Filter element burr detection method based on visual detection

Abstract

The invention discloses a filter element burr detection method based on visual detection, and the method comprises the steps: providing a test fixture which comprises a plurality of cameras for obtaining the images of an object at a test point; a to-be-detected filter element is placed at the test point position, edge detection is carried out on the filter element images obtained by the cameras, filter element edge lines of the to-be-detected filter element in the filter element images are obtained, and the filter element edge lines are used for representing the edges of the to-be-detected filter element; filter element geometric parameters of the to-be-detected filter element are obtained according to the filter element edge line, wherein the filter element geometric parameters are used for representing the size and the shape of the to-be-detected filter element; and determining whether the difference value between each filter element geometric parameter and the corresponding standard geometric parameter is in a corresponding preset range, and if not, determining that the filter element to be detected is a filter element with burrs on the surface. The filter element with burrs on the surface can be removed through the steps, and the test efficiency of the filter element can be improved.

Description

Technical Field

[0001] This application relates to the field of filter cartridge inspection, and more particularly to a method for detecting burrs on filter cartridges based on visual inspection. Background Technology

[0002] The filter elements involved in this application are widely used in biomedicine, life sciences, clinical diagnostics, chemical analysis, sample processing, gas filtration, and other fields to adsorb target particulate matter in sample solutions. When separating target substances in sample solutions through filtration, extraction, or other processes, different filter elements are required to achieve the desired separation effect. However, the shape, size, and other specifications of filter elements vary significantly depending on the application scenario. Manually determining the quality of a filter element requires at least confirming that the filter element surface is free of burrs, the filter element size meets requirements, and there are no other impurities within the filter element, resulting in low detection efficiency. Therefore, an automated method for detecting the quality of filter elements is needed. Summary of the Invention

[0003] The purpose of this application is to provide a visual inspection-based method for detecting burrs on filter cartridges, which can remove filter cartridges with burrs on their surface and improve inspection efficiency.

[0004] To achieve the above objectives, this application provides a visual inspection-based method for detecting burrs in filter cartridges, comprising: A test fixture is provided, the test fixture including a plurality of cameras for acquiring images of objects at test points; The filter element to be tested is placed at the test point, and edge detection is performed on the filter element images acquired by each of the cameras to obtain the filter element edge line in each filter element image. The filter element edge line is used to represent the edge of the filter element to be tested. The filter element geometric parameters are obtained from the edge line of the filter element, and the filter element geometric parameters are used to represent the size and shape of the filter element under test; Confirm whether the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on the surface.

[0005] In this application, when performing edge detection, if the surface of the filter element under test has burrs, the edge line obtained through edge detection will differ significantly from that obtained from a normal filter element. This will result in differences between the filter element's geometric parameters obtained from the edge line and the standard geometric parameters. Therefore, by confirming that the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within a preset range, filter elements with burrs can be rejected, thus improving the testing efficiency of the filter elements.

[0006] Optionally, obtaining the standard geometric parameters includes: A standard mold is placed at the test point, and edge detection is performed on the mold images acquired by each camera to obtain the mold edge line of the standard mold in each mold image. The mold edge line is used to represent the edge of the standard mold. The standard geometric parameters are obtained based on the edge line of the mold, and the standard geometric parameters are used to represent the size and shape of the standard mold.

[0007] Optionally, the camera includes a first camera and a plurality of second cameras; The first camera is used to acquire an image of the object at the test point from a top-down angle; Each of the second cameras is used to acquire images of the object at the test point from different side-view angles.

[0008] Optionally, the filter image includes a first filter image and a second filter image, and the filter edge line includes a first filter edge line and a second filter edge line; The step of performing edge detection on the filter images acquired by each of the cameras to obtain the filter edge line of the filter under test in each of the filter images includes: The filter element to be tested is placed at the test point, and the first filter element edge line of the filter element to be tested in each of the first filter element images is obtained by using the first camera and each of the second cameras. After flipping the filter element under test upside down and rotating it laterally by a preset angle, it is placed at the test point. The filter element under test is placed at the test point, and the edge line of the filter element under test in each of the second filter element images is obtained by using the second filter element images obtained by the first camera and each of the second cameras.

[0009] Optionally, the mold image includes a first mold image and a second mold image, and the mold edge line includes a first mold edge line and a second mold edge line; The step of performing edge detection on the mold images acquired by each of the cameras to obtain the mold edge line of the standard mold in each of the mold images includes: The standard mold is placed at the test point, and the first mold edge line of the standard mold in each of the first mold images is obtained using the first mold images acquired by the first camera and each of the second cameras. After flipping the standard mold upside down and rotating it laterally by the preset angle, it is placed at the test point. The standard mold is placed at the test point, and the edge line of the standard mold in each of the second mold images is obtained using the images of the second mold acquired by the first camera and each of the second cameras.

[0010] Optionally, confirming whether the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within the corresponding preset range includes: Confirm whether the difference between the standard geometric parameter corresponding to the edge line of the first mold and the filter element geometric parameter corresponding to the edge line of the first filter element, and the difference between the standard geometric parameter corresponding to the edge line of the second mold and the filter element geometric parameter corresponding to the edge line of the second filter element, are both within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on its surface.

[0011] Optionally, each of the second cameras is rotate symmetrically arranged around the test point. Attached Figure Description

[0012] Figure 1 This is a schematic flowchart of the filter element burr detection method based on visual inspection according to an embodiment of this application.

[0013] Figures 2 to 5 This is a partial flowchart of the filter burr detection method based on visual inspection according to an embodiment of this application. Detailed Implementation

[0014] To explain in detail the technical content, structural features, objectives and effects of this application, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0015] Please see Figure 1 This application discloses a method for detecting burrs in filter cartridges based on visual inspection, including: S1 provides a test fixture, which includes several cameras for acquiring images of objects at test points.

[0016] S3. Place the filter element to be tested at the test point and perform edge detection on the filter element images acquired by each camera to obtain the filter element edge line in each filter element image. The filter element edge line is used to represent the edge of the filter element to be tested. In other words, the filter element edge line can reflect the approximate outline of the filter element to be tested in the filter element image to a certain extent.

[0017] S5. Obtain the filter element geometric parameters of the filter element to be tested based on the filter element edge line. The filter element geometric parameters are used to represent the size and shape of the filter element to be tested, such as the length of each side of the filter element, the included angle between the side wall and the bottom wall, etc.

[0018] S7. Confirm whether the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on the surface. For example, taking the filter element geometric parameters including the height of the sidewall of the filter element to be tested as an example, assuming that the standard height of the sidewall of the filter element is 5mm and the error range is ±1mm, then when the measured height of the sidewall of the filter element to be tested is 7mm, 7-5=2mm>1mm, the filter element to be tested is a filter element with burrs on the surface.

[0019] The principle of removing burrs from filter elements in this application is briefly described below: Taking a cylindrical filter element as an example, the filter element's geometric parameters include the radii and roundness of both ends, as well as the inclination of the sidewalls. Since the filter element's edge lines can reflect its approximate outline to a certain extent, these geometric parameters can usually be obtained by extracting the outline of the filter element from its edge lines using contour extraction algorithms (such as polygon approximation), thereby determining the geometric parameters of each filter element. During this process, if burrs are present in the filter element under test, the edge detection algorithm will mistakenly identify the burrs as part of the filter element's edge line. The edge line obtained by edge detection will then differ from the edge line of a normal filter element in the image. For example, the edge line of a filter element with burrs in some filter element images will be more distorted than that of a normal filter element. For instance, if a normal filter element has a circular edge line, the edge line of a filter element with burrs might bulge outwards from the circular shape, forming several protrusions. Therefore, when using the algorithm... The outline of the filter element area extracted from the edge line of the filter element area (which can be extracted using polygon approximation and other outline extraction algorithms) will differ from the outline of a normal filter element. This will ultimately lead to a difference between the obtained filter element geometric parameters and the standard geometric parameters. For example, the radius and roundness of the two ends of the filter element under test will be too large, or the inclination of the side wall of the filter element under test will be too large. This will cause the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters to exceed the corresponding preset range. In this way, filter elements with burrs can be rejected (when rejecting, an air gun can be used to blow the unqualified filter elements off the production line).

[0020] It is understood that in steps S3-S7, one of the filter images can be detected first to confirm that the difference between the filter geometric parameters corresponding to the filter image and its corresponding standard geometric parameters is within the corresponding preset range before detecting the next filter image; or the filter geometric parameters of the filter to be tested can be extracted from all the filter images first, and then it can be confirmed whether the difference between each filter geometric parameter and its corresponding standard geometric parameter is within the corresponding preset range.

[0021] It is understandable that there may be other steps between steps S1 to S7. For example, executing step S1 does not necessarily mean executing step S3. Similarly, executing step S3 does not necessarily mean executing step S5, and executing step S5 does not necessarily mean executing step S7.

[0022] In this application, when performing edge detection, if the surface of the filter element under test has burrs, the edge line obtained through edge detection will differ significantly from that obtained from a normal filter element. This will result in differences between the filter element's geometric parameters obtained from the edge line and the standard geometric parameters. Therefore, by confirming that the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within a preset range, filter elements with burrs can be rejected, thus improving the testing efficiency of the filter elements.

[0023] Please see Figure 2 In some embodiments, obtaining standard geometric parameters includes: S21, Place the standard mold at the test point and perform edge detection on the mold images acquired by each camera to obtain the mold edge lines of the standard mold in each mold image. The mold edge lines are used to represent the edges of the standard mold. In other words, the mold edge lines can reflect the approximate outline of the standard mold in the mold image to a certain extent.

[0024] S22, standard geometric parameters are obtained based on the edge line of the mold. The standard geometric parameters are used to represent the size and shape of the standard mold.

[0025] Optionally, the standard mold and the filter element under test are placed in the same way, including the placement angle of the standard mold and the filter element under test. A positioning device can be set at the test point to ensure that the standard mold and the filter element under test are placed in the same way, which helps to eliminate measurement errors caused by differences in shooting angle.

[0026] Specifically, the camera includes a first camera and several second cameras. The first camera is used to acquire images of the object at the test point from a top-down perspective, while each of the second cameras is used to acquire images of the object at the test point from different side-view perspectives. This allows for observation of the object at the test point from both top-down and side-view perspectives, resulting in more comprehensive data.

[0027] It is understandable that standard geometric parameters only need to be obtained once, and the corresponding standard geometric parameters can be used directly when testing the same type of filter element.

[0028] It is understandable that the step numbers above are only for ease of understanding. In reality, there is no specific order in which the standard geometric parameters and the filter element geometric parameters are obtained. It is also possible to obtain the filter element geometric parameters of the filter element to be tested first, and then obtain the standard geometric parameters of the standard mold.

[0029] It should be explained that since the first camera can only acquire a top view of the object at the test point, and each of the second cameras can only acquire a side view of the object at the test point, but the bottom view of the object cannot be acquired, a method is provided below to enable the first camera to acquire the bottom view of the object at the test point, thereby making the image acquired by the camera more comprehensive.

[0030] Please see Figure 3 More specifically, the filter image includes a first filter image and a second filter image, and the filter edge line includes a first filter edge line and a second filter edge line; Edge detection is performed on the filter images acquired by each camera to obtain the filter edge lines of the filter under test in each filter image, including: S31, Place the filter element to be tested in the test point, and use the first filter element images obtained by the first camera and each of the second cameras to obtain the first filter element edge line of the filter element to be tested in each of the first filter element images; S32, after flipping the filter element to be tested upside down and rotating it laterally by a preset angle, place it at the test point. Place the filter element to be tested in the test point and use the second filter element images obtained by the first camera and each second camera to obtain the second filter element edge line of the filter element to be tested in each second filter element image.

[0031] Please see Figure 2 and Figure 4 More specifically, the mold image includes a first mold image and a second mold image, and the mold edge line includes a first mold edge line and a second mold edge line; Accordingly, step S21 includes: S211, Place the standard mold at the test point, and use the first mold image obtained by the first camera and each second camera to obtain the first mold edge line of the standard mold in the first mold image; S212, after flipping the standard mold up and down and rotating it laterally by a preset angle, place it at the test point. Place the standard mold at the test point and use the second mold images obtained by the first camera and each second camera to obtain the second mold edge line of the standard mold in the second mold image.

[0032] Please see Figure 5 Furthermore, step S7 includes: S71, confirm whether the difference between the standard geometric parameters corresponding to the edge line of the first mold and the filter element geometric parameters corresponding to the edge line of the first filter element, and the difference between the standard geometric parameters corresponding to the edge line of the second mold and the filter element geometric parameters corresponding to the edge line of the second filter element are all within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on the surface.

[0033] Optionally, the preset angle is 180°. After flipping vertically, the object's front and back also rotate 180°. At this point, rotating the object laterally by 180° can keep the object's front and back relationship consistent with that before the vertical flip, which helps ensure the accuracy of the measurement results from the first camera.

[0034] Specifically, the second cameras are set up symmetrically around the test point, which can reduce the overlap of the viewing angles observed by the second cameras and make the viewing angles of the second cameras more comprehensive.

[0035] Optionally, the number of second cameras is three. Looking down from above the test point, the positions of the second cameras connected by straight lines form an equilateral triangle.

[0036] The filter element involved in this application can be a filter element made of sintered ultra-high molecular weight polyethylene particles. Therefore, under normal circumstances, the color of the filter element involved in this application is a solid color (e.g., pure white).

[0037] In some embodiments, the visual inspection-based filter burr detection method of this application further includes: S81 uses an edge detection algorithm to confirm that the area corresponding to the filter element in the filter element image is the filter element area.

[0038] Specifically, the filter region can be obtained based on the filter edge lines obtained by the edge detection algorithm. Since the edges of each filter element cooperate to represent the approximate outline of the filter element in the filter image, for a given filter image, there are boundaries between the filter element and its surrounding environment within the filter edge lines. These boundaries are selected from the various filter edge lines, and connecting them sequentially forms the filter region. Of course, using edge detection algorithms to extract the location of an object in an image is not limited to this, and will not be elaborated further.

[0039] S82, confirm whether there are any abnormal filter element colors in each filter element area. If so, the filter element to be tested is an abnormal filter element. Among them, the abnormal point may be impurities, so the filter element needs to be rejected.

[0040] Specifically, the filter to be tested is pure white. Points with abnormal filter colors are those with pixel values ​​less than the preset color value. Taking the RGB channel as an example, the pixel value for white is (255, 255, 255). Considering that there may be color deviations when the camera actually acquires images (for example, a pixel with a value of (253, 253, 253) may actually appear white to the naked eye), the preset color value can be set to 250. When the pixel value of a pixel in all channels is less than 250, that pixel is determined to be a point with an abnormal filter color. In actual examples, points with abnormal filter colors are usually black. Taking the RGB channel as an example, the preset color value can also be set to a smaller value such as 10 to directly determine whether the pixel is a black point.

[0041] In some embodiments, the visual inspection-based filter burr detection method of this application further includes: S91, confirm that the weight deviation of the filter element to be tested is within the preset weight deviation value.

[0042] Specifically, when the filter element to be tested passes the weight sensor on the conveyor belt, the weight of the filter element to be tested is detected, and the filter element to be tested with a weight deviation exceeding the preset weight deviation value (such as the preset weight deviation value of 1%) is rejected.

[0043] S92, confirm that the airflow of the filter element under test meets the preset requirements.

[0044] Specifically, the filter element to be tested enters the test pipe of the airflow testing fixture from the conveyor belt, and one end of the test pipe is connected to the airflow tester. After testing, the filter element to be tested is removed from the airflow testing fixture, and any filter elements that do not meet the preset requirements are discarded.

[0045] S93, for irregular filter elements under test, determine whether the internal condition of the filter element under test is normal.

[0046] Specifically, a telescopic lens is used to extend into the filter element under test to observe the internal condition of the filter element and remove any abnormal filter elements.

[0047] It is understandable that the execution of steps S91 to S93 has no specific order. For example, step S93 can be executed first, followed by step S91.

[0048] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the scope of this application shall still fall within the scope of this application.

Claims

1. A method for detecting burrs in filter cartridges based on visual inspection, characterized in that, include: A test fixture is provided, the test fixture including a plurality of cameras for acquiring images of objects at test points; The filter element to be tested is placed at the test point, and edge detection is performed on the filter element images acquired by each of the cameras to obtain the filter element edge line in each filter element image. The filter element edge line is used to represent the edge of the filter element to be tested. The filter element geometric parameters are obtained from the edge line of the filter element, and the filter element geometric parameters are used to represent the size and shape of the filter element under test; Confirm whether the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on the surface.

2. The filter element burr detection method based on visual inspection as described in claim 1, characterized in that, The acquisition of the standard geometric parameters includes: A standard mold is placed at the test point, and edge detection is performed on the mold images acquired by each camera to obtain the mold edge line of the standard mold in each mold image. The mold edge line is used to represent the edge of the standard mold. The standard geometric parameters are obtained based on the edge line of the mold, and the standard geometric parameters are used to represent the size and shape of the standard mold.

3. The filter element burr detection method based on visual inspection as described in claim 2, characterized in that, The camera includes a first camera and several second cameras; The first camera is used to acquire an image of the object at the test point from a top-down angle; Each of the second cameras is used to acquire images of the object at the test point from different side-view angles.

4. The filter element burr detection method based on visual inspection as described in claim 3, characterized in that, The filter element image includes a first filter element image and a second filter element image, and the filter element edge line includes a first filter element edge line and a second filter element edge line; The step of performing edge detection on the filter images acquired by each of the cameras to obtain the filter edge line of the filter under test in each of the filter images includes: The filter element to be tested is placed at the test point, and the first filter element edge line of the filter element to be tested in each of the first filter element images is obtained by using the first camera and each of the second cameras. After flipping the filter element under test upside down and rotating it laterally by a preset angle, it is placed at the test point. The filter element under test is placed at the test point, and the edge line of the filter element under test in each of the second filter element images is obtained by using the second filter element images obtained by the first camera and each of the second cameras.

5. The filter cartridge burr detection method based on visual inspection as described in claim 4, characterized in that, The mold image includes a first mold image and a second mold image, and the mold edge line includes a first mold edge line and a second mold edge line; The step of performing edge detection on the mold images acquired by each of the cameras to obtain the mold edge line of the standard mold in each of the mold images includes: The standard mold is placed at the test point, and the first mold edge line of the standard mold in each of the first mold images is obtained using the first mold images acquired by the first camera and each of the second cameras. After flipping the standard mold upside down and rotating it laterally by the preset angle, it is placed at the test point. The standard mold is placed at the test point, and the edge line of the standard mold in each of the second mold images is obtained using the images of the second mold acquired by the first camera and each of the second cameras.

6. The filter element burr detection method based on visual inspection as described in claim 5, characterized in that, The step of confirming whether the difference between the geometric parameters of each filter element and the corresponding standard geometric parameters is within the corresponding preset range includes: Confirm whether the difference between the standard geometric parameter corresponding to the edge line of the first mold and the filter element geometric parameter corresponding to the edge line of the first filter element, and the difference between the standard geometric parameter corresponding to the edge line of the second mold and the filter element geometric parameter corresponding to the edge line of the second filter element, are both within the corresponding preset range. If not, the filter element to be tested is a filter element with burrs on its surface.

7. The filter element burr detection method based on visual inspection as described in claim 3, characterized in that, Each of the second cameras is rotated symmetrically around the test point.

Images