Plastic product diagonal plane defect detection mechanism
By staggering the optical axes of the ring light source and the telecentric lens and combining them with deep learning algorithms, the problems of uneven imaging and material damage in the inspection of oblique sections of plastic products were solved, achieving efficient and stable defect detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- DONGGUAN AUSPICIOUS IMAGE INTELLIGENCE TECH CO LTD
- Filing Date
- 2025-05-10
- Publication Date
- 2026-04-28
AI Technical Summary
Existing technologies are insufficient for effectively detecting defects in the beveled surfaces of plastic products, especially due to uneven imaging of the beveled surfaces and the ease with which materials are damaged during handling.
By misaligning the center of the ring light source with the central optical axis of the telecentric lens, and using an XY-axis adjustment device to move the ring light source, combined with a deep learning algorithm, the uniformity of grayscale on the beveled surface and the avoidance of overexposure at the edges are achieved, allowing for direct defect detection within the material tray.
It improves the accuracy and stability of bevel surface defect detection, avoids damage to materials during handling, and simplifies the detection process.
Smart Images
Figure CN224176431U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of plastic product testing equipment, and in particular to a plastic product bevel surface defect detection mechanism. Background Technology
[0002] Plastic injection molding is one of the molding processes for plastic products. Molten plastic is injected into a plastic product mold under pressure, and then cooled and solidified to obtain the predetermined plastic part.
[0003] Due to the continuous development of optical structures, simple planar cuts can no longer meet the needs of customization, so a large number of injection molded products with beveled cuts have emerged.
[0004] The common method for detecting this type of beveled cut surface on the market is...
[0005] The first method involves adjusting the material angle so that it faces the telecentric lens optical axis, and then detecting it using a coaxial point light source. However, due to gaps in the fixture holes of plastic products and occasional deviations during material handling, industrial black-and-white cameras often produce out-of-focus images and misses in inspections. Additionally, during material handling and inspection, the material may scrape against the side of the material within the tray, potentially damaging it.
[0006] The second method is to use conventional thinking, aligning the center of the ring light source with the central optical axis of the telecentric lens, and taking an image for inspection. However, since the top of the beveled surface of the plastic product is convex, the convex part will be overexposed and bright, resulting in very poor image uniformity of the entire beveled surface, which makes it difficult to detect defects in this beveled surface area. Utility Model Content
[0007] The main purpose of this invention is to propose a defect detection mechanism for the beveled surface of plastic products. The mechanism aims to offset the center of the ring light source from the central optical axis of the telecentric lens, move the ring light source in the XY direction by a motor, observe that the gray scale of the beveled surface of the plastic product is relatively uniform, and that the edge brightness is not overexposed, and then record the position of that point.
[0008] Through algorithms such as deep learning, defects in the entire beveled surface can be detected effectively. Therefore, there is no need to pick up or put down the material; defects in the beveled surface can be detected directly in the material tray, avoiding the material scraping caused by picking up and putting down the material repeatedly.
[0009] To achieve the above objectives, this utility model proposes a defect detection mechanism for the beveled surface of plastic products, comprising:
[0010] An optical experimental platform, which is used to place the object to be tested;
[0011] A visual inspection device, comprising a vertical adjustment device, a camera disposed on the vertical adjustment device, and a telecentric lens disposed below the camera, wherein the vertical adjustment device is used to adjust the distance between the telecentric lens and the object being measured.
[0012] A ring light source is located between the bottom of the camera and the top of the telecentric lens. The ring light source is mounted on an XY-axis adjustment device, which is used to adjust the relative position between the light coverage axis of the ring light source and the central optical axis of the telecentric lens.
[0013] In the actual design, the vision inspection device is set on the vertical adjustment device, so vertical adjustment can be achieved, and the ring light source is set on the XY axis adjustment device, so the relative horizontal position of the ring light source can be adjusted.
[0014] Therefore, by adjusting the relative positions of the light coverage axis and the central optical axis, it can be observed that the gray scale of the beveled surface of the plastic product is relatively uniform. After the edge brightness is not overexposed, the ring light source is lit above the object being tested, and the LED beads emit uniform light that is projected onto the object being tested. The light is reflected by the object being tested and refracted by the lens, and then imaged on the photosensitive element of the industrial black and white camera.
[0015] This allows the defects on the beveled surface of the tested object to form a sharp contrast with the entire beveled surface. Through algorithms such as deep learning (which are based on not overexposing the material, and thus achieve relative position adjustment of the telecentric lens and / or ring light source), the defects on the entire beveled surface can be detected very well. Therefore, there is no need to pick up or put down the material; the defects can be detected directly in the material tray, avoiding the material scraping phenomenon caused by picking up and putting down the material. Attached Figure Description
[0016] Figure 1 This is a three-dimensional schematic diagram of the present invention;
[0017] Figure 2 This is a cross-sectional view of the present invention;
[0018] Figure 3 Product image showing the misalignment of the light coverage axis of the ring light source and the central optical axis of the telecentric lens.
[0019] In the picture,
[0020] 1 represents the optical experimental platform, and 10 represents the object being tested.
[0021] 2 is a visual inspection device, 21 is a camera, and 22 is a telecentric lens.
[0022] 3 is a ring light source.
[0023] 4 is the XY axis adjustment device.
[0024] 5 is the vertical adjustment device, 61 is the light coverage axis, and 62 is the central optical axis. Detailed Implementation
[0025] The technical solutions of the embodiments of this utility model will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this utility model, and not all of them. Based on the embodiments of this utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this utility model.
[0026] It should be noted that if any directional indication (such as up, down, left, right, front, back, top, bottom, inside, outside, vertical, horizontal, longitudinal, counterclockwise, clockwise, circumferential, radial, axial, etc.) is involved in the embodiments of this utility model, the directional indication is only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.
[0027] Furthermore, if the embodiments of this utility model involve descriptions such as "first" or "second," such descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this utility model.
[0028] like Figures 1 to 3 As shown, a defect detection mechanism for beveled surfaces of plastic products includes:
[0029] Optical experimental platform 1, which is used to place the object to be tested 10;
[0030] The visual inspection device 2 includes a vertical adjustment device 5, a camera 21 disposed on the vertical adjustment device 5, and a telecentric lens 22 disposed below the camera 21. The vertical adjustment device 5 is used to adjust the distance between the telecentric lens 22 and the object being measured 10.
[0031] A ring light source 3 is positioned between the bottom of the camera 21 and the top of the telecentric lens 22, and is mounted on the XY-axis adjustment device 4.
[0032] The XY axis adjustment device 4 is used to adjust the relative position between the light coverage axis of the ring light source 3 and the central optical axis of the telecentric lens 22.
[0033] In the actual design, the visual inspection device 2 is mounted on the vertical adjustment device 5, so vertical adjustment can be achieved. The ring light source 3 is mounted on the XY axis adjustment device 4, so the relative horizontal position of the ring light source 3 can be adjusted.
[0034] The ring light source 3 is lit above the object under test 10, and the LED beads emit uniform light that is projected onto the object under test 10. The light is reflected by the object under test 10 and refracted by the telecentric lens 22, forming an image on the photosensitive element of the industrial monochrome camera 21. By adjusting the relative position of the light coverage axis of the ring light source 3 and the central optical axis of the telecentric lens 22 through the XY axis adjustment device 4, it can be observed that the grayscale of the beveled surface of the plastic product is relatively uniform. After the edge brightness is not overexposed, the camera 21 is controlled by the camera software to form an image on the photosensitive element of the camera 21.
[0035] This makes the defects on the beveled surface of the tested object 10 stand out in stark contrast to the entire beveled surface. Through algorithms such as deep learning (which are based on not overexposing the material, and thus achieve relative position adjustment of the telecentric lens 22 and / or the ring light source 3), the defects on the entire beveled surface can be detected very well. Therefore, there is no need to pick up or put down the material; the defects can be detected directly in the material tray, avoiding the material scraping phenomenon caused by picking up and putting down the material.
[0036] Specifically, the ring light source 3 is connected to a light source extension line, which is connected to a brightness interface. The brightness knob of the brightness interface is adjusted to make the light intensity of the ring light source 3 moderate. The brightness interface can be wirelessly controlled and can be set on the optical experimental platform 1 or designed for convenience.
[0037] In this embodiment of the invention, the light coverage axis of the ring light source 3 and the central optical axis of the telecentric lens 22 are misaligned, which allows for relatively uniform grayscale on the beveled surface of the plastic product, preventing overexposure of the edge brightness and effectively detecting defects across the entire beveled surface. Figure 3 This helps reduce the pressure on subsequent image algorithm detection and improves the stability of defect detection.
[0038] Specifically, the illumination surface of the ring light source 3 is located at the bottom wall of the lamp holder, and the illumination surface is inclined. The inclined illumination surface can facilitate the adjustment of the position of the optical axis, thereby avoiding problems such as overexposure and reflection.
[0039] In this embodiment of the invention, the tilt angle of the irradiation surface is 60 degrees. Of course, specific angles of 20 degrees, 30 degrees, 40 degrees, and 70 degrees can also achieve the same technical effect, depending on the specific product being designed.
[0040] Specifically, the camera 21 is a black and white camera or a color camera. By adjusting the relative optical axis position of the ring light source 3, the grayscale of the oblique cut surface of the object under test 10 can be effectively observed, while the edge brightness is not overexposed.
[0041] In this embodiment of the utility model, the XY axis adjustment device 4 includes an X-axis motor and a Y-axis motor disposed on the X-axis motor, and the ring light source 3 is disposed on the Y-axis motor. Of course, in specific embodiments, manual adjustment can also be used to achieve the same technical effect.
[0042] In this embodiment of the invention, the diameter of the ring light source 3 is 150mm. Of course, in specific embodiments, the diameter can be selected according to the size of the product.
[0043] Specifically, the vertical adjustment device 5 uses a lead screw pair to adjust the vertical position or uses manual adjustment of the vertical position.
[0044] This application preferably uses a ring light source 3 with a diameter of 150mm and an LED angle of 60°. The ring light source 3 is positioned below the industrial monochrome camera and above the bottom of the telecentric lens 22, with the bottom of the ring light source 3 75mm away from the bottom of the telecentric lens 22. The ring light source 3 is moved by an XY motor, causing it to synchronously shift in the horizontal XY direction. While moving, observation is performed until the grayscale of the beveled surface of the plastic product is observed to be relatively uniform and the edge brightness is not overexposed. The position of this point is then recorded. Through algorithms such as deep learning, defects across the entire beveled surface can be effectively detected. Therefore, there is no need to pick up or put down the material; defects can be detected directly within the material tray, avoiding material scraping caused by repeated handling.
[0045] By rotating the hand crank of the rotating optical experimental platform 1, the industrial monochrome camera 21 and the telecentric lens 22 are moved up and down via a worm gear transmission and a fixed block. Adjusting the distance between the telecentric lens 22 and the object 10 ensures that the object 10 is clearly displayed in the industrial monochrome camera software window on the computer monitor.
[0046] For example, adjust the exposure parameters of the industrial monochrome camera software to 20ms for taking a picture.
[0047] The above description is only a preferred embodiment of the present utility model and does not limit the patent scope of the present utility model. All equivalent structural transformations made under the inventive concept of the present utility model using the contents of the present utility model specification and drawings, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present utility model.
Claims
1. A defect detection mechanism for beveled surfaces of plastic products, characterized in that, include: An optical experimental platform, which is used to place the object to be tested; A visual inspection device, comprising a vertical adjustment device, a camera disposed on the vertical adjustment device, and a telecentric lens disposed below the camera, wherein the vertical adjustment device is used to adjust the distance between the telecentric lens and the object being measured. A ring light source is positioned between the bottom of the camera and the top of the telecentric lens, and is mounted on the XY-axis adjustment device. The XY axis adjustment device is used to adjust the relative position between the light coverage axis of the ring light source and the central optical axis of the telecentric lens.
2. The plastic product bevel section defect detection mechanism as described in claim 1, characterized in that: The ring light source is connected to a light source extension line, which is connected to a brightness interface.
3. The plastic product bevel surface defect detection mechanism as described in claim 1, characterized in that: The light from the ring light source is positioned so that the optical axis of the ring light source and the central optical axis of the telecentric lens are misaligned.
4. The plastic product bevel surface defect detection mechanism as described in claim 1, characterized in that: The illumination surface of the ring light source is located at the bottom wall of the lamp holder, and the illumination surface is set at an angle.
5. The plastic product bevel surface defect detection mechanism as described in claim 4, characterized in that: The tilt angle of the irradiated surface is 60 degrees.
6. The plastic product bevel defect detection mechanism as described in claim 1, characterized in that: The camera is either a black-and-white camera or a color camera.
7. The plastic product bevel defect detection mechanism as described in claim 1, characterized in that: The XY axis adjustment device includes an X-axis motor and a Y-axis motor located on the X-axis motor, and the ring light source is located on the Y-axis motor.
8. The plastic product bevel surface defect detection mechanism as described in claim 1, characterized in that: The diameter of the ring light source is 150mm.
9. The plastic product bevel surface defect detection mechanism as described in claim 1, characterized in that: The vertical adjustment device uses a lead screw to adjust the vertical position or can be manually adjusted.