Detection device and detection method based on ultra-low pressure gas leakage detection agent

Abstract

The invention provides a detection device and a detection method based on an ultra-low pressure gas leakage detection agent, and relates to the field of component detection, and the method comprises the steps: coating the ultra-low pressure gas leakage detection agent on a to-be-detected part of a component, and obtaining an image of the to-be-detected part in real time to generate a detection image set; acquiring bubble information corresponding to elements in the detection image set to obtain two-dimensional data of each bubble in the image to be detected at different moments; segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble; according to the three-dimensional data of each bubble, determining leakage severity corresponding to elements in the detection image set; and evaluating the leakage severity of the to-be-detected part of the component according to the leakage severity corresponding to each element so as to obtain an air tightness detection result of the to-be-detected part of the component. According to the invention, missing detection and false detection caused by subjective reasons such as fatigue and attention distraction of people can be avoided.

Description

Technical Field

[0001] The present invention relates to the field of component detection, and in particular to a detection device and a detection method based on an ultra-low pressure gas leakage detection agent. Background Art

[0002] Air tightness testing of components is crucial to ensure product quality. At present, according to the usage scenario and needs, pump-suction combustible gas detectors can be used for testing, or bubble method, gas membrane method or sonar method can be used for testing. However, sonar method testing is difficult to apply to the detection of ultra-low pressure gas leaks, and the bubble method or gas membrane method both require the use of leak detection agents, such as covering the natural gas pipeline with leak detection agents to detect whether there is a natural gas leak in the natural gas pipeline.

[0003] Leak detection agents can be divided into ordinary gas leak detection agents and ultra-low pressure gas leak detection agents. Ultra-low pressure gas leak detection agents are a new type of gas leak visualization detection material that can sensitively respond to gas flow and concentration changes in an ultra-low pressure environment. Users generally judge whether there is a gas leak by whether bubbles are generated.

[0004] At present, the detection and judgment of air tightness based on ultra-low pressure gas leakage detection agents mainly rely on human naked eye observation of bubble generation at the location to be detected. This method is prone to missed detection and false detection, which may lead to deviations in the conclusions of air tightness detection of components. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a detection device and a detection method based on an ultra-low pressure gas leakage detection agent, which can avoid missed detection and false detection caused by subjective reasons such as fatigue and distraction, realize accurate and stable detection of the air tightness of components, and improve the detection efficiency of the air tightness of components.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows: In a first aspect, a detection method based on an ultra-low pressure gas leakage detection agent comprises: Applying an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquiring an image of the location to be detected in real time to generate a detection image set; Acquire bubble information corresponding to the elements in the detection image set, so as to obtain two-dimensional data of each bubble in the image to be detected at different times; Segmenting individual bubbles according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the individual bubbles to obtain the three-dimensional data of each bubble; Determining the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; The leakage severity of the component to be tested is evaluated according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested.

[0007] Furthermore, after evaluating the leakage severity of the component to be tested according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested, it also includes: According to the bubble information corresponding to the first and last elements in the detection image set, the leakage point of the component to be detected is traced.

[0008] Furthermore, the bubble information corresponding to the elements in the detection image set is obtained to obtain the two-dimensional data of each bubble in the image to be detected at different times, including: Numbering the elements in the detection image set in time sequence, wherein the first and last elements in the detection image set are not numbered; Traversing each digitally labeled element, extracting the digitally labeled element and the adjacent elements in the corresponding detection image set, and generating an associated image set corresponding to each digitally labeled element, wherein the elements in the associated image set are arranged along a time axis; Calculating the difference in pixel values ​​at the same position between adjacent elements in the associated image set to obtain two groups of bubble contour pixel points; Merging the two groups of bubble outline pixel points to obtain a first bubble outline image corresponding to the digitally labeled element; Perform edge detection on digitally labeled elements and mark all detected edge lines; Taking a union of the edge lines and the first bubble contour image, and connecting the interruption points of the first bubble contour image to obtain a first complete contour image; The first complete outline image is marked, and each connected area is extracted to obtain complete outline images of each independent bubble in the digitally labeled element.

[0009] Further, a single bubble is segmented according to the two-dimensional data of each bubble, and the three-dimensional data of the single bubble is calculated to obtain the three-dimensional data of each bubble, including: Initializing a background model, wherein the background model includes a plurality of sample values; The background model is updated according to the associated image set corresponding to the digitally labeled elements to generate a static background area; Extracting the outline of the static background area to obtain a mask of the static background area in the digitally labeled element, and performing image segmentation on the digitally labeled element through the mask to extract a background image; Performing a differential operation on the background image and the complete outline of each bubble to determine a single bubble area; The bright spot characteristics of each bubble area are obtained to obtain the bright spot distribution, and the three-dimensional volume parameters of each bubble are determined according to the bright spot distribution.

[0010] Further, determining the leakage severity corresponding to the element in the detection image set according to the three-dimensional data of each bubble includes: Perform wavelet decomposition on the three-dimensional data of each bubble to obtain sub-bands in different directions of low frequency and high frequency; In the low-frequency subband, the volume value is calculated, and the volume variation coefficients in the three axes are counted to obtain the global morphological characteristics of the bubble; In the high-frequency sub-band, the volume distribution histograms of different angle intervals in three axial planes are counted to obtain the detailed features of the local volume; On each subband, extract the volume local descriptor to obtain a three-dimensional feature representation; Concatenate the feature vectors of all sub-bands into a high-dimensional feature vector and perform normalization; The weight vector is set to perform weighted fusion on the normalized high-dimensional feature vector to obtain the index value of the leakage severity corresponding to each bubble; The average index value of all bubbles is calculated, and the leakage severity score corresponding to each element in the detection image set is determined according to the average index value.

[0011] Furthermore, the leakage severity of the component to be tested is evaluated according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested, including: Obtaining a first variation sequence by calculating the variation of leakage severity of adjacent elements in the detection image set; Perform sliding window smoothing on the initial variation sequence to obtain a second variation sequence; Locating the inflection point where the sign of the variation in the second variation sequence changes, and selecting the elements in the detection image set corresponding to before and after the inflection point as the key first evaluation element; Selecting an element in the detection image set corresponding to a local maximum point in the second variation sequence as a second evaluation element; Selecting the remaining elements in the detection image set as the third evaluation element, and setting the first evaluation weight, the second evaluation weight and the third evaluation weight to correspond to the first evaluation element, the second evaluation element and the third evaluation element respectively, wherein the first evaluation weight ≥ the second evaluation weight > the third evaluation weight; The leakage severity scores corresponding to the elements in the detection image set are weighted and fused according to the first evaluation element, the second evaluation element and the third evaluation element to obtain the airtightness detection result of the component to be detected.

[0012] In a second aspect, a detection device based on an ultra-low pressure gas leakage detection agent comprises: A generation module, used for applying an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquiring an image of the location to be detected in real time to generate a detection image set; An acquisition module, used to acquire bubble information corresponding to the elements in the detection image set, so as to obtain two-dimensional data of each bubble in the image to be detected at different times; A calculation module, used for segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble; A determination module, used to determine the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; The evaluation module is used to evaluate the leakage severity of the component to be tested according to the leakage severity corresponding to each element, so as to obtain the airtightness test result of the component to be tested.

[0013] Furthermore, the detection device based on the ultra-low pressure gas leakage detection agent also includes: The tracing module is used to trace the leakage point of the component to be detected according to the bubble information corresponding to the first and last elements in the detection image set.

[0014] According to a third aspect, a computing device includes: one or more processors; The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the above method.

[0015] In a fourth aspect, a computer-readable storage medium stores a program, and the program implements the above method when executed by a processor.

[0016] The above solution of the present invention includes at least the following beneficial effects: The above-mentioned scheme of the present invention realizes rapid and efficient detection of the air tightness of components by using an ultra-low pressure gas leakage detection agent. Compared with the traditional air tightness detection method, the method can avoid missed detection and false detection caused by subjective reasons such as fatigue and distraction, and realizes accurate and stable detection of the air tightness of components, thereby improving the detection efficiency of the air tightness of components; at the same time, the present invention can determine the gas leakage point without sealing the components, thereby reducing the difficulty of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of a detection method based on an ultra-low pressure gas leakage detection agent provided in an embodiment of the present invention.

[0018] Figure 2 It is a schematic diagram of a detection device based on an ultra-low pressure gas leakage detection agent provided in an embodiment of the present invention. Implementation

[0019] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0020] like Figure 1 As shown, an embodiment of the present invention provides a detection method based on an ultra-low pressure gas leakage detection agent, the method comprising: Step 1, applying an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquiring an image of the location to be detected in real time to generate a detection image set; Step 2, obtaining bubble information corresponding to the elements in the detection image set, so as to obtain two-dimensional data of each bubble in the image to be detected at different times; Step 3, segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble; Step 4, determining the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; Step 5, evaluating the leakage severity of the component to be tested according to the leakage severity corresponding to each element, so as to obtain the airtightness test result of the component to be tested.

[0021] In the embodiment of the present invention, the above step 5, after evaluating the leakage severity of the component to be tested according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested, further includes: Step 6: According to the bubble information corresponding to the first and last elements in the detection image set, the leakage point of the component to be detected is traced.

[0022] In the above steps of the embodiment of the present invention, an ultra-low pressure gas leakage detection agent is used, which can quickly reveal even tiny leaks and detect airtightness problems more sensitively; by acquiring images of the location to be detected in real time, the generation and evolution of bubbles can be dynamically observed, and the leakage process can be reflected in time; by analyzing the bubble information at each moment, the size, quantity, change, etc. of a single bubble can be determined, and the severity of the leakage can be evaluated; by calculating the three-dimensional volume parameters of each bubble, the actual size of the bubble can be more accurately determined; by combining the leakage detection results at multiple moments, the overall airtightness performance can be evaluated to increase the comprehensiveness of the judgment; by comparing the bubble information at the beginning and the end, the source of the leakage can be located and the problem area can be found; the present detection method as a whole realizes dynamic evaluation of the airtightness performance of components and leak location, and improves detection efficiency and accuracy; compared with traditional methods, the present method is more sensitive to tiny leaks, has more accurate positioning, and has more reliable results.

[0023] In a preferred embodiment of the present invention, the above step 2, obtaining the bubble information corresponding to the elements in the detection image set, so as to obtain the two-dimensional data of each bubble in the image to be detected at different times, includes: Step 21, numerically labeling the elements in the detection image set in time axis order, wherein the first and last elements in the detection image set are not numbered; Step 22, traversing each digitally labeled element, extracting the digitally labeled element and the adjacent elements in the corresponding detection image set, and generating an associated image set corresponding to each digitally labeled element, wherein the elements in the associated image set are arranged on a time axis; Step 23, calculating the difference in pixel values ​​at the same position between adjacent elements in the associated image set to obtain two groups of bubble contour pixel points; Step 24, merging the two groups of bubble outline pixel points to obtain a first bubble outline image corresponding to the digitally labeled element; Step 25, performing edge detection on the digitally labeled elements and marking all detected edge lines; Step 26, taking a union of the edge lines and the first bubble contour image, and connecting the interruption points of the first bubble contour image to obtain a first complete contour image; Step 27, marking the first complete outline image, and extracting each connected area to obtain complete outline images of each independent bubble in the digitally labeled element.

[0024] In the above steps of the embodiment of the present invention, generating a set of associated images can establish time series data of bubbles, which is conducive to the subsequent use of time series analysis methods to extract dynamic evolution features; the image difference method is used to extract bubble contours, which has a small amount of calculation and is easy to implement real-time processing. There may be omissions, but edge detection can well supplement the shortcomings of the difference method. The combination of the two can extract a more complete bubble contour; obtaining a complete contour is the basis for calculating two-dimensional features such as bubble area and perimeter, which can be used to assess the severity of the leak; aligning the bubble contours at different times can calculate the movement trajectory of the bubble, which is conducive to locating the source of the leak; the time series changes in the number and size of bubbles reflect the dynamic process of gas leakage; the two-dimensional contour image is also the input for the subsequent segmentation of each bubble area, which is the basis for leak location; compared with a single moment, the comprehensive use of information in the time dimension can more comprehensively assess the leakage condition.

[0025] In a preferred embodiment of the present invention, the above step 27, marking the first complete contour image and extracting each connected area to obtain complete contour images of each independent bubble in the digitally labeled element, includes: Step 271, initializing a flag array, wherein the size of the flag array is the same as that of the first complete contour image, and all element values ​​are set to 0; Step 272, starting from the upper left corner of the first complete contour image, scanning is performed. If the element value of the flag array corresponding to a pixel is 0, it is determined that the pixel is not extracted into any contour area and is recorded as the first zero-returned pixel; Step 273, extracting a connected area with the first zeroed pixel as the starting point, and setting the pixels in the connected area of ​​the first zeroed pixel as the first mark value in the mark array; Step 274, continue scanning, if another pixel is found whose corresponding element value of the flag array is 0, it is recorded as the second zeroed pixel, and a connected area is extracted with the second zeroed pixel as the starting point, and the pixels in the connected area of ​​the second zeroed pixel are set to the second flag value in the flag array; repeat until there is no element value of the default value 0 in the flag array; Step 275 , aggregate the connected areas of different label values ​​to obtain the complete bubble outline of each bubble in the digitally labeled element.

[0026] In the above steps of the embodiment of the present invention, the marking array can be used to clearly identify each extracted and to-be-extracted bubble area, avoid repeated extraction and omission, and improve efficiency; starting scanning from the upper left corner, all pixels can be traversed in order to ensure a comprehensive inspection; the extracted area is marked, and the unmarked pixels are used as a new starting point, and the contours of all independent bubbles can be gradually extracted; different marking values ​​represent different bubble contours, and finally the contour of each bubble can be correctly summarized and output; it can effectively deal with complex situations where the number of bubbles in the image is unknown and the overlap is serious, and ensure that all bubbles are extracted; compared with the direct segmentation method, this embodiment can provide an independent and complete contour for each bubble, which provides a guarantee for the subsequent accurate calculation of bubble area, perimeter and other parameters; the accurately extracted contour of each bubble is conducive to constructing a three-dimensional digital model of the bubble and accurately calculating the bubble volume.

[0027] In a preferred embodiment of the present invention, the above step 3, segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble, includes: Step 31, initializing a background model, wherein the background model includes a plurality of sample values; Step 32, updating the background model according to the associated image set corresponding to the digitally labeled element to generate a static background area; Step 33, extracting the outline of the static background area, obtaining a mask of the static background area in the digitally labeled element, and performing image segmentation on the digitally labeled element through the mask to extract a background image; Step 34, performing a differential operation on the background image and the complete outline of each bubble to determine a single bubble area; Step 35, acquiring the bright spot characteristics of each bubble region to obtain bright spot distribution, and determining the three-dimensional volume parameters of each bubble according to the bright spot distribution.

[0028] In the above steps of the embodiment of the present invention, the background modeling technology can be used to effectively extract the static background area of ​​the image, providing a guarantee for the accurate segmentation of each bubble area; the mask image represents the background contour and can be directly used for image segmentation to simplify the processing flow; the difference between the background and the bubble contour can eliminate background interference and accurately separate each bubble area; the segmented bubble image can be used for feature extraction and three-dimensional modeling to calculate the bubble volume; the bright spot feature reflects the three-dimensional structure of the bubble surface, and the three-dimensional data of the bubble can be restored based on the bright spot; the three-dimensional surface of the bubble can be fitted using the three-dimensional data, and the precise volume of the bubble can be calculated; compared with two-dimensional images, the three-dimensional volume parameters can more accurately judge the size of a single bubble and evaluate the severity of the leakage; the three-dimensional parameters of the bubble can also be applied to three-dimensional reconstruction and stereoscopic display to perform three-dimensional visualization of the leakage process.

[0029] In a preferred embodiment of the present invention, the above step 32, updating the background model according to the associated image set corresponding to the digitally labeled element to generate a static background area, includes: Step 321, setting the elements in the associated image set corresponding to the digitally labeled elements to I1, I2, and I3 respectively, where I2 is the digitally labeled element; Step 322, for each pixel point (x, y) in I1, assign its value to a sample S(x, y) and put it into the background model; Step 323, for each pixel (x, y) in I2, search whether there is a sample S (x, y) matching its value in the neighborhood centered on the background model. If so, the point is considered background and updated to the sample corresponding to the background model. If not, it is considered foreground. Step 324, for each pixel (x, y) in I3, search whether there is a sample S (x, y) matching its value in the neighborhood centered at the background model. If so, the point is considered background and updated to the sample corresponding to the background model. If not, it is considered foreground. Step 325, repeat until all sample values ​​in the background model no longer change; Step 326: All sample positions (x, y) included in the background model are regarded as static background areas.

[0030] In the above steps of the embodiment of the present invention, multiple frames of continuous images are set to establish an accurate image time series, which is conducive to the iterative optimization of the background model; the first frame image is globally initialized, and all pixels are added to the model as background samples to achieve a complete global description; in subsequent frames, an adaptive mechanism based on sample value matching is used to update the background model, which can adapt to dynamic changes in the environment such as lighting changes; iterative updates of multiple frames can continuously optimize the parameters of the background model and gradually improve the accuracy and robustness of the background description; when the model tends to be stable, all background sample points are connected to accurately extract the static background area of ​​the entire image sequence; compared with single-frame background extraction, the present invention can achieve adaptive modeling of dynamic environments and greatly improve the quality of background extraction; the background area extracted with high precision can effectively support subsequent target precise segmentation, recognition and analysis, and improve the overall processing effect.

[0031] In a preferred embodiment of the present invention, the above step 35, obtaining the bright spot characteristics of each bubble region to obtain the bright spot distribution, and determining the three-dimensional volume parameters of each bubble according to the bright spot distribution, includes: Step 351, by formula: , denoise the bubble area image, where is the denoised bubble area image, is the bubble area image before denoising, and are domain filter kernel and value filter kernel respectively, for Normalization coefficient of pixel weights in the neighborhood of a point, is the pixel coordinate in the bubble area image before denoising, is located at pixel The pixel coordinates in the neighborhood, for The pixel value of for The pixel value of Step 352, by formula: , detect the corner point R with large local curvature change, where H is the Hessian matrix of the bubble area image, reflecting the local second-order derivative information of the bubble area image. , , are the second-order derivative of the bubble region image in the x direction, the second-order derivative of the bubble region image in the x and y directions, and the second-order derivative of the bubble region image in the y direction. R is the Harris response function, which is calculated by the determinant and trace of the Hessian matrix. k is an empirical parameter. When the R value is large, it means that the point may be a corner point. Step 353, non-maximum suppression and thresholding are used on R to extract the local maximum bright spot and form a bright spot set ; Step 354, by formula: , get the height coordinate z of each bright spot, where, is the basis function, are the horizontal and vertical coordinates of the two-dimensional coordinate points that need to be interpolated, is the two-dimensional coordinate value of the i-th bright spot, is the weight coefficient of the i-th bright spot; Step 355: Fit the three-dimensional surface function of the bubble using the moving least squares method according to the three-dimensional coordinates of the bright spot. ; Step 356, by formula: , calculate the surface area of ​​the i-th bubble region ,in, is the range of the i-th bubble area on the image plane, for The gradient of For Double integrate each point above; Step 357, by formula: , approximate calculation of the three-dimensional volume of each bubble ,in, is the proportionality coefficient.

[0032] In the above steps of the embodiment of the present invention, the combination of domain filtering and value filtering can effectively reduce noise and improve the reliability of subsequent processing; the Hessian matrix detects bright spots and finds feature points with strong light reflection on the bubble surface; non-maximum suppression can extract locally significant bright spots to avoid duplication; using the spatial distribution information of the bright spots, the three-dimensional shape of the bubble surface can be restored by interpolation; moving least squares fitting can make the fitted three-dimensional surface more consistent with the actual shape of the bubble; the surface area parameters of the bubble can be accurately calculated based on the three-dimensional surface; the introduction of the surface area and volume ratio coefficient can establish a mapping model between the two and approximately calculate the volume; ultimately, the three-dimensional volume of the bubble can be accurately calculated through image processing without relying on complex equipment; the three-dimensional parameters can more comprehensively reflect the actual size of the bubble and accurately judge the severity of the leakage.

[0033] In a preferred embodiment of the present invention, the above step 4, determining the leakage severity corresponding to the element in the detection image set according to the three-dimensional data of each bubble, includes: Step 41, performing wavelet decomposition on the three-dimensional data of each bubble to obtain sub-bands in different directions of low frequency and high frequency; Step 42, in the low-frequency sub-band, calculate the numerical value of the volume, and count the volume variation coefficients in three axes to obtain the global morphological characteristics of the bubble; Step 43, in the high frequency sub-band, counting volume distribution histograms in different direction angle intervals in three axial planes to obtain detailed features of the local volume; Step 44, extracting a volume local descriptor on each sub-band to obtain a three-dimensional feature representation; Step 45, concatenate the feature vectors of all sub-bands into a high-dimensional feature vector and perform normalization processing; Step 46, setting a weight vector to perform weighted fusion on the normalized high-dimensional feature vector to obtain an index value of the leakage severity corresponding to each bubble; Step 47, calculating the average index value of all bubbles, and determining the leakage severity score corresponding to each element in the detection image set according to the average index value.

[0034] In the above steps of the embodiment of the present invention, wavelet decomposition can obtain three-dimensional information of bubbles at different frequencies and perform multi-scale analysis; the low-frequency part reflects the global three-dimensional shape, calculates the volume and coefficient of variation, and judges the overall trend; the high-frequency part reflects the local three-dimensional structure, and the distribution in different directions is statistically analyzed to judge the local state; three-dimensional local descriptors are extracted in each frequency band to obtain rich three-dimensional features; multi-scale three-dimensional features are spliced ​​and normalized to obtain stable feature expressions; weighted fusion of multi-source three-dimensional features can comprehensively judge the severity of bubble leakage; statistically calculate the exponential average of all bubbles to judge the leakage state of the image as a whole; compared with a single feature, the present invention can perform three-dimensional multi-scale analysis to improve the accuracy of leakage judgment.

[0035] In a preferred embodiment of the present invention, the above step 5, evaluating the leakage severity of the component to be tested according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested, includes: Step 51, obtaining a first variation sequence by calculating the variation of leakage severity of adjacent elements in the detection image set; Step 52, performing sliding window smoothing on the initial variation sequence to obtain a second variation sequence; Step 53, locating the inflection point where the sign of the variation in the second variation sequence changes, and selecting the elements in the detection image set corresponding to before and after the inflection point as the key first evaluation element; Step 54, selecting the element in the detection image set corresponding to the local maximum point in the second variation sequence as the second evaluation element; Step 55, selecting the remaining elements in the detection image set as the third evaluation element, and setting the first evaluation weight, the second evaluation weight and the third evaluation weight to correspond to the first evaluation element, the second evaluation element and the third evaluation element respectively, wherein the first evaluation weight ≥ the second evaluation weight > the third evaluation weight; Step 56 , weightedly fusion the leakage severity scores corresponding to the elements in the detection image set according to the first evaluation element, the second evaluation element and the third evaluation element, so as to obtain the airtightness detection result of the to-be-detected location of the component.

[0036] In the above steps of the embodiment of the present invention, the leakage change at adjacent moments is calculated to determine the dynamic process of the leakage change; smoothing processing can filter noise to make the change curve smoother; analyzing the inflection point of the change, the key turning point of the leakage change trend can be found; finding the peak point of the change can locate the moment when the leakage is most serious; setting different weights makes the key and peak moments have a greater impact on the overall result; weighted fusion of the detection results of multiple moments can perform dynamic evaluation; compared with a single moment, the present invention makes full use of time dimension information and can more accurately judge the overall air tightness.

[0037] In a preferred embodiment of the present invention, the above step 6, tracing the leakage point of the component to be detected according to the bubble information corresponding to the first and last elements in the detection image set, includes: Step 61, recording the center coordinates and bright spot distribution of each bubble respectively contained in the first and last elements in the detection image set; Step 62, comparing the center coordinate distribution diagrams of each bubble contained in the first and last elements, and locating the center coordinate point of the newly generated bubble; Step 63, determining a possible leakage area according to the newly generated bubble center coordinate point; Step 64, in the possible leakage area, compare the bright spot distribution diagrams of each bubble contained in the first and last elements, locate the area with the largest bright spot distribution change, and take the center of the area with the largest bright spot distribution change as the gas leakage source point.

[0038] In the above steps of the embodiment of the present invention, the position coordinates and bright spot features of the bubbles at critical moments are recorded to provide basic data for leak tracing; by comparing the changes in coordinate distribution at different moments, the possible leakage area where the newly generated bubbles are located can be located; the newly generated bubbles and their areas are often closer to the leakage source, which can narrow the positioning range; within the possible area, the changes in bright spot distribution at different phases are further analyzed; the area with the largest change in bright spot distribution has more severe bubble deformation and is often closest to the leakage source; the precise positioning of the leakage source can be achieved by combining the newly generated coordinate points and the bright spot change area; the present invention makes full use of time dimension information to clarify the location of the leakage source.

[0039] In a preferred embodiment of the present invention, the above step 63, determining the possible leakage area according to the newly generated bubble center coordinate point, includes: Step 631, calculating the distance d between each newly generated bubble center coordinate point and each initial bubble center coordinate point; Step 632, taking each newly generated bubble center coordinate point as the center of the circle, setting a distance threshold R, if the distance d from the initial bubble center coordinate point to the newly generated bubble center coordinate point < R, then the corresponding initial bubble center coordinate point is within the neighborhood of the newly generated bubble center coordinate point; Step 633: The area connecting all the initial bubble center coordinate points in the neighborhood of the newly generated bubble center coordinate point is determined as a possible leakage area.

[0040] In the above steps of the embodiment of the present invention, the distance between the new and old coordinate points is calculated to determine the related old coordinate points that may exist around the new coordinate point; a reasonable distance threshold is set to control the search range of the related old coordinate points; the old coordinate points around the new coordinate point that are less than the threshold distance are connected to locate the possible leakage area; the bubble coordinate information is fully utilized to effectively narrow the scope of the leakage area; compared with direct positioning within the entire detection area, the efficiency of locating the leakage source can be greatly improved.

[0041] In a preferred embodiment of the present invention, the above step 64, in the possible leakage area, compares the bright spot distribution diagrams of each bubble contained in the first and last elements, locates the area with the largest change in bright spot distribution, and takes the center of the area with the largest change in bright spot distribution as the gas leakage source point, including: Step 641, dividing the possible leakage area into multiple sub-areas, and calculating the number of bright spots in the first and last elements of each sub-area; Step 642, calculating the difference Δn between the number of bright spots in the first and last elements of each sub-region, wherein the larger the absolute value of Δn is, the more obvious the change in the bright spot distribution of the sub-region is; Step 643, mapping the absolute value of Δn of each sub-region to the interval [0, 1] as the bright spot change index of the sub-region; Step 644, selecting the sub-region with the largest bright spot change index as the target sub-region, and taking the center of the target sub-region as the target leakage point; Step 645: If the bright spot change indexes of multiple sub-areas are similar, the center points of the multiple sub-areas are connected as the target leakage area.

[0042] In the above steps of the embodiment of the present invention, the possible area is divided into sub-areas, so that a more detailed local analysis can be performed; the change amount of the bright spot in each sub-area at different times is counted, so that the key sub-area with more drastic changes can be located; the change amount of the bright spot is mapped to an index, so that a standardized change index can be obtained; the sub-area with the largest change index is screened out, and the center of the sub-area is most likely to be the leakage point; for multiple areas with similar change indexes, joint positioning can be performed to increase robustness; the present invention makes full use of the bubble bright spot information to achieve high-precision positioning of the leakage point; compared with overall statistics, local sub-area analysis can greatly improve the positioning accuracy.

[0043] like Figure 2 As shown, a detection device based on an ultra-low pressure gas leakage detection agent comprises: A generating module 10 is used to apply an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquire an image of the location to be detected in real time to generate a detection image set; The acquisition module 20 is used to acquire the bubble information corresponding to the elements in the detection image set, so as to obtain the two-dimensional data of each bubble in the image to be detected at different times; A calculation module 30, for segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble; A determination module 40, for determining the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; The evaluation module 50 is used to evaluate the leakage severity of the component to be tested according to the leakage severity corresponding to each element, so as to obtain the airtightness test result of the component to be tested.

[0044] In a preferred embodiment of the present invention, the detection device based on the ultra-low pressure gas leakage detection agent further includes: The tracing module is used to trace the leakage point of the component to be detected according to the bubble information corresponding to the first and last elements in the detection image set.

[0045] It should be noted that the device is a device corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0046] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0047] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0048] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0049] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0050] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0051] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0052] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0053] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0054] In addition, it should be pointed out that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understandable that all or any steps or components of the method and apparatus of the present invention can be implemented in hardware, firmware, software or a combination thereof in any computing device (including processors, storage media, etc.) or a network of computing devices, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0055] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0056] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A detection method based on an ultra-low pressure gas leakage detection agent, characterized in that: The method comprises: Applying an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquiring an image of the location to be detected in real time to generate a detection image set; Acquire bubble information corresponding to the elements in the detection image set, so as to obtain two-dimensional data of each bubble in the image to be detected at different times; Segmenting individual bubbles according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the individual bubbles to obtain the three-dimensional data of each bubble; Determining the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; The leakage severity of the component to be tested is evaluated according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested.

2. The detection method based on the ultra-low pressure gas leakage detection agent according to claim 1, characterized in that: After evaluating the leakage severity of the component to be tested according to the leakage severity corresponding to each element to obtain the airtightness test result of the component to be tested, it also includes: According to the bubble information corresponding to the first and last elements in the detection image set, the leakage point of the component to be detected is traced.

3. The detection method based on the ultra-low pressure gas leakage detection agent according to claim 2 is characterized in that: Acquiring bubble information corresponding to the elements in the detection image set to obtain two-dimensional data of each bubble in the image to be detected at different times, including: Numbering the elements in the detection image set in time sequence, wherein the first and last elements in the detection image set are not numbered; Traversing each digitally labeled element, extracting the digitally labeled element and the adjacent elements in the corresponding detection image set, and generating an associated image set corresponding to each digitally labeled element, wherein the elements in the associated image set are arranged along a time axis; Calculating the difference in pixel values ​​at the same position between adjacent elements in the associated image set to obtain two groups of bubble contour pixel points; Merging the two groups of bubble outline pixel points to obtain a first bubble outline image corresponding to the digitally labeled element; Perform edge detection on digitally labeled elements and mark all detected edge lines; Taking a union of the edge lines and the first bubble contour image, and connecting the interruption points of the first bubble contour image to obtain a first complete contour image; The first complete outline image is marked, and each connected area is extracted to obtain complete outline images of each independent bubble in the digitally labeled element.

4. The detection method based on the ultra-low pressure gas leakage detection agent according to claim 3 is characterized in that: The individual bubbles are segmented according to the two-dimensional data of each bubble, and the three-dimensional data of the individual bubbles are calculated to obtain the three-dimensional data of each bubble, including: Initializing a background model, wherein the background model includes a plurality of sample values; The background model is updated according to the associated image set corresponding to the digitally labeled elements to generate a static background area; Extracting the outline of the static background area to obtain a mask of the static background area in the digitally labeled element, and performing image segmentation on the digitally labeled element through the mask to extract a background image; Performing a differential operation on the background image and the complete outline of each bubble to determine a single bubble area; The bright spot characteristics of each bubble area are obtained to obtain the bright spot distribution, and the three-dimensional volume parameters of each bubble are determined according to the bright spot distribution.

5. The detection method based on the ultra-low pressure gas leakage detection agent according to claim 4 is characterized in that: Determining the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble includes: Perform wavelet decomposition on the three-dimensional data of each bubble to obtain sub-bands in different directions of low frequency and high frequency; In the low-frequency subband, the volume value is calculated, and the volume variation coefficients in the three axes are counted to obtain the global morphological characteristics of the bubble; In the high-frequency sub-band, the volume distribution histograms of different angle intervals in three axial planes are counted to obtain the detailed features of the local volume; On each subband, extract the volume local descriptor to obtain a three-dimensional feature representation; Concatenate the feature vectors of all sub-bands into a high-dimensional feature vector and perform normalization; The weight vector is set to perform weighted fusion on the normalized high-dimensional feature vector to obtain the index value of the leakage severity corresponding to each bubble; The average index value of all bubbles is calculated, and the leakage severity score corresponding to each element in the detection image set is determined according to the average index value.

6. The detection method based on the ultra-low pressure gas leakage detection agent according to claim 5, characterized in that: The leakage severity of the component to be tested is evaluated according to the leakage severity corresponding to each element to obtain the airtightness test results of the component to be tested, including: Obtaining a first variation sequence by calculating the variation of leakage severity of adjacent elements in the detection image set; Perform sliding window smoothing on the initial variation sequence to obtain a second variation sequence; Locating the inflection point where the sign of the variation in the second variation sequence changes, and selecting the elements in the detection image set corresponding to before and after the inflection point as the key first evaluation element; Selecting an element in the detection image set corresponding to a local maximum point in the second variation sequence as a second evaluation element; Selecting the remaining elements in the detection image set as the third evaluation element, and setting the first evaluation weight, the second evaluation weight and the third evaluation weight to correspond to the first evaluation element, the second evaluation element and the third evaluation element respectively, wherein the first evaluation weight ≥ the second evaluation weight > the third evaluation weight; The leakage severity scores corresponding to the elements in the detection image set are weighted and fused according to the first evaluation element, the second evaluation element and the third evaluation element to obtain the airtightness detection result of the component to be detected.

7. A detection device based on an ultra-low pressure gas leakage detection agent, characterized in that: include: A generation module, used for applying an ultra-low pressure gas leakage detection agent to a location to be detected of a component, and acquiring an image of the location to be detected in real time to generate a detection image set; An acquisition module, used to acquire bubble information corresponding to the elements in the detection image set, so as to obtain two-dimensional data of each bubble in the image to be detected at different times; A calculation module, used for segmenting a single bubble according to the two-dimensional data of each bubble, and calculating the three-dimensional data of the single bubble to obtain the three-dimensional data of each bubble; A determination module, used to determine the leakage severity corresponding to the elements in the detection image set according to the three-dimensional data of each bubble; The evaluation module is used to evaluate the leakage severity of the component to be tested according to the leakage severity corresponding to each element, so as to obtain the airtightness test result of the component to be tested.

8. The detection device based on the ultra-low pressure gas leakage detection agent according to claim 7 is characterized in that: Also includes: The tracing module is used to trace the leakage point of the component to be detected according to the bubble information corresponding to the first and last elements in the detection image set.

9. A computing device, characterized in that include: one or more processors; one or more processors; A storage device, used for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 6.

Images