for recognition, making use of the powerful image analysis and pattern recognition capabilities of artificial intelligence technology. It can automatically identify and analyze the feature information in the infrared image, accurately judge the airtightness state of the object to be measured. Compared with the traditional method of manually analyzing images, it greatly improves the detection speed and accuracy, and reduces the interference and error of human factors. Using an infrared camera for detection, which is a non-contact detection method, will not cause physical damage or interference to the object to be measured, and is applicable to objects to be measured of various shapes and materials. Especially for some sensitive or precision components that are not suitable for contact detection, it has good applicability and expands the application range of the detection method.

[0064] The entire detection process, from obtaining information of the object to be measured, controlling inflation, pressure detection to infrared image detection and result output, can be automatically completed by the system without a large amount of manual intervention, improving the automation level of detection, reducing labor costs, and at the same time improving the consistency and repeatability of detection, and being able to ensure the stability of detection quality in large-scale detection.

[0065] In addition to the above reasons, during the specific implementation, in the process of measuring the battery pack, such a situation will occur. Although the overall leakage of the battery pack exceeds the set value, it meets the airtightness requirements because the battery pack has multiple leakage points, but each of these multiple leakage points only leaks a little, which makes the overall leakage a little more, but each leaking a little does not have any impact on the overall airtightness. In order to achieve the above precise detection, the method of using infrared for precise detection is adopted.

[0066] As a complex system, the internal structure of the battery pack contains many components and connection parts, and these parts may all become potential leakage points. For example, the sealed parts of the battery modules, the connection parts of the electrical interfaces, and the splicing parts of the outer shell, etc., may all have tiny gaps due to manufacturing processes, material properties or long-term use, thus causing leakage. Usually, during the detection process, a threshold value of overall leakage is preset as the standard for judging whether the airtightness is qualified. However, this set value is determined based on general situations and experience and cannot fully cover the complex leakage situations of the battery pack. In actual situations, when multiple leakage points exist simultaneously and the leakage amount of each leakage point is small, it may occur that the overall leakage amount exceeds the set value, but the leakage of each leakage point does not affect the overall airtightness.

[0067] When there are multiple leakage points in a battery pack, although each leakage point only leaks a small amount of gas, over time, these small leakage amounts will gradually accumulate, eventually causing the overall leakage amount to exceed the set value. For example, if there are 10 leakage points in a battery pack and each leakage point leaks 0.01 milliliters of gas per second, after 100 seconds, the overall leakage amount reaches 10 × 0.01 × 100 = 10 milliliters. If the set overall leakage threshold is 5 milliliters, the overall leakage amount has exceeded the standard at this time. Although the overall leakage amount exceeds the set value, the impact on airtightness cannot be judged solely based on the leakage amount. Airtightness mainly focuses on whether gas leakage will have a substantial impact on the internal environment of the battery pack, such as whether it will cause an increase in humidity and affect battery performance, or whether it will allow external gas to enter and pose a safety hazard. If the gas leakage amount from each leakage point is extremely small and insufficient to change the key environmental parameters inside the battery pack and does not pose a threat to the normal operation and safety of the battery pack, then from a practical functional perspective, the battery pack can still be considered to meet the airtightness requirements.

[0068] In the airtightness detection method based on artificial intelligence, first, the pressure change information of the object to be measured over a period of time is obtained through a pressure sensor to preliminarily judge the airtightness. However, this method can only detect the overall pressure change situation and cannot accurately identify the existence of multiple small leakage points. When multiple small leakage points act together to cause the overall pressure change to exceed the first set threshold, relying solely on pressure detection may misjudge that the airtightness of the battery pack is unqualified. By introducing an infrared camera to take infrared detection images and using an airtightness detection model for identification, the leakage points can be more accurately located and analyzed. Through the analysis of the infrared images, not only can the positions of the leakage points be determined, but also the leakage degree of each leakage point can be evaluated according to the image features. For example, in the infrared image, the leakage points will show different temperature distribution characteristics from the surrounding areas. By identifying and analyzing these characteristics, the leakage situation of each leakage point can be judged. Even if the overall leakage amount exceeds the set value, but if the infrared image detection shows that the leakage degree of each leakage point is slight and has no obvious impact on the overall performance of the battery pack, then it can be comprehensively judged that the battery pack meets the airtightness requirements. This method combining multiple detection means can more comprehensively and accurately evaluate the airtightness status of the battery pack and avoid misjudgment caused by the limitations of a single detection method.

[0069] More preferably, as Figure 2 shown. The obtaining of the infrared detection image of the object to be measured taken by the infrared camera and inputting the infrared detection image into the airtightness detection model for identification to determine the airtightness status of the corresponding object to be measured and output the corresponding airtightness status includes:

[0070] S1041: Obtain the detection image information set of the object under test captured by the infrared camera within a period of time, and input each detection image information in the detection image information set into the temperature detection model for detection to obtain the temperature state change at each time point;

[0071] S1042: Input the temperature state change into the airtight timing detection model for identification to determine whether the object under test meets the airtightness requirement. If so, determine that the airtightness of the corresponding object under test meets the requirement and determine it as a qualified state. If not, determine that the corresponding object under test does not meet the airtightness requirement and determine it as an unqualified state, and output the airtightness detection result of the corresponding leakage point. The airtightness detection result includes leakage point position information and leakage point size information.

[0072] In the embodiment of the present invention, by obtaining the detection image information set of the object under test within a period of time and inputting each detection image information into the temperature detection model, the temperature state change at each time point can be accurately obtained. When the battery pack leaks, a temperature difference will be generated at the leakage point due to gas flow and other reasons. Such an operation can capture these subtle temperature changes, providing detailed data support for more accurate judgment of airtightness subsequently.

[0073] Input the temperature state change into the airtight timing detection model, which can analyze using time series information. Since airtightness problems often show specific change rules over time, this model can better judge whether the object under test meets the airtightness requirement through time series analysis of the temperature state change, improving the accuracy and reliability of the detection.

[0074] If it is detected that the object under test does not meet the airtightness requirement, the airtightness detection result including leakage point position information and leakage point size information can be output. This is crucial for subsequent repair and improvement work. The staff can quickly locate the problem based on this precise information, understand the severity of the leakage point, and carry out targeted repairs or optimize the design, saving time and cost.

[0075] This detection process constructs a comprehensive and detailed airtightness evaluation system from the capture of temperature changes, to the analysis of the timing model, and then to the output of specific leakage point information. Compared with a single detection method, it can understand the airtightness condition of the battery pack more deeply, effectively avoiding overall performance problems caused by neglecting local small leakage points, and ensuring the quality and safety of the battery pack.

[0076] The airtight timing detection model in the embodiment of the present invention is constructed through the following steps:

[0077] Data collection: Collect an infrared detection image information set of battery packs in a large number of different states, including battery packs with good airtightness and those with leakage problems of different degrees. These data need to cover different battery pack models, different usage environments (such as temperature, humidity, pressure, etc.), and different usage time stages to ensure the diversity and comprehensiveness of the data. For each sample, ensure that a continuous infrared detection image sequence within a period of time is collected so that the time series information of the temperature state can be extracted.

[0078] Image preprocessing: Standardize the infrared detection images, such as operations like cropping, normalization, and denoising, to improve the image quality and consistency. For images of different sizes, they can be scaled to a unified size to ensure the consistency of subsequent model inputs.

[0079] Time series generation: Input the image information of each sample at different time points into the temperature detection model to obtain the corresponding temperature states, and then arrange these temperature states in chronological order to form time series data of the temperature state.

[0080] Basic statistical features: Extract basic statistical features from the time series of the temperature state, such as mean, variance, maximum value, minimum value, median, quartiles, etc. These features can reflect the overall distribution and dispersion degree of the temperature state.

[0081] Trend features: Calculate the trend features of the time series, such as linear trend, seasonal trend, etc. Methods such as linear regression and moving average can be used to analyze the change trend of temperature over time to help the model understand the long-term trend of the temperature state.

[0082] Frequency domain features: Perform Fourier transform on the time series to extract frequency domain features, including the amplitudes and phases of different frequency components, to identify periodic temperature change patterns, which are very useful for analyzing possible periodic phenomena (such as intermittent leakage) during the leakage process.

[0083] Dynamic features: Calculate the dynamic features of the time series, such as autocorrelation coefficient, partial autocorrelation coefficient, etc. These features can reflect the correlation of the temperature state at different time lags and provide information about the dynamic characteristics of the time series for the model.

[0084] Deep learning models such as long short-term memory network (LSTM) or gated recurrent unit (GRU) can be selected. They have excellent performance in processing time series data. These models can automatically learn the long-term dependencies in the time series and are very effective in capturing the complex change patterns of the temperature state over time.

[0085] The dataset after preprocessing and feature extraction is divided into a training set, a validation set, and a test set. Generally, it is divided according to a certain ratio (such as 7:1:2) to ensure the effectiveness of model training, validation, and evaluation.

[0086] For deep learning models, the training set data is used for training, and the model parameters are updated through the backpropagation algorithm to minimize the loss function (such as cross-entropy loss, mean squared error, etc.). At the same time, the validation set is used to monitor the model's performance to prevent overfitting, and the hyperparameters of the model (such as learning rate, number of hidden layers, number of neurons, etc.) are adjusted according to the performance on the validation set. For traditional time series prediction models and machine learning models, the model parameters are adjusted through optimization algorithms (such as gradient descent, stochastic gradient descent, etc.) to achieve optimal performance on the training set and are verified on the validation set.

[0087] More preferably, before obtaining the set of detection image information of the object to be measured captured by the infrared camera over a period of time, it further includes:

[0088] Obtain the image information of the object to be measured captured by the camera, and input the image information of the object to be measured into the detection and recognition model to determine the position information of each detection point in the image information of the object to be measured; or, transmit the image information of the object to be measured captured by the camera to the user terminal, so that the user can circle the corresponding detection area on the user terminal and transmit the corresponding detection area to the detection terminal;

[0089] The step of inputting each detection image information in the set of detection image information into the airtightness detection model for detection to obtain the change of the leakage state at each time point includes:

[0090] Determine the corresponding infrared detection area in each detection image information in the set of detection image information according to the position information of each detection point or the detection area;

[0091] Input the infrared detection area into the airtightness detection model for detection to determine the corresponding temperature change information.

[0092] In the embodiment of the present invention, by inputting the image information of the object to be measured into the detection and recognition model, the position information of each detection point in the image information of the object to be measured can be automatically determined. This automated positioning method helps to improve the accuracy and consistency of detection, and avoids the errors that may be brought by manual marking. For objects to be measured with complex structures, the detection and recognition model can accurately find the key detection points based on its training data and algorithms, providing an accurate position reference for subsequent airtightness detection, and ensuring the scientificity and reliability of the detection process.

[0093] In addition, precise detection point positioning can guide subsequent infrared detection, ensuring that the infrared camera can focus on key areas, thereby improving detection efficiency and avoiding wasting time and resources in irrelevant areas.

[0094] The image information of the object to be measured captured by the camera is transmitted to the user terminal, enabling the user to circle the corresponding detection area, thus realizing the interaction between the user and the detection system. This method provides a customized detection experience for the user. The user can select the areas they consider to be the most critical or in need of attention as the detection areas according to their professional knowledge, experience or specific requirements, enhancing the pertinence and flexibility of the detection. The user's participation makes the detection system more in line with the actual application scenarios, meeting the special requirements of different users for the detection area under different circumstances, and improving the applicability and practicality of the detection system.

[0095] According to the automatically or user-specified detection point position information or detection area, the corresponding infrared detection area can be accurately determined from the detection image information set. This avoids indiscriminate analysis of the entire image, concentrates the detection resources on key areas, and improves the efficiency of infrared detection. At the same time, this targeted area determination method can reduce interference factors and improve the accuracy of detection, because different areas may have different backgrounds and environmental noises, and only focusing on key areas helps to exclude the interference of irrelevant information.

[0096] For some objects to be measured with large sizes or complex structures, only performing infrared detection on key areas can speed up the detection, enable the detection system to obtain valuable information faster, shorten the detection cycle, and improve the overall efficiency of production or detection.

[0097] Inputting the infrared detection area into the airtightness detection model for detection can accurately determine the corresponding temperature change information. Since the detection is performed on the pre-determined key areas, the obtained temperature change information can better reflect the characteristics related to airtightness, reducing the influence of temperature fluctuations in other areas on the detection results, and making the final temperature change information more representative and reliable.

[0098] Such an operation can improve the accuracy of subsequent judgment of the airtightness state based on temperature changes, more precisely detect potential leakage points and leakage situations, contribute to improving the performance of the entire airtightness detection system, and provide better data support for product quality assessment and fault diagnosis. Through these steps, the entire airtightness detection system is not only more flexible and precise in the positioning of detection points and the determination of detection areas, but also more accurate and effective in obtaining the final temperature change information, and can better meet the needs of different users and the characteristics of different objects to be measured, providing better technical guarantee for the airtightness detection of products.

[0099] More preferably, the obtaining of the infrared detection image of the object to be measured captured by the infrared camera and inputting the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured further includes:

[0100] S104a: Obtain the infrared detection image of the object to be measured captured by the infrared camera, and process the infrared detection image to obtain preprocessed infrared detection data;

[0101] S104b: Determine the temperature parameters of each pixel point in the infrared detection data, and determine the gray - scale difference parameter and the gradient difference parameter between adjacent pixel points according to the temperature parameters of each pixel point;

[0102] S104c: Determine the corresponding change node information according to the temperature parameters of each pixel point, the gray - scale difference parameter and the gradient difference parameter between the corresponding data point and adjacent pixel points;

[0103] S104d: Determine the change parameter information between each pixel point and other pixel points according to the change node information of each pixel point in the obtained infrared detection data, and determine the association information between each pixel point and the area of other pixel points according to the transformation parameter information;

[0104] S104e: Perform a clustering operation on the infrared detection data according to the association information between each pixel point and the area of other pixel points to obtain a clustered image result, extract features from the image result to obtain corresponding recognition features, and input the recognition features into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured.

[0105] In the embodiment of the present invention, processing the infrared detection image to obtain preprocessed infrared detection data helps to eliminate noise, distortion and other interference factors in the original image. Through image preprocessing, the contrast and clarity of the image can be enhanced, making the subsequent feature extraction and analysis process more accurate and reliable, and providing a better data basis for subsequent airtightness detection.

[0106] Determining the temperature parameters of each pixel point in the infrared detection data and calculating the gray - scale difference parameter and the gradient difference parameter between adjacent pixel points provide multi - dimensional information for subsequent airtightness analysis. The temperature parameter can directly reflect the thermal distribution on the object surface, while the gray - scale difference and gradient difference parameters can more finely reveal the change of the temperature field, which helps to more sensitively capture possible leakage areas because there are usually sudden changes in the temperature gradient around the leakage point.

[0107] Determining the change node information based on the temperature, grayscale difference, and gradient difference parameters of pixel points can accurately locate the positions in the image where abnormal temperature changes may occur. These change nodes may be the key clues to potential leakage points, helping to narrow the detection range and improve the accuracy and efficiency of detection. By determining the change parameter information and correlation information between each pixel point and other pixel points, a more in-depth analysis of the spatial structure of the entire infrared detection data can be carried out. The determination of this correlation information can grasp the distribution and change law of the temperature field from a global perspective, help identify the diffusion and connection of temperature anomaly regions that may be caused by leakage, avoid looking at each pixel point in isolation, and provide more relevant and systematic data support for subsequent clustering operations.

[0108] Performing a clustering operation on the infrared detection data and classifying similar pixel point regions into one category helps to simplify and integrate complex image information. The clustered image result is more regular, facilitating subsequent feature extraction. By extracting features from the clustered image result, the obtained recognition features can more prominently reflect the information that may be related to airtightness, improving the representativeness and effectiveness of the features input into the airtightness detection model.

[0109] Inputting the extracted recognition features into the airtightness detection model for recognition, and finally determining the airtightness state of the object to be measured. Due to the above series of processing and feature extraction, the model can make judgments based on more representative and systematic features, thereby improving the accuracy and reliability of airtightness detection, reducing the possibility of misjudgment, and providing a more scientific and accurate basis for airtightness assessment.

[0110] More preferably, the obtaining of the information of the object to be measured and the determination of the corresponding control parameter information according to the volume of the object to be measured and the deformation parameter of the object to be measured include:

[0111] S1011: Obtaining the point cloud basic data of the object to be measured through a 3D scanner, identifying the voxels occupied by the object to be measured through image processing technology, and calculating the number of voxels;

[0112] S1012: Judging whether the object to be measured is a known detection object according to the point cloud basic data. If so, directly retrieving the volume parameter and deformation state parameter associated with the corresponding object to be measured. If not, identifying the number of voxels occupied by the object to be measured through image processing technology, determining its volume parameter according to the number of voxels, and determining the corresponding deformation state parameter according to the volume parameter and the point cloud basic data. The deformation state parameter refers to the influence of the air pressure parameter on the deformation of the object to be measured;

[0113] S1013: Determining the corresponding control parameter information according to the volume parameter and the deformation state parameter.

[0114] In the embodiment of the present invention, the point cloud basic data of the object to be measured is obtained through a 3D scanner, and image processing technology is used to identify the voxels occupied by the object to be measured and calculate the number of voxels. This method can provide very accurate geometric information of the object to be measured. Compared with traditional measurement methods, it can comprehensively and meticulously describe the shape and size of the object to be measured from a three-dimensional perspective, avoiding inaccurate data caused by the limitations of manual measurement or two-dimensional measurement, providing more accurate basic data for subsequent detection, especially having significant advantages for objects to be measured with complex shapes.

[0115] According to the point cloud basic data, it is judged whether the object to be measured is a known detection object. For known objects, the corresponding volume parameters and deformation state parameters can be directly retrieved, which can save calculation time and resources. Because for some common objects to be measured, the system can pre-store their relevant parameters without repeated calculation, and can quickly determine the required information, improving the efficiency of the entire detection process.

[0116] For unknown objects to be measured, the volume parameters are determined according to the number of voxels through image processing technology, and the deformation state parameters are determined in combination with the point cloud basic data, ensuring the adaptability of the system to different types of objects to be measured. Even when faced with new types of objects to be measured that have not been detected before, their volume and deformation state parameters can be accurately calculated, expanding the applicable range of the detection system.

[0117] According to the volume parameters and deformation state parameters, the corresponding control parameter information is determined, which can achieve precise control of the inflation process. The volume parameters and deformation state parameters can help determine control parameters such as air pressure, inflation volume, and inflation time suitable for the object to be measured, ensuring that the inflation operation not only meets the detection requirements but also does not cause damage to the object to be measured or affect the detection results. This precise control helps to ensure the safety of the inflation process and the accuracy of the detection, and at the same time provides more stable experimental conditions for subsequent airtightness detection.

[0118] More preferably, the airtightness detection method further includes:

[0119] During the inflation process, the point cloud dynamic data collected by the 3D scanner is obtained, and the voxel association relationship between the point cloud basic data and the point cloud dynamic data is constructed;

[0120] According to the voxel association, the similarity parameter between the point cloud basic data and the point cloud dynamic data is calculated, and the similarity parameter is input into a pre-constructed deformation detection model to determine the corresponding deformation result and output the corresponding deformation result.

[0121] During the inflation process, dynamic point cloud data is obtained, and the voxel association relationship between the basic point cloud data and the dynamic point cloud data is constructed, enabling real-time monitoring of the geometric shape changes of the object under test during inflation. By calculating the similarity parameter between the two, the deformation degree and trend of the object under test can be quantitatively described, providing accurate data basis for subsequent deformation result analysis. This real-time monitoring helps to detect some subtle and imperceptible deformations, improving the perception ability of the deformation of the object under test.

[0122] Inputting the similarity parameter into the pre-constructed deformation detection model, the powerful analysis ability of the model can be used to judge the deformation of the object under test. Since the model is pre-constructed, it can be trained based on a large amount of data, and can learn different deformation states corresponding to different similarity parameters, thus accurately outputting the corresponding deformation results. This avoids the subjectivity and inaccuracy of traditional manual judgment, making the judgment of deformation results more objective and scientific. Through accurate judgment of deformation results, abnormal deformations caused by over-inflation or the object's own structural problems can be detected in advance. For some objects under test that are sensitive to deformation, timely measures can be taken to avoid damage or safety accidents during the detection process, improving the safety and reliability of the entire sealing detection process.

[0123] This technology adds a new dimension to the sealing detection, incorporating the deformation results into the evaluation system. Combining the previous pressure detection and infrared image detection, the object under test can be comprehensively evaluated from multiple angles, providing richer information for judging the airtightness of the object under test. For example, abnormal deformations may imply potential leakage locations or structural defects of the object under test, thus further assisting in judging whether the object under test meets the airtightness requirements.

[0124] More preferably, the operation of controlling the gas source assembly and the inflation ratio valve according to the control parameter information to inflate the object under test according to the set inflation standard includes:

[0125] Determining the target value of the inflation pressure according to the control parameter information, and measuring the actual output pressure of the inflation system in real time through a pressure sensor;

[0126] Comparing the target value of the inflation pressure with the actual output pressure detected by the pressure sensor to calculate the corresponding error parameter;

[0127] Calculating the corresponding control quantity according to the error parameter and the parameter set of the PID controller, the parameter combination includes the proportional coefficient, integral coefficient and differential coefficient;

[0128] Adjusting the inflation parameters of the inflation system with the calculated control quantity to make the actual measured value gradually approach the target value and inflate the object under test.

[0129] In the embodiment of the present invention, the actual output pressure of the inflation system is measured in real time by a pressure sensor and compared with the target value of the inflation pressure, realizing real-time monitoring of the inflation process. This real-time monitoring can ensure that during the inflation operation, the system always grasps the actual situation of the inflation pressure, avoiding errors that may occur when inflating only relying on the set value, and providing an accurate data basis for subsequent pressure adjustment. For example, when inflating some pressure-sensitive objects to be measured, accurate pressure monitoring can prevent damage to the objects to be measured caused by excessive pressure or inability to meet the test requirements due to too low pressure.

[0130] Calculating the error parameter between the target value of the inflation pressure and the actual output pressure can quantify the pressure deviation. This provides a clear basis for subsequent pressure control adjustment, enabling the system to clearly understand the gap between the current inflation pressure and the target pressure, and helping to adopt targeted control strategies subsequently.

[0131] Calculating the control quantity according to the error parameter and the parameter set of the PID controller (including the proportional coefficient, integral coefficient, and differential coefficient) can achieve precise control of the inflation system. Proportional (P) control: The proportional coefficient can adjust the inflation system according to the magnitude of the current error, enabling the inflation system to respond quickly. The larger the error, the greater the adjustment amplitude, which helps to quickly reduce the deviation. Integral (I) control: The integral coefficient takes into account the cumulative effect of the error and can eliminate the steady-state error of the system. For small errors that exist for a long time, integral control can continuously adjust the output of the system, ultimately making the actual pressure reach the target pressure and avoiding long-term deviation of the system. Differential (D) control: The differential coefficient is adjusted according to the rate of change of the error, which helps to predict the trend of the error, make early adjustments to the inflation system, prevent overshoot, and make the system more stable.

[0132] Using the calculated control quantity to adjust the inflation parameters of the inflation system, making the actual measured value gradually approach the target value and performing the inflation operation on the object to be measured, realizes a dynamic and continuously optimized inflation process. This dynamic adjustment can make the inflation operation more stable and accurate. No matter what kind of interference or system fluctuation occurs during the inflation process, it can be continuously adjusted by the PID controller to ensure the stability and accuracy of the inflation process. For objects to be measured with different volumes and different materials, efficient, stable, and precise inflation operations can be achieved according to the set control parameter information and the adjustment of the PID controller, improving the adaptability and reliability of the inflation process.

[0133] In the invention embodiment, fuzzy calculation is also combined to optimize the input of the PID controller. Input variables: The error of the gas pressure and the rate of change of the error are selected as the input quantities of the fuzzy controller. Output variable: The output of the fuzzy controller is the increment Δu of the inflation valve flow correction amount, which is used to adjust the input parameters of the PID controller. Fuzzification process: The input variables are fuzzified to determine the fuzzy levels and membership functions. Fuzzy rules: A fuzzy rule base is established, and based on the fuzzy values of the input variables, the fuzzy value of the output variable is obtained through fuzzy inference. Defuzzification: The center-of-gravity method or the maximum membership function method is used to defuzzify the fuzzy value of the output variable to obtain the accurate correction amounts ΔK_p, ΔK_i, and ΔK_d.

[0134] In the embodiment of the present invention, the artificial intelligence-based airtightness detection method automatically determines control parameters according to the information of the object to be measured, thereby controlling the working states of the gas source assembly and the inflation proportional valve. This process is highly automated, reducing the dependence on manual operations and improving the detection efficiency. At the same time, the airtightness detection model is used to identify the infrared detection image, further enhancing the intelligence level of the detection.

[0135] Embodiment 2

[0136] Please refer to Figure 5 , Figure 5 which is the structural schematic diagram of the artificial intelligence-based airtightness detection system disclosed in the embodiment of the present invention. As Figure 5 shown, the artificial intelligence-based airtightness detection system may include:

[0137] Acquisition module 21: It is used to acquire the information of the object to be measured and determine the corresponding control parameter information according to the volume of the object to be measured and the deformation parameter of the object to be measured in the acquired information of the object to be measured;

[0138] Inflation module 22: It is used to control the working states of the gas source assembly and the inflation proportional valve according to the control parameter information to perform an inflation operation on the object to be measured according to the set inflation standard, wherein, the inflation operation on the object to be measured includes inflating the cavity where the object to be measured is placed or inflating the cavity of the object to be measured;

[0139] Pressure detection module 23: It is used to obtain the pressure change information of the object to be measured within a period of time through a pressure sensor after the inflation operation is completed. If the pressure change information is less than the first set threshold, it is determined that the airtightness of the corresponding object to be measured meets the requirements and is determined to be in a qualified state. If the pressure change information is greater than the first set threshold, the next step is executed;

[0140] Infrared detection module 24: It is used to obtain the infrared detection image of the object to be measured captured by the infrared camera, input the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and output the corresponding airtightness state.

[0141] In the embodiment of the present invention, the artificial intelligence-based airtightness detection method obtains the information of the object to be measured and automatically determines the control parameters according to this information, so as to control the working states of the gas source assembly and the inflation ratio valve. This process is highly automated, reducing the dependence on manual operations and improving the detection efficiency. At the same time, the airtightness detection model is used to identify the infrared detection image, further enhancing the intelligent level of the detection.

[0142] Embodiment III

[0143] Please refer to Figure 6 , Figure 6 which is a schematic structural diagram of an electronic device disclosed in the embodiment of the present invention. The electronic device can be a computer, a server, etc. Of course, in certain cases, it can also be intelligent devices such as mobile phones, tablet computers, and monitoring terminals, as well as an image acquisition device with processing functions. As Figure 6 shown, the electronic device may include:

[0144] A memory 510 storing executable program code;

[0145] A processor 520 coupled to the memory 510;

[0146] Among them, the processor 520 calls the executable program code stored in the memory 510 and executes some or all of the steps in the artificial intelligence-based airtightness detection method in Embodiment I.

[0147] The embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program. Among them, the computer program enables a computer to execute some or all of the steps in the artificial intelligence-based airtightness detection method in Embodiment I.

[0148] The embodiment of the present invention also discloses a computer program product. Among them, when the computer program product runs on a computer, it enables the computer to execute some or all of the steps in the artificial intelligence-based airtightness detection method in Embodiment I.

[0149] The embodiment of the present invention also discloses an application publishing platform. Among them, the application publishing platform is used to publish a computer program product. Among them, when the computer program product runs on a computer, it enables the computer to execute some or all of the steps in the artificial intelligence-based airtightness detection method in Embodiment I.

[0150] In various embodiments of the present invention, it should be understood that the magnitude of the serial numbers of the various processes does not necessarily imply the order of execution. The order of execution of the various processes should be determined based on their functions and internal logic, and should not impose any limitation on the implementation process of the embodiments of the present invention.

[0151] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0152] In addition, in each embodiment of the present invention, the various functional units may be integrated in one processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0153] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc., specifically, the processor in the computer device) to execute some or all of the steps of the methods described in the various embodiments of the present invention.

[0154] In the embodiments provided by the present invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0155] Those of ordinary skill in the art can understand that some or all of the steps in the various methods of the embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The storage medium includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc memories, magnetic disc memories, magnetic tape memories, or any other computer-readable medium capable of carrying or storing data.

[0156] The above has introduced in detail the artificial intelligence-based seal tightness detection method, system, electronic device and storage medium disclosed in the embodiments of the present invention. Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A sealing detection method based on artificial intelligence, characterized in that: include: Acquire information of the object to be measured, and determine corresponding control parameter information according to the volume of the object to be measured and the deformation parameters of the object to be measured; Controlling the working state of the gas source component and the inflation proportional valve according to the control parameter information to inflate the object to be measured according to the set inflation standard, wherein the inflation operation on the object to be measured includes inflating the cavity in which the object to be measured is placed or inflating the cavity of the object to be measured; After the inflation operation is completed, the pressure change information of the object under test within a period of time is obtained through the pressure sensor. If the pressure change information is less than the first set threshold, it is determined that the air tightness of the corresponding object under test meets the requirements and is determined to be in a qualified state. If the pressure change information is greater than the first set threshold, the next step is executed; Acquire an infrared detection image of the object to be measured taken by an infrared camera, input the infrared detection image into an airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and output the corresponding airtightness state; Acquire an infrared detection image of the object to be measured taken by an infrared camera, input the infrared detection image into an airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and also include: Acquire an infrared detection image of the object to be detected captured by an infrared camera, and process the infrared detection image to obtain preprocessed infrared detection data; Determine the temperature parameter of each pixel in the infrared detection data, and determine the grayscale difference parameter and the gradient difference parameter of adjacent pixels according to the temperature parameter of each pixel; According to the temperature parameters of each pixel point, the grayscale difference parameter and the gradient difference parameter of the corresponding data point and the adjacent pixel point are determined to determine the corresponding change node information; Determine the change parameter information between each pixel point and other pixel points according to the change node information of each pixel point in the obtained infrared detection data, and determine the association information between each pixel point and other pixel point areas according to the change parameter information; The infrared detection data is clustered according to the association information between each pixel point and other pixel point areas to obtain a clustered image result, and feature extraction is performed on the image result to obtain corresponding identification features, which are input into the airtightness detection model for identification to determine the airtightness status of the corresponding object under test.

2. The artificial intelligence-based sealing detection method according to claim 1, characterized in that: The method of obtaining an infrared detection image of the object to be tested taken by an infrared camera, inputting the infrared detection image into an airtightness detection model for identification to determine the airtightness state of the corresponding object to be tested, and outputting the corresponding airtightness state includes: Acquire a detection image information set of the object under test captured by an infrared camera within a period of time, and input each detection image information in the detection image information set into a temperature detection model for detection to obtain the temperature state change at each time point; The temperature state change is input into the airtight timing detection model for identification to determine whether the object under test meets the airtightness requirements. If so, it is determined that the airtightness of the corresponding object under test meets the requirements and is determined to be in a qualified state. If not, it is determined that the corresponding object under test does not meet the airtightness requirements and is determined to be in an unqualified state, and the airtightness detection result of the corresponding leakage point is output, and the airtightness detection result includes the leakage point location information and the leakage point size information.

3. The sealing detection method based on artificial intelligence as claimed in claim 2, characterized in that: Before acquiring the detection image information set of the object to be detected captured by the infrared camera within a period of time, the method further includes: Acquire the image information of the object to be tested captured by the camera, and input the image information of the object to be tested into the detection and recognition model to determine the position information of each detection point in the image information of the object to be tested; or, transmit the image information of the object to be tested captured by the camera to the user end, so that the user can circle the corresponding detection area on the user end, and transmit the corresponding detection area to the detection end; The step of inputting each detection image information in the detection image information set into the airtightness detection model for detection to obtain leakage state changes at each time point includes: Determine the corresponding infrared detection area of ​​each detection image information in the detection image information set according to the position information or the detection area of ​​each detection point; The infrared detection area is input into the airtightness detection model for detection to determine the corresponding temperature change information.

4. The sealing detection method based on artificial intelligence according to claim 1, characterized in that: The step of obtaining the measured object information and determining corresponding control parameter information according to the measured object volume and the measured object deformation parameters of the measured object information includes: The point cloud basic data of the object to be measured is obtained through a 3D scanner, and the voxels occupied by the object to be measured are identified through image processing technology, and the number of voxels is calculated; Determine whether the object to be tested is a known test object according to the point cloud basic data. If so, directly retrieve the volume parameters and deformation state parameters associated with the corresponding object to be tested. If not, identify the number of voxels occupied by the object to be tested through image processing technology, determine its volume parameters according to the number of voxels, and determine the corresponding deformation state parameters according to the volume parameters and the point cloud basic data. The deformation state parameters refer to the influence of air pressure parameters on the deformation of the object to be tested. Corresponding control parameter information is determined according to the volume parameters and the deformation state parameters.

5. The sealing detection method based on artificial intelligence according to claim 4, characterized in that: The sealing detection method further comprises: During the inflation process, the point cloud dynamic data collected by the 3D scanner is obtained, and a voxel correlation relationship between the point cloud basic data and the point cloud dynamic data is constructed; The similarity parameters between the point cloud basic data and the point cloud dynamic data are calculated according to the voxel association, and the similarity parameters are input into a pre-built deformation detection model to determine the corresponding deformation result, and the corresponding deformation result is output.

6. The sealing detection method based on artificial intelligence according to claim 1, characterized in that: The controlling the working state of the gas source component and the inflation proportional valve according to the control parameter information to inflate the object to be measured according to the set inflation standard includes: Determine the target value of the inflation pressure according to the control parameter information, and measure the actual output pressure of the inflation system in real time through a pressure sensor; Comparing the target value of the inflation pressure with the actual output pressure detected by the pressure sensor to calculate a corresponding error parameter; Calculating a corresponding control amount according to the error parameter and a set of parameters of the PID controller, wherein the set of parameters includes a proportional coefficient, an integral coefficient and a differential coefficient; The calculated control amount is used to adjust the inflation parameters of the inflation system so that the actual measured value gradually approaches the target value and the inflation operation is performed on the object to be measured.

7. A sealing detection system based on artificial intelligence, characterized in that: include: Acquisition module: used to acquire information of the object to be measured, and determine corresponding control parameter information according to the volume of the object to be measured and the deformation parameters of the object to be measured; Inflation module: used to control the working state of the air source component and the inflation proportional valve according to the control parameter information to inflate the object to be measured according to the set inflation standard, wherein the inflation operation on the object to be measured includes inflating the cavity in which the object to be measured is placed or inflating the cavity of the object to be measured; Pressure detection module: used to obtain the pressure change information of the object under test within a period of time through the pressure sensor after the inflation operation is completed. If the pressure change information is less than the first set threshold, it is determined that the air tightness of the corresponding object under test meets the requirements and is determined to be in a qualified state. If the pressure change information is greater than the first set threshold, the next step is executed; Infrared detection module: used to obtain the infrared detection image of the object to be measured taken by the infrared camera, and input the infrared detection image into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and output the corresponding airtightness state; the infrared detection image of the object to be measured taken by the infrared camera is obtained, and the infrared detection image is input into the airtightness detection model for identification to determine the airtightness state of the corresponding object to be measured, and also includes: Acquire an infrared detection image of the object to be detected captured by an infrared camera, and process the infrared detection image to obtain preprocessed infrared detection data; Determine the temperature parameter of each pixel in the infrared detection data, and determine the grayscale difference parameter and the gradient difference parameter of adjacent pixels according to the temperature parameter of each pixel; According to the temperature parameters of each pixel point, the grayscale difference parameter and the gradient difference parameter of the corresponding data point and the adjacent pixel point are determined to determine the corresponding change node information; Determine the change parameter information between each pixel point and other pixel points according to the change node information of each pixel point in the obtained infrared detection data, and determine the association information between each pixel point and other pixel point areas according to the change parameter information; The infrared detection data is clustered according to the association information between each pixel point and other pixel point areas to obtain a clustered image result, and feature extraction is performed on the image result to obtain corresponding identification features, which are input into the airtightness detection model for identification to determine the airtightness status of the corresponding object under test.

8. An electronic device, characterized in that: include: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the artificial intelligence-based sealing detection method described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program enables a computer to execute the artificial intelligence-based sealing detection method according to any one of claims 1 to 6.

Images