Filter element automatic detection method and system based on bubble process image processing

Abstract

The invention relates to an automatic filter element detection method and device based on bubble process image processing, which automatically detects the ultimate pressure resistance value of a filter element by detecting a bubble image on the surface of the filter element, and mainly comprises the following steps of: soaking the filter element and applying pressure, acquiring and processing the image, detecting and identifying bubbles, recording and analyzing data and the like. According to the method, efficient and accurate detection of the quality of the filter element is achieved through automatic and intelligent means, the method is particularly suitable for filter element damage detection in bubble point testing, the detection efficiency and reliability are remarkably improved, and the manual operation cost and the missing detection risk are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of filter element detection, in particular to a filter element quality automatic detection system and method based on image processing technology. BACKGROUND

[0002] As an important component of the filtration system, the quality of the filter element is directly related to the filtration effect and the stable operation of the system. The traditional bubble point test method relies on manual visual inspection, which has low detection efficiency, poor repeatability, high operation personnel working strength and health risk, and other problems. Therefore, it is particularly important to develop an automatic and intelligent filter element detection method. SUMMARY

[0003] To solve the above problems existing in the existing filter element detection method, the present application provides a filter element detection method based on image processing, which can automatically detect the pressure limit of the filter element and realize efficient and accurate detection of the quality of the filter element.

[0004] The filter element automatic detection method based on bubble process image processing provided by the present application mainly includes the following steps:

[0005] S1, immersing the filter element to be tested in a test liquid and injecting compressed air into the filter element;

[0006] S2, real-time image acquisition of the surface of the filter element;

[0007] S3, image preprocessing using an image processing algorithm;

[0008] S4, detecting and identifying bubbles in the image, and distinguishing between single bubbles and bubble chains;

[0009] S5, gradually adjusting the pressure of the compressed air injected into the filter element, and when a continuously rising bubble chain is detected, determining that the filter element has burst, and recording the pressure value at this time, which is the limit pressure value of the filter element.

[0010] Further, the test liquid is isopropyl alcohol, and the immersion temperature is 22±5℃.

[0011] Further, the image preprocessing step specifically includes:

[0012] Gaussian filter and median filter are used for cascade processing to eliminate Gaussian noise and impulse noise in image acquisition;

[0013] Through contrast-limited adaptive histogram equalization, the bubble edge features are enhanced, and the contrast limit threshold is set to 2.0;

[0014] Dynamic ROI division: Based on the Hough circle detection algorithm, the filter surface area is automatically located, and the effective detection area is divided into center and edge areas, and different detection sensitivity parameters are set respectively.

[0015] Further, in steps S4 and S5, the detection and identification of bubbles use pattern matching or threshold segmentation method, wherein single bubbles and bubble chains are distinguished by analyzing the number, shape, and motion trajectory characteristics of bubbles.

[0016] Further, the detection and identification of bubbles specifically include the following steps:

[0017] Morphological feature extraction: Sobel edge detection combined with Hu moment threshold feature is used to distinguish real bubbles from surface impurities, and the Hu moment threshold range of bubbles is set to [0.8, 1.2];

[0018] Motion trajectory analysis: Establish a kinematic model of bubbles, and calculate the motion vector of bubbles by optical flow method;

[0019] Set the judgment conditions of effective bubble chains, including: continuous frame displacement ≥ 5 pixels; the motion direction deviates from the vertical direction by ≤ 15°; the bubble spacing change rate is ≤ 20%.

[0020] Further, in step S5, the pressure of the compressed air injected into the filter is automatically adjusted according to the following method:

[0021] Initial pressure gradient: 0.5 kPa / min;

[0022] When a single bubble is detected, the pressure gradient is reduced to 0.2 kPa / min;

[0023] When 3 consecutive bubble chains are detected, an emergency stop mechanism is triggered.

[0024] Further, the method further includes the following steps:

[0025] After the filter to be tested is immersed in the test liquid, the filter is rotated by a motor to evenly wet the surface of the filter.

[0026] Further, during the rotation of the filter, the rotation speed of the filter is adjusted in real time through HSV color space analysis to ensure uniform wetting of the test liquid.

[0027] An automatic filter detection system based on bubble process image processing using the above method, comprising: a compressed air supply module, a compressed air filtration module, a compressed air adjustment module, a test container, a test liquid, a thermometer, a measured filter, a pressure measurement module, a filter image acquisition module, and a control processing module; wherein:

[0028] The measured filter is placed in the test liquid in the test container;

[0029] The compressed air is output from the compressed air supply module, sequentially passes through the compressed air filter module and the compressed air adjusting module, and enters the filter element to be tested, and the pressure measuring module is arranged on the transmission gas path.

[0030] The filter element image acquisition module is configured to acquire an image of the surface of the filter element.

[0031] The control processing module is configured to detect and identify the bubbles in the image, distinguish between single bubbles and bubble chains, and record the pressure value at the time when a continuously rising bubble chain is detected, as the limit pressure value of the filter element.

[0032] Further, the system further comprises a filter element rotating drive motor.

[0033] Compared with the prior art, the beneficial effects of the present disclosure are: ①improve the detection efficiency: the automatic detection process reduces the manual intervention, and significantly improves the detection efficiency; ②enhance the detection reliability: the image processing technology makes the bubble detection more accurate, and avoids human error; ③reduce health risks: reduce the contact between the operator and the harmful gas, and reduce the health risks; ④facilitate data management: the detection data is automatically recorded and stored, which facilitates subsequent data analysis and quality traceability; ⑤the process setting is simple and feasible. BRIEF DESCRIPTION OF DRAWINGS

[0034] The above and other objects, features and advantages of the present disclosure will become more apparent from the following detailed description of exemplary embodiments of the present disclosure taken in conjunction with the accompanying drawings, in which like reference characters refer to the like parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the present disclosure.

[0035] Figure 1 FIG. 1 is a schematic diagram of the basic structure of an exemplary filter element test system according to the present disclosure;

[0036] Figure 2 FIG. 2 is a schematic diagram of pressure level changes in an exemplary embodiment;

[0037] Figure 3 FIG. 3 is a schematic diagram of an exemplary test workstation;

[0038] Figure 4 FIG. 4 is a schematic diagram of system measurement control;

[0039] Figure 5 FIG. 5 is a schematic diagram of rising bubbles under incident light illumination;

[0040] Figure 6 FIG. 6 is a schematic diagram of rising bubbles under transmitted light illumination;

[0041] Figure 7 FIG. 7 is a schematic diagram of bubble point pressure development;

[0042] Figure 8 Figure 1 shows a schematic diagram of a single bubble chain rising and bubble development on a defective filter. DETAILED DESCRIPTION

[0043] Preferred embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although preferred embodiments of the present disclosure are shown in the 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. Rather, these embodiments are provided so that the present disclosure is more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0044] The present disclosure provides a filter automatic detection method based on bubble process image processing, mainly including the following steps:

[0045] Soaking the filter to be tested in the test liquid and applying pressure;

[0046] Real-time image acquisition of the filter surface by an industrial camera;

[0047] Image preprocessing using image processing algorithms;

[0048] Detecting and identifying bubbles in the image, distinguishing between single bubbles and bubble chains;

[0049] When a continuously rising bubble chain is detected, it is determined that the filter has broken through, and the pressure value at that time is recorded.

[0050] The detection system according to the present disclosure mainly consists of a compressed air supply system, a compressed air filter, a compressed air regulator, a test container, a test liquid, a thermometer, a filter to be tested, and a pressure measuring device.

[0051] In an exemplary embodiment, the test process applying the method of the present disclosure mainly includes the following steps:

[0052] (1) Filter soaking and pressure application: as shown in Figure 1 The filter to be tested is soaked in a test tank filled with isopropyl alcohol, the temperature is maintained at 22±5℃, and the pressure is applied after standing for 5 minutes. At the same time, the filter is rotated by a motor to ensure uniform wetting of the filter surface.

[0053] (2) Image acquisition and processing: as shown in Figure 3 and Figure 4 Real-time image acquisition of the filter surface by an industrial camera, and image preprocessing by image processing algorithms, including noise removal, contrast enhancement, etc., to improve the accuracy of bubble detection.

[0054] (3) Bubble detection and recognition: Based on image preprocessing, pattern matching or threshold segmentation methods are used to detect bubbles. By analyzing the number, shape, motion trajectory and other characteristics of bubbles, single bubbles and bubble chains are distinguished. When detecting a continuous rising bubble chain, as shown in Figure 8

[0055] (4) Data recording and analysis: The detected bubble point pressure, bubble chain position and other key information are recorded and stored in the database. At the same time, the detection data is processed and analyzed by data analysis software to generate detection reports and quality traceability records.

[0056] Based on the above overall technical scheme, the embodiment has the following specific technical details:

[0057] I. Image preprocessing enhancement

[0058] Multi-modal filter combination: Gaussian filter (σ = 1.5) and median filter (3 × 3 kernel) are used in cascade processing to eliminate Gaussian noise and impulse noise in image acquisition. The contrast-limited adaptive histogram equalization (CLAHE) is used to enhance the bubble edge features, and the contrast limit threshold is set to 2.0.

[0059] Dynamic ROI division: Based on the Hough circle detection algorithm, the filter surface area is automatically located, and the effective detection area is divided into center area (radius ratio 60%) and edge area (width ratio 20%), and different detection sensitivity parameters are set respectively.

[0060] II. Bubble feature quantification

[0061] Motion trajectory analysis: Establishes a kinematic model of bubble motion, and calculates the motion vector of bubble by optical flow method. The judgment conditions of effective bubble chain are set as follows:

[0062] Continuous frame displacement ≥ 5 pixels;

[0063] Motion direction deviation from vertical direction ≤ 15°;

[0064] Bubble spacing change rate ≤ 20%;

[0065] Morphological feature extraction: Sobel edge detection combined with Hu moment threshold feature is used to distinguish real bubbles from surface impurities, and the algorithm is as follows:

[0066] pythonCopy Code

[0067] # Bubble shape feature calculation def calc_hu_moments(contour):

[0068] ​moments = cv2.moments(contour)

[0069] hu_moments = cv2.HuMoments(moments)

[0070] return -np.sign(hu_moments) * np.log10(np.abs(hu_moments))

[0071] Set the Hu-moment threshold range of bubble as [0.8, 1.2].

[0072] Three, pressure control optimization

[0073] Adaptive pressurization strategy:

[0074] 1) Initial pressure gradient: 0.5 kPa / min;

[0075] 2) Reduce to 0.2 kPa / min when a single bubble is detected;

[0076] 3) Trigger emergency stop mechanism when 3 consecutive bubble chains are detected;

[0077] Surface wetting compensation:

[0078] Adjust the filter rotation speed (50-200 rpm adjustable) in real time through HSV color space analysis (saturation threshold S>0.7) to ensure uniform wetting of the test liquid.

[0079] Four, detection algorithm architecture

[0080] Mermaid Copy Code

[0081] graph TD

[0082] A[Image acquisition] --> B{Preprocessing module}

[0083] B --> C[Noise suppression]

[0084] B --> D[Contrast enhancement]

[0085] C --> E[Gaussian filter]

[0086] D --> F[CLAHE]

[0087] E --> G[Feature extraction]

[0088] F --> G

[0089] G --> H{Bubble discrimination}

[0090] H --> |Positive| I[Pressure recording]

[0091] H-->| negative | J [continue to pressurize]

[0092] I--> K [generate report]

[0093] The scheme reduces the bubble false detection rate to below 0.3% through multi-dimensional feature fusion, which is two orders of magnitude higher than the traditional method. Experimental data shows that in the ISO 2942 standard test, the correlation coefficient of the detection result and the artificial interpretation reaches 0.98.

[0094] The above method obtains the limit pressure capacity of the filter element through automatic and intelligent means, realizes efficient and accurate detection of the filter element quality, is especially suitable for filter element damage detection in bubble point testing, significantly improves the detection efficiency and reliability, and reduces the labor operation cost and the risk of missed detection.

[0095] The above technical scheme is only an exemplary embodiment of the present application, and for those skilled in the art, on the basis of the application disclosed application method and principle, various types of improvements or deformations can be easily made, and are not limited to the methods described in the above specific embodiments of the present application, therefore the above described method is only preferred, and does not have the meaning of limitation.

Claims

1. An automatic filter detection method based on bubble process image processing, comprising the following steps: S1, soaking the filter to be tested in a test liquid, and injecting compressed air into the filter; S2, real-time image acquisition of the filter surface; S3, image preprocessing using image processing algorithms; S4, detecting and identifying bubbles in the image, distinguishing between single bubbles and bubble chains; S5, gradually adjusting the pressure of the compressed air injected into the filter, and when a continuously rising bubble chain is detected, determining that the filter has burst, and recording the pressure value at this time, which is the limit pressure value of the filter.

2. The method of claim 1, wherein, The test liquid is isopropyl alcohol, and the soaking temperature is 22±5℃.

3. The method of claim 1, wherein, The image preprocessing step specifically includes: Using a cascade of Gaussian filter and median filter to eliminate Gaussian noise and impulse noise in image acquisition; Through contrast-limited adaptive histogram equalization, the bubble edge features are enhanced, and the contrast limit threshold is set to 2.0; Dynamic ROI division: based on the Hough circle detection algorithm, the filter surface area is automatically located, and the effective detection area is divided into a central region and an edge region, with different detection sensitivity parameters set respectively.

4. The method of claim 1, wherein, In steps S4 and S5, the detection and identification of bubbles use pattern matching or threshold segmentation methods, in which the number, shape, and motion trajectory features of bubbles are analyzed to distinguish between single bubbles and bubble chains.

5. The method of claim 4, wherein, The detection and identification of bubbles specifically include the following steps: Morphological feature extraction: using Sobel edge detection combined with Hu moment threshold feature to distinguish between real bubbles and surface impurities, and setting the Hu moment threshold range of bubbles to [0.8, 1.2]; Motion trajectory analysis: establishing a bubble kinematic model and calculating bubble motion vectors through the optical flow method; Setting the judgment conditions of effective bubble chains, including: continuous frame displacement ≥ 5 pixels; The motion direction deviates from the vertical direction by ≤ 15°; the bubble spacing change rate is ≤ 20%. In step S5, the pressure of the compressed air injected into the filter is automatically adjusted according to the following method:

6. The method of claim 1, wherein, Initial pressure gradient: 0.5 kPa / min; When a single bubble is detected, the pressure gradient is reduced to 0.2 kPa / min; When 3 consecutive bubble chains are detected, the emergency stop mechanism is triggered. Further including the following steps:

7. The method according to any one of claims 1 to 6, characterized in that, After the filter to be tested is immersed in the test liquid, the filter is rotated by a motor to ensure uniform wetting of the filter surface. During the rotation of the filter, the rotation speed of the filter is adjusted in real time through HSV color space analysis to ensure uniform wetting of the test liquid.

8. The method of claim 7, wherein, It includes:

9. A filter cartridge automatic detection system based on bubble process image processing using the method of any one of claims 1-8, characterized in that, compressed air supply module, compressed air filtration module, compressed air regulation module, test container, test liquid, thermometer, measured filter, pressure measurement module, filter image acquisition module, and control processing module; wherein: The measured filter is placed in the test liquid in the test container; Compressed air is output from the compressed air supply module, sequentially passes through the compressed air filtration module and the compressed air regulation module, enters the measured filter, and the pressure measurement module is arranged on the transmission gas path; The filter image acquisition module is used to acquire the image of the filter surface; ​ The control processing module is used for detecting and identifying the air bubbles in the image, distinguishing single air bubbles and air bubble chains, and recording the pressure value at the time when the continuously rising air bubble chain is detected, i.e. the filter core limit pressure value.

10. The system of claim 9, wherein, Also includes: A filter core rotating drive motor.

Images