Visual inspection method for welding quality of reverse osmosis membrane

Abstract

The invention discloses a visual inspection method for welding quality of a reverse osmosis membrane, and relates to the technical field of visual inspection. The visual inspection method comprises the following steps: acquiring a top surface backlight image, a top surface light-facing image and a bottom surface backlight image of the reverse osmosis membrane to be detected; performing feature extraction and image segmentation on the preprocessed image through a trained deep learning model; and judging whether the following defects exist or not according to results of feature extraction and image segmentation: surface damage, narrow welding edge width, multiple punching, multiple layers, weld penetration, tearing or welding edge defects. Writing a detection result into a radio frequency identification (RFID) tag corresponding to the reverse osmosis membrane to be detected; and reading data of the RFID tag, and classifying the reverse osmosis membrane to be detected into a qualified area or an unqualified area according to the data of the RFID tag. Therefore, the detection efficiency and the detection precision of the reverse osmosis membrane can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of visual inspection, and in particular to a method for visual inspection of reverse osmosis membrane welding quality. Background Art

[0002] DTRO membrane, also known as disc-tube reverse osmosis membrane, has achieved efficient and stable treatment of landfill leachate with its unique disc-type design. Compared with traditional reverse osmosis technology, DTRO membrane has higher flux, lower energy consumption and longer service life, and is particularly good at treating high-concentration, highly polluted and complex landfill leachate. Its unique flow channel design effectively reduces concentration polarization, improves the interception rate of salt and pollutants, ensures that the effluent water quality is stable and meets the standards, and lays a solid foundation for subsequent wastewater reuse or safe discharge.

[0003] DTRO reverse osmosis membrane is mainly used for water treatment and purification, and has the following characteristics: Material: usually polyamide (PA) or polyester (PE) materials, which have excellent chemical stability and water resistance; Application: widely used in seawater desalination, industrial wastewater treatment and drinking water purification; Importance of quality: high-quality diaphragms can effectively remove dissolved substances and pollutants in water, affecting the efficiency of water treatment and the quality of effluent water; Welding process: diaphragms are usually connected by hot pressing or ultrasonic welding process to ensure the sealing and durability of diaphragms and membrane elements; Precise manufacturing and welding processes are crucial to the performance and life of diaphragms.

[0004] Based on this, the present invention proposes a method for visually inspecting the welding quality of a reverse osmosis membrane. Summary of the invention

[0005] The present invention provides a method for visually inspecting the welding quality of a reverse osmosis membrane, which solves the above-mentioned technical problem of inspecting the welding quality of a reverse osmosis membrane.

[0006] According to one aspect of the present disclosure, a method for visually inspecting welding quality of a reverse osmosis membrane is provided, the method comprising: Acquire a top surface backlight image, a top surface light source image and a bottom surface backlight image of the reverse osmosis membrane to be detected collected by a top surface backlight detection device, a front surface light source detection device and a bottom surface backlight detection device; Preprocessing the top backlit image, the top facing light image and the bottom backlit image; Perform feature extraction and image segmentation on the preprocessed images through the trained deep learning model; Based on the results of feature extraction and image segmentation, determine whether the following defects exist: surface damage, narrow weld edge width, multiple punching, multiple layers, weld penetration, tearing or weld edge defects; If there is a defect, the test result is unqualified, and if there is no defect, the test result is qualified; The test results are written into the RFID tag corresponding to the reverse osmosis membrane to be tested; The data of the radio frequency identification RFID tag is read, and the reverse osmosis membrane to be tested is classified into a qualified area or a unqualified area according to the data of the radio frequency identification RFID tag.

[0007] In one possible implementation, the presence of the following defects is determined based on the results of feature extraction and image segmentation, including: identifying the size and position of the defect based on the grayscale changes of the top backlit image, the top surface lighted image or the bottom backlit image, including: determining the size of the minimum circumscribed rectangle of the damaged position, and determining it as unqualified when the size of one of the sides is greater than 1 mm.

[0008] In a possible implementation, judging whether the following defects exist based on the results of feature extraction and image segmentation includes: dividing the edge of the reverse osmosis membrane to be inspected in the image segmentation results of the top backlit image, the top surface facing light image or the bottom backlit image into small segments of 10 mm, wherein as long as there is a segment where the weld edge width is less than 3 mm, it is judged as unqualified, and the weld edge widths of all segments are greater than or equal to 3 mm, they are judged as qualified.

[0009] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: determining the aspect ratio of the punched segmented rectangular frame in the image segmentation results of the top backlit image, the top surface illuminated image, or the bottom backlit image, and if the aspect ratio is greater than a threshold, determining that there are multiple punchings.

[0010] In one possible implementation, whether the following defects exist is determined based on the results of feature extraction and image segmentation, including: determining whether the reverse osmosis membrane to be tested is multi-layered based on the grayscale average value obtained by feature extraction of the top backlight image or the bottom backlight image, including: if the grayscale average value is between 40-80, it is determined that it is not multi-layered.

[0011] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: determining whether the reverse osmosis membrane to be inspected is welded through based on the grayscale average value obtained by feature extraction of the top backlit image, the top surface facing light image or the bottom backlit image, including: if the grayscale average value is greater than a preset value, it is determined to be welded through.

[0012] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: if the grayscale average value obtained by feature extraction of the top backlit image, the top surface facing light image or the bottom backlit image is between 110 and 120, it is determined that the reverse osmosis membrane to be inspected is a leaky weld.

[0013] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: The welding area in the image segmentation results of the top surface backlit image, the top surface illuminated image and the bottom surface backlit image is segmented into sections of 10 mm each, and the grayscale average value of each section is calculated. If the grayscale average value of one section is greater than 240, the weld is judged to be large.

[0014] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: It is determined that a crack defect exists if the length of the 90-degree corner crack in the image segmentation results of the top surface backlight image, the top surface light image and the bottom surface backlight image is greater than 1 mm.

[0015] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: The welding area in the image segmentation result of the top backlit image or the bottom backlit image is segmented into sections of 10 mm each, and the average grayscale value of each section is determined. If the average grayscale value is less than 140 or greater than 180, it is judged as unqualified.

[0016] Compared with the prior art, the present invention has the following beneficial effects: A method for visually inspecting the welding quality of a reverse osmosis membrane according to an embodiment of the present disclosure is provided. The method collects a backlit image of the top surface, a front-lit image, and a backlit image of the bottom surface of the reverse osmosis membrane to be inspected, and performs feature extraction and image segmentation on the preprocessed image through a trained deep learning model; based on the results of feature extraction and image segmentation, it is determined whether the following defects exist: surface damage, narrow weld edge width, multiple punching holes, multiple layers, weld penetration, tearing, or weld edge quality. The inspection result is written into the radio frequency identification RFID tag corresponding to the reverse osmosis membrane to be inspected; the data of the radio frequency identification RFID tag is read, and the reverse osmosis membrane to be inspected is classified into a qualified area or an unqualified area based on the data of the radio frequency identification RFID tag. In this way, the inspection efficiency and inspection accuracy of the reverse osmosis membrane can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of a method for visually inspecting welding quality of a reverse osmosis membrane according to an embodiment of the present disclosure is shown.

[0018] Figure 2 The images of the reverse osmosis membrane to be inspected are shown, (a) shows the top surface light image of the reverse osmosis membrane to be inspected with damage defects, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be inspected with damage defects.

[0019] Figure 3 The images of the reverse osmosis membrane to be tested are shown, (a) shows the top surface light image of the reverse osmosis membrane to be tested related to the weld edge width detection, and (b) is the processed image of the top surface light image of the reverse osmosis membrane to be tested related to the weld edge width detection.

[0020] Figure 4 The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the image after the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests is processed.

[0021] Figure 5 The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be tested related to multi-layer tests.

[0022] Figure 6 The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be tested related to weld enlargement tests.

[0023] Figure 7 The images of the reverse osmosis membrane to be inspected are shown. (a) shows the top surface light image of the reverse osmosis membrane to be inspected related to multiple punching inspections, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be inspected related to weld edge defect inspection. DETAILED DESCRIPTION

[0024] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0025] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0026] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0027] A method for visually inspecting the welding quality of a reverse osmosis membrane, the method comprising: S01, obtaining a top surface backlight image, a top surface light source image and a bottom surface backlight image of a reverse osmosis membrane to be detected collected by a top surface backlight detection device, a front surface light source detection device and a bottom surface backlight detection device; S02, preprocessing the top backlit image, the top surface illuminated image and the bottom backlit image; S03, feature extraction and image segmentation of the preprocessed image through the trained deep learning model; S04, judging whether there are the following defects based on the results of feature extraction and image segmentation: surface damage, narrow weld edge width, multiple punching, multiple layers, weld penetration, tearing or weld edge defects; S05, if there is a defect, the test result is unqualified, if there is no defect, the test result is qualified; S06, writing the test result into the radio frequency identification RFID tag corresponding to the reverse osmosis membrane to be tested; S07, reading the data of the radio frequency identification RFID tag, and classifying the reverse osmosis membrane to be tested into a qualified area or a unqualified area according to the data of the radio frequency identification RFID tag.

[0028] A method for visual inspection of the welding quality of reverse osmosis membranes according to an embodiment of the present disclosure is provided. The method collects a backlit image of the top surface, a front-lit image and a backlit image of the bottom surface of the reverse osmosis membrane to be inspected, and performs feature extraction and image segmentation on the preprocessed image through a trained deep learning model; based on the results of feature extraction and image segmentation, it is determined whether the following defects exist: surface damage, narrow weld edge width, multiple punching holes, multiple layers, weld penetration, tearing or weld edge defects. The inspection result is written into the radio frequency identification RFID tag corresponding to the reverse osmosis membrane to be inspected; the data of the radio frequency identification RFID tag is read, and the reverse osmosis membrane to be inspected is classified into a qualified area or an unqualified area according to the data of the radio frequency identification RFID tag. In this way, the inspection efficiency and inspection accuracy of the reverse osmosis membrane can be improved.

[0029] In a possible implementation, judging whether the following defects exist based on the results of feature extraction and image segmentation includes: identifying the size and position of the defect based on the grayscale change of the top backlit image, the top surface light image or the bottom backlit image, including: judging the size of the smallest circumscribed rectangle of the damaged position, and judging it as unqualified when the size of one side is greater than 1 mm. The specific detection limit size is 1 mm for each side, and as long as the size of one side is greater than 1 mm, it is judged to be unqualified.

[0030] In a possible implementation, judging whether the following defects exist based on the results of feature extraction and image segmentation includes: dividing the edge of the reverse osmosis membrane to be inspected in the image segmentation results of the top backlit image, the top surface facing light image or the bottom backlit image into small segments of 10 mm, wherein as long as there is a segment where the weld edge width is less than 3 mm, it is judged as unqualified, and the weld edge widths of all segments are greater than or equal to 3 mm, they are judged as qualified.

[0031] For example, along the entire circumference of the weld edge, the circumference is divided into small segments of 10 mm in length. If there is only one segment with a weld edge width less than 3 mm, it is judged as unqualified. If all segments are greater than or equal to 3 mm, they are judged as qualified.

[0032] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: determining the aspect ratio of the punched segmented rectangular frame in the image segmentation results of the top backlit image, the top surface illuminated image, or the bottom backlit image, and if the aspect ratio is greater than a threshold, determining that there are multiple punchings.

[0033] You can also determine whether multiple punching is required by judging the area of ​​the center hole.

[0034] In a possible implementation, judging whether the following defects exist according to the results of feature extraction and image segmentation includes: judging whether the reverse osmosis membrane to be tested is multi-layered according to the grayscale average value obtained by feature extraction of the top backlight image or the bottom backlight image, including: if the grayscale average value is between 40 and 80, judging that it is not multi-layered. The principle of judging multi-layered is that the light transmittance of the image at the multi-layer position is weak.

[0035] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: determining whether the reverse osmosis membrane to be inspected is welded through based on the grayscale average value obtained by feature extraction of the top backlit image, the top surface facing light image or the bottom backlit image, including: if the grayscale average value is greater than a preset value, it is determined to be welded through.

[0036] Alternatively, the minimum detectable weld penetration size is 0.5mm on one side. As long as one side of the hole is larger than 0.5mm, it will be judged as unqualified.

[0037] In one possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: if the grayscale average value obtained by feature extraction of the top backlit image, the top surface facing light image or the bottom backlit image is between 110 and 120, it is determined that the reverse osmosis membrane to be inspected is a leaky weld.

[0038] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: The welding area in the image segmentation results of the top surface backlit image, the top surface illuminated image and the bottom surface backlit image is segmented into sections of 10 mm each, and the grayscale average value of each section is calculated. If the grayscale average value of one section is greater than 240, the weld is judged to be large.

[0039] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: It is determined that a crack defect exists if the length of the 90-degree corner crack in the image segmentation results of the top surface backlight image, the top surface light image and the bottom surface backlight image is greater than 1 mm.

[0040] In a possible implementation, it is determined whether the following defects exist based on the results of feature extraction and image segmentation, including: The welding area in the image segmentation result of the top backlit image or the bottom backlit image is segmented into sections of 10 mm each, and the average grayscale value of each section is determined. If the average grayscale value is less than 140 or greater than 180, it is judged as unqualified.

[0041] Figure 1 A flowchart of a method for visually inspecting welding quality of a reverse osmosis membrane according to an embodiment of the present disclosure is shown.

[0042] Figure 2 The images of the reverse osmosis membrane to be inspected are shown, (a) shows the top surface light image of the reverse osmosis membrane to be inspected with damage defects, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be inspected with damage defects.

[0043] Figure 3 The images of the reverse osmosis membrane to be tested are shown, (a) shows the top surface light image of the reverse osmosis membrane to be tested related to the weld edge width detection, and (b) is the processed image of the top surface light image of the reverse osmosis membrane to be tested related to the weld edge width detection.

[0044] Figure 4 The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the image after the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests is processed.

[0045] Figure 5 The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be tested related to multi-layer tests.

[0046] Figure 6The images of the reverse osmosis membrane to be tested are shown. (a) shows the top surface light image of the reverse osmosis membrane to be tested related to multiple punching tests, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be tested related to weld enlargement tests.

[0047] Figure 7 The images of the reverse osmosis membrane to be inspected are shown. (a) shows the top surface light image of the reverse osmosis membrane to be inspected related to multiple punching inspections, and (b) shows the processed image of the top surface light image of the reverse osmosis membrane to be inspected related to weld edge defect inspection.

[0048] refer to Figure 1-Figure 7 ,from Figure 2 It can be seen from the figure that the red box in (a) is the damaged area, and the blue box in (b) is the damaged area detected in the processed image; Figure 3 It can be seen from the figure that the red box in (a) is the weld edge area, and the green lines in (b) are the weld edge areas detected in the processed image; Figure 4 It can be seen that the red box in (a) is the area with multiple punching, and (b) is the area with multiple punching detected in the processed image; Figure 5 (a) is the image before processing, (b) is the image after processing; Figure 6 It can be seen that the position marked with a red frame is the area with large welding defects; Figure 7 Pictures of weld edge defects before and after treatment can be seen in the figure.

[0049] The embodiments of the present disclosure are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present disclosure. The description of the above embodiments is only used to help understand the method of the present disclosure and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present disclosure, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present disclosure.

[0050] Various aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0051] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0052] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for visually inspecting the welding quality of a reverse osmosis membrane, characterized in that: The method comprises: Acquire a top surface backlight image, a top surface light source image and a bottom surface backlight image of the reverse osmosis membrane to be detected collected by a top surface backlight detection device, a front surface light source detection device and a bottom surface backlight detection device; Preprocessing the top backlit image, the top facing light image and the bottom backlit image; Perform feature extraction and image segmentation on the preprocessed images through the trained deep learning model; Based on the results of feature extraction and image segmentation, determine whether the following defects exist: surface damage, narrow weld edge width, multiple punching, multiple layers, weld penetration, tearing or weld edge defects; If there is a defect, the test result is unqualified, and if there is no defect, the test result is qualified; The test results are written into the RFID tag corresponding to the reverse osmosis membrane to be tested; The data of the radio frequency identification RFID tag is read, and the reverse osmosis membrane to be tested is classified into a qualified area or a unqualified area according to the data of the radio frequency identification RFID tag.

2. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Determine whether the following defects exist based on the results of feature extraction and image segmentation, including: identifying the size and position of the defect based on the grayscale changes of the top backlit image, the top surface illuminated image or the bottom backlit image, including: determining the size of the minimum circumscribed rectangle of the damaged position, and when the size of one of the sides is greater than 1 mm, it is judged to be unqualified.

3. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Judging whether the following defects exist based on the results of feature extraction and image segmentation, includes: based on the edge of the reverse osmosis membrane to be inspected in the image segmentation results of the top backlit image, the top surface lighted image or the bottom backlit image, dividing the edge into small segments of 10 mm, where as long as there is a segment with a weld edge width less than 3 mm, it is judged as unqualified, and all segments with a weld edge width greater than or equal to 3 mm are judged as qualified.

4. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Judging whether the following defects exist based on the results of feature extraction and image segmentation, including: judging the aspect ratio of the punched segmented rectangular frame in the image segmentation results of the top backlit image, the top surface illuminated image or the bottom backlit image, and judging multiple punchings if the aspect ratio is greater than a threshold.

5. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Judging whether the following defects exist based on the results of feature extraction and image segmentation, including: judging whether the reverse osmosis membrane to be tested is multi-layered based on the grayscale average value obtained by feature extraction of the top backlight image or the bottom backlight image, including: if the grayscale average value is between 40-80, it is judged as not multi-layered.

6. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Judging whether the following defects exist based on the results of feature extraction and image segmentation, including: judging whether the reverse osmosis membrane to be inspected is welded through based on the grayscale average value obtained by feature extraction of the top backlit image, the top face-lighted image or the bottom backlit image, including: judging as welded through if the grayscale average value is greater than a preset value.

7. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Based on the results of feature extraction and image segmentation, it is judged whether the following defects exist, including: if the grayscale average value obtained by feature extraction of the top backlit image, the top facing light image or the bottom backlit image is between 110 and 120, it is judged that the reverse osmosis membrane to be tested is a leaky weld.

8. A method for visually inspecting the welding quality of a reverse osmosis membrane according to claim 1, characterized in that: Based on the results of feature extraction and image segmentation, determine whether the following defects exist, including: The welding area in the image segmentation results of the top surface backlit image, the top surface illuminated image and the bottom surface backlit image is segmented into sections of 10 mm each, and the grayscale average value of each section is calculated. If the grayscale average value of one section is greater than 240, the weld is judged to be large.

9. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Based on the results of feature extraction and image segmentation, determine whether the following defects exist, including: It is determined that a crack defect exists if the length of the 90-degree corner crack in the image segmentation results of the top surface backlight image, the top surface light image and the bottom surface backlight image is greater than 1 mm.

10. A method for visually inspecting welding quality of reverse osmosis membrane according to claim 1, characterized in that: Based on the results of feature extraction and image segmentation, determine whether the following defects exist, including: The welding area in the image segmentation result of the top backlit image or the bottom backlit image is segmented into sections of 10 mm each, and the average grayscale value of each section is determined. If the average grayscale value is less than 140 or greater than 180, it is judged as unqualified.

Images