Surface defect detection device

Abstract

The surface defect detection device comprises a moving platform, a line light source, a line array camera and a processor, the processor is electrically connected with the line array camera; the line light source is distributed on the same side of the moving platform relative to the line array camera; the moving platform carries a part to be detected, the moving platform drives the part to be detected to move in a preset moving mode, and the line light source emits linear light to the part to be detected; the line array camera scans the part to be detected row by row to obtain image data of the part to be detected, and the processor receives the image data and correspondingly outputs a surface defect detection result of the part to be detected. The device obtains the image data of the part to be detected through the line array camera, monitors the image data based on the processor, obtains the surface defect detection result of the part to be detected, and can greatly improve the accuracy of surface defect detection compared with the traditional manual detection mode.

Description

Technical Field

[0001] This disclosure relates to the field of industrial production, and in particular to a surface defect detection device. Background Technology

[0002] During the production and use of industrial parts, various surface defects may occur due to factors such as collisions or friction, thus affecting their use. Surface defect detection is one of the important steps in the industrial production process.

[0003] Traditional industrial parts surface defect detection mainly relies on workers to judge whether there are surface defects by visual inspection. However, this method is affected by various factors such as worker skill level and fatigue, resulting in low detection accuracy. Utility Model Content

[0004] The technical problem to be solved by this disclosure is to overcome the low accuracy of manual surface defect detection of parts in the prior art, and to provide a surface defect detection device.

[0005] This disclosure solves the above-mentioned technical problems through the following technical solution:

[0006] This disclosure provides a surface defect detection device, which includes a mobile platform, a line light source, a line scan camera, and a processor, wherein the processor is electrically connected to the line scan camera.

[0007] The line light source and the moving platform are distributed on the same side relative to the line scan camera;

[0008] The mobile platform carries the part to be tested. The mobile platform moves the part to be tested according to a preset moving method. The line light source emits linear light towards the part to be tested.

[0009] The line scan camera scans the part to be inspected line by line to obtain image data of the part to be inspected. The processor receives the image data and outputs the surface defect detection results of the part to be inspected accordingly.

[0010] Optionally, a fixture is provided on the mobile platform, and the mobile platform fixes the part to be inspected by the fixture.

[0011] Optionally, the clamp includes movable jaws;

[0012] The movable jaw passes through the inner hole of the part to be inspected to fix the part. The movable jaw moves radially outward to form an interference fit with the inner hole.

[0013] Optionally, each movable gripper is covered with a silicone pad.

[0014] Optionally, the line light source is a coaxial line light source;

[0015] A coaxial light source is positioned between the line scan camera and the moving platform. A through-hole is provided at the intersection of the coaxial light source and the imaging path of the line scan camera. The coaxial light source emits linear light rays in a direction coaxial with the optical axis of the line scan camera.

[0016] Optionally, the line light source is a dome line light source;

[0017] The dome-shaped line light source is positioned directly above the moving platform, and it diffusely reflects linear light onto the part to be inspected.

[0018] Optionally, linear rays include blue light rays.

[0019] Optionally, the surface defect detection device further includes a filter cover with an opening, through which the blue light emitted by the line light source can pass and be projected onto the part to be inspected.

[0020] Optionally, the surface defect detection device further includes a first base and a first support;

[0021] The first bracket is set on the first base, and the line scan camera is movably mounted on the first bracket;

[0022] Optionally, the surface defect detection device also includes a second base and a second support;

[0023] The second bracket is mounted on the second base, and the line light source is movably mounted on the second bracket.

[0024] Optionally, the mobile platform includes a turntable on which the part to be inspected is carried, and the turntable drives the part to be inspected to rotate at a constant speed along the central axis of the part to be inspected.

[0025] Optionally, the frame rate of the line scan camera is matched to the movement speed of the mobile platform.

[0026] Optionally, the surface defect detection device also includes a display;

[0027] The display is electrically connected to the processor and is used to display the received image data and / or surface defect detection results.

[0028] The positive progress of this disclosure is that it proposes a surface defect detection device. The device acquires image data of the part to be inspected through a line scan camera and monitors the image data based on a processor to obtain the surface defect detection results of the part to be inspected. Compared with the traditional manual inspection method, it can greatly improve the accuracy of surface defect detection. Attached Figure Description

[0029] Figure 1 A first structural schematic diagram of a surface defect detection device provided as an exemplary embodiment of this disclosure;

[0030] Figure 2 A second structural schematic diagram of a surface defect detection device provided as an exemplary embodiment of this disclosure;

[0031] Figure 3 A third structural schematic diagram of a surface defect detection device provided as an exemplary embodiment of this disclosure;

[0032] Figure 4 This is a schematic diagram of the structure of a clamp provided for an exemplary embodiment of the present disclosure. Detailed Implementation

[0033] The present disclosure is further illustrated below by way of embodiments, but the present disclosure is not limited to the scope of the embodiments described herein.

[0034] An exemplary embodiment of this disclosure provides a surface defect detection device, see [link to example]. Figure 1 The device includes a mobile platform 11, a line light source 12, a line scan camera 13, and a processor 14. The processor 14 is electrically connected to the line scan camera 13.

[0035] The line light source 12 and the moving platform 11 are arranged on the same side relative to the line scan camera 13, generally located directly in front of the lens of the line scan camera 13. The moving platform 11 carries the part to be inspected 2, and the moving platform 11 moves the part to be inspected 2 according to a preset moving mode. The line light source 12 emits linear light rays to the part to be inspected 2. The line scan camera 13 scans the part to be inspected 2 line by line to obtain image data of the part to be inspected 2. The processor 14 receives the image data and outputs the surface defect detection results of the part to be inspected 2 accordingly.

[0036] The line light source 12 is an illumination device that produces a long, thin, uniform light band. The emitted linear light is bright and uniform, suitable for detecting subtle features (such as scratches and textures). The line scan camera 13 is a camera that scans the part 2 to be inspected line by line using a single row of photosensitive elements and finally synthesizes a complete image. The line scan camera 13 features high-speed imaging, suitable for detecting fast-moving objects, and also features high resolution and low distortion, reducing errors caused by image noise during surface defect detection.

[0037] The preset movement method can be selected according to the shape of the part 2 to be inspected, including but not limited to translation from left to right, translation from top to bottom, and rotation at clockwise or counterclockwise angles. Rotation is usually used to ensure that every face of the part 2 to be inspected can be fully detected. Alternatively, for cubic parts to be inspected, a combination of translation and rotation can be used.

[0038] The surface defect detection results can include whether the part 2 under test has surface defects, the location of the surface defects of the part 2 under test, the type of surface defects of the part 2 under test, etc., but are not limited to these. The specific settings should be made according to the actual situation.

[0039] In this embodiment, a surface defect detection device is proposed. The device acquires image data of the part 2 to be inspected through a line scan camera 13, and monitors the image data based on a processor 14 to obtain the surface defect detection results of the part 2. Compared with traditional manual inspection methods, this significantly improves the accuracy of surface defect detection. Simultaneously, the moving platform 11 can guide the part 2 to be inspected during the imaging process of the line scan camera 13, ensuring the comprehensiveness of the surface defect detection. Furthermore, the linear illumination provided by the line light source 12 highlights the defects on the part 2, further improving the accuracy and efficiency of surface defect detection.

[0040] In an optional embodiment, mobile platform 11 is described, see [link to relevant documentation]. Figure 2 and Figure 3 The mobile platform 11 is equipped with a clamp 111, which is generally used to fix the part 2 to be tested.

[0041] There are generally no major restrictions on the form of the clamp 111, such as a chuck or a sleeve; the specific choice can be made according to the actual application scenario. For the preset movement method of acquiring image data by rotating the part 2 to be inspected, this embodiment also provides another specific implementation of the clamp 111, see [link to relevant documentation]. Figure 4 The clamp 111 includes a movable gripper 1111.

[0042] The movable gripper 1111 can extend and retract radially to fix the part 2 to be inspected. Specifically, the movable gripper 1111 passes through the inner hole of the part 2 to fix the part 2 to be inspected. The movable gripper 1111 forms an interference fit with the inner hole by moving radially outward.

[0043] Meanwhile, the radial extension and retraction of the movable gripper 1111 can also match the inner hole dimensions of different parts 2 to be inspected, ensuring the concentricity of the movable gripper 1111 and the parts 2 to be inspected during rotation. Concentricity is the geometric tolerance of whether the axes of two or more rotating bodies coincide. On the one hand, a large concentricity will cause collisions between the contact surfaces of the movable gripper 1111 and the parts 2 to be inspected, leading to wear on the parts 2 to be inspected. On the other hand, a large concentricity will cause a mismatch between the movement trajectory of the parts 2 to be inspected and the movement trajectory of the moving platform 11, affecting the imaging effect when imaging by the line scan camera 13. Generally speaking, image data with high concentricity will show obvious jitter and ripples compared to image data with low concentricity. This type of image data is prone to interfering with the surface defect detection results when applied to the detection of some small surface defects in the parts 2 to be inspected.

[0044] In addition, in some application scenarios, each movable gripper 1111 can be wrapped with a silicone pad to avoid damage to the part 2 to be inspected by the movable gripper 1111.

[0045] In one specific embodiment, the mobile platform 11 that acquires image data by rotating the part 2 to be inspected is further described, see [link to relevant documentation]. Figure 2 and Figure 3 The mobile platform 11 includes a turntable 112, on which the part to be inspected 2 is mounted. The turntable 112 drives the part to be inspected 2 to rotate at a constant speed. When performing surface defect inspection on some cylindrical parts using a line scan camera 13 in conjunction with the turntable 112, the side of the cylinder can be unfolded into a rectangle, avoiding the influence of image distortion on cylindrical parts and reducing the high cost of acquiring image data.

[0046] In one specific embodiment, the line scan camera 13 acquires image data by scanning line by line, with each row of pixels being exposed sequentially. When the part to be inspected 2 is moving, the line scan camera 13 continuously captures images of different positions of the part to be inspected 2. If the frame rate and the moving speed are not matched, it may cause the image to be stretched or compressed, resulting in distortion. Therefore, in this embodiment, the frame rate of the line scan camera 13 is set to match the moving speed of the moving platform 11. The frame rate refers to the number of complete image frames generated by the line scan camera per second, measured in fps (frames per second).

[0047] For example, suppose the frame rate of the line scan camera is 4000 frames per second, the single frame resolution of the line scan camera is 0.1 mm / frame (the actual length corresponding to each frame scanned by the line scan camera), and the moving platform moves at a speed of 500 mm / s.

[0048] For a line scan camera, the time required to scan one frame is s, During the time it takes for the linear scan camera to scan one frame, the moving platform moves the part to be inspected. .

[0049] At this time, due to the distance traveled A frame resolution greater than 0.1mm / frame for a line scan camera results in gaps between adjacent frames, causing image artifacts such as gaps, blurring, and loss of detail. To avoid this, the movement speed can be set to match the frame rate; specifically, the movement speed can be set to equal the frame resolution. Frame rate.

[0050] In the example above, the movement speed of the mobile platform can be set to... In this way, during the scanning time of one frame by the linear array camera, the moving platform leads the part to be inspected to move 0.1mm, so there will be no gaps between adjacent frames, and the image data will be continuous and complete.

[0051] In an optional embodiment, the line light source 12 is described as follows:

[0052] In many application scenarios, since the objects to be detected are usually materials with highly reflective surfaces such as metals, these materials are prone to uneven lighting under ordinary light sources, which reduces image quality. Therefore, a line light source 12 that can shine parallel and directly is required.

[0053] In one specific implementation, when it is necessary to highlight surface defects such as pits and rust on the object to be inspected, see [reference needed]. Figure 2 The line light source 12 is preferably a coaxial line light source 121.

[0054] For the positional arrangement of the line light source 12 as a coaxial line light source 121, please refer to... Figure 2 A coaxial light source 121 is positioned between the line scan camera 13 and the moving platform 11. The coaxial light source 121 has a through-hole at its intersection with the imaging path of the line scan camera 13, ensuring that the coaxial light source 121 does not obstruct the imaging of the line scan camera 13. Specifically, the coaxial light source 121 includes a light-emitting component and a beam splitter. The light-emitting component emits linear light rays, which, after being split by the beam splitter, illuminate the part 2 to be inspected along a direction coaxial with the optical axis of the line scan camera 13.

[0055] In one specific embodiment, to further eliminate multi-angle reflection interference, see [link to relevant documentation]. Figure 3 To reduce noise in image data, the line light source 12 is preferably a dome line light source 123.

[0056] The dome-shaped line light source 123 is positioned directly above the moving platform 11. The dome-shaped line light source 123 generally includes a light-emitting array and a ring-shaped dome. The ring-shaped dome is covered with a diffuser material. The light-emitting array emits linear light rays, which then pass through the ring-shaped dome and illuminate the part 2 to be inspected. Because the ring-shaped dome converts direct light into diffused light, it eliminates local bright spots and constructs an omnidirectional uniform light field within a limited space, resulting in better imaging of the final image data.

[0057] In one specific embodiment, the linear light emitted by the line light source 12 can be blue light. This is because blue light has a shorter wavelength and higher energy, enabling it to penetrate the surface of the part 2 to be inspected, thus reducing glare. Furthermore, blue light can create excellent contrast on the surface of the object being inspected, making surface defects more visible in the image data.

[0058] In addition, when the linear light source is blue light, the surface defect detection device also includes a filter 122 to prevent damage to the eyes. The filter 122 has an opening, through which the blue light emitted by the linear light source 12 can pass and be projected onto the part 2 to be inspected. The filter 122 is generally made of polycarbonate (PC) or polyester (PET) with added blue light-absorbing dyes, coated glass with a metal oxide coating, interference filter film, etc., and the specific material can be selected according to the actual situation. The filter 122 is used to filter out blue light.

[0059] by Figure 2 For example, the opening can be located on the side close to the line scan camera 13, and the size of the opening matches the size of the coaxial light source 121. The linear light emitted by the coaxial light source 121 can pass through the opening and be projected onto the part 2 to be inspected, filtering out the influence of linear blue light on the human eye without affecting the camera's imaging.

[0060] by Figure 3 For example, the opening may include a first opening and a second opening. The first opening is located on the side close to the dome line light source 123, and the size of the first opening matches the size of the dome line light source 123 to ensure that the linear light emitted by the dome line light source 123 can pass through the first opening and be emitted onto the part to be inspected. The second opening is located on the side close to the line scan camera 13, and the size of the second opening matches the size of the imaging path of the line scan camera 13 to avoid affecting the imaging of the line scan camera 13.

[0061] In an optional embodiment, the processor 14 includes, but is not limited to, any of a central processing unit, a computer, a microcontroller, or a server. Those skilled in the art can use the processor 14 to implement any surface defect detection method, thereby achieving surface defect detection of the part to be inspected. The program corresponding to the surface defect detection method is not claimed in this disclosure.

[0062] The processor 14 is loaded with a recognition model for surface defect detection. Image data is input into the recognition model to obtain the surface defect detection result. Regarding the recognition model, this embodiment does not impose any particular limitation; any existing recognition model can be selected and loaded into the processor 14 according to the recognition requirements. The recognition model can be trained based on image data with defects or on normal image data. An example of a recognition model trained on normal image data is provided below. The recognition model provided in this embodiment is only provided as a possible implementation method for reference. This disclosure does not seek protection for the corresponding program of the recognition model.

[0063] The recognition model provided in this embodiment mainly consists of the following hierarchical structure: image cropping layer, feature extraction layer, and data partitioning layer.

[0064] The image cropping layer is used to identify the region where the part 2 to be inspected is located in the normal image data and crop it to obtain cropped image data. The preferred image cropping layer is YOLO-V8 (a target detection model). The image cropping layer can effectively reduce the image size, reduce background interference, and improve the success rate and speed of defect recognition.

[0065] The feature extraction layer is used to extract features from the cropped image data. These features include lines, textures, edges, etc. The preferred feature extraction layer is the first and second layers of a ResNet (residual network). Because only the first and second layers of the ResNet network are used, these layers have a smaller receptive field compared to deeper layers, and can only capture small, localized features in the image data. Specifically, the first layer can capture features such as edges and lines, while the second layer can capture features such as structural shapes. Besides ResNet, other convolutional networks can also be used.

[0066] The data partitioning layer is used to identify features of normal image data. The data partitioning layer preferably uses one-class SVM (an unsupervised learning algorithm). One-class SVM finds a hyperplane to separate defective features from defective features. This hyperplane can contain defective features to the greatest extent.

[0067] In this embodiment, the recognition model first identifies and crops out the area where the part to be detected 2 is located through the image cropping layer for the input image data. Then, it extracts the features from the cropped image data through the feature extraction layer. Next, it divides the data through the data partitioning layer to determine whether the extracted features are defective features. Generally speaking, defective features are located in the hyperplane of the data partitioning layer.

[0068] In one specific implementation, see Figure 2 and Figure 3The surface defect detection device may also include a display 17. The display 17 is electrically connected to the processor 14 and is used to display the received image data and / or surface defect detection results so that the operator can promptly determine the surface defect detection results of the part 2 to be inspected.

[0069] In an optional embodiment, see Figure 2 and Figure 3 The surface defect detection device also includes a first base 151 and a first bracket 152. The first bracket is disposed on the first base, and the line scan camera 13 is movably mounted on the first bracket. The position of the line scan camera 13 in the vertical direction can be adjusted by adjusting the first bracket 152, and the position of the line scan camera 13 in the horizontal direction can be adjusted by adjusting the first base 151.

[0070] In an optional embodiment, see Figure 2 The surface defect detection device also includes a second base 161 and a second bracket 162. The second bracket 162 is disposed on the second base 161, and the line light source 12 is movably mounted on the second bracket. The position of the line light source 12 in the vertical direction can be adjusted by adjusting the second bracket 162, and the position of the line scan camera 13 in the horizontal direction can be adjusted by adjusting the second base 161.

[0071] The following is combined Figure 2 The usage process of the surface defect detection device in this embodiment is described as follows: Parts produced are sampled according to a set ratio. The sampled parts to be inspected are placed on a turntable 112. The movable gripper 1111 on the turntable 112 is inserted into the inner hole of the part to be inspected 2. The movable gripper 1111 moves radially outward, forming an interference fit with the inner hole, thus fixing the part to be inspected 2 on the turntable 112. The operator can adjust the positions of the line scan camera 13 and the line light source 12 according to the position of the part to be inspected 2. Specifically, this can be achieved by adjusting the first support 152 and the first base 151 where the line scan camera 13 is located, and the second support 162 and the second base 161 where the line light source 12 is located. After fixing the part to be inspected 2 and adjusting the positions of the line scan camera 13 and the line light source 12, the moving speed of the turntable 112 is set according to the frame rate of the line scan camera 13. The turntable 112 rotates around the central axis of the part to be inspected 2 according to the set moving speed. At the same time, the line light source 12 starts to provide linear illumination, and the line scan camera 13 starts to scan the part to be inspected 2 line by line. When the line scan camera 13 finishes scanning, it obtains the image data of the part to be inspected 2. The processor 14 receives the image data and determines the surface defect detection result of the part to be inspected 2 based on the received image data.

[0072] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.

Claims

1. A surface defect detection device, characterized in that, The device includes a mobile platform, a line light source, a line scan camera, and a processor, wherein the processor is electrically connected to the line scan camera. The line light source and the moving platform are arranged on the same side relative to the line scan camera; The mobile platform carries the part to be tested, the mobile platform drives the part to be tested to move according to a preset moving method, and the line light source emits linear light to the part to be tested; The line scan camera scans the part to be inspected line by line to obtain image data of the part to be inspected, and the processor receives the image data and outputs the surface defect detection result of the part to be inspected accordingly.

2. The surface defect detection device as described in claim 1, characterized in that, The mobile platform is equipped with a clamp, which is used to fix the part to be tested.

3. The surface defect detection device as described in claim 2, characterized in that, The clamp includes movable jaws; The movable gripper passes through the inner hole of the part to be tested to fix the part to be tested, and the movable gripper forms an interference fit with the inner hole by moving radially outward.

4. The surface defect detection device as described in claim 3, characterized in that, Each of the said active grippers is covered with a silicone pad.

5. The surface defect detection device as described in claim 1, characterized in that, The line light source is a coaxial line light source; The coaxial light source is disposed between the line scan camera and the moving platform. A through hole is provided on the coaxial light source at the intersection with the imaging path of the line scan camera. The coaxial light source emits linear light rays in a direction coaxial with the optical axis of the line scan camera. And / or, the line light source is a dome line light source; The dome-shaped line light source is positioned directly above the moving platform, and diffusely reflects linear light rays onto the part to be inspected.

6. The surface defect detection device as described in claim 1, characterized in that, The linear ray is blue light.

7. The surface defect detection device as described in claim 6, characterized in that, The surface defect detection device also includes a filter cover with an opening, through which the blue light emitted by the line light source can pass and be projected onto the part to be inspected.

8. The surface defect detection device as described in claim 1, characterized in that, The surface defect detection device further includes a first base and a first support; The first bracket is disposed on the first base, and the line scan camera is movably mounted on the first bracket; And / or, the surface defect detection device further includes a second base and a second support; The second bracket is mounted on the second base, and the line light source is movably mounted on the second bracket.

9. The surface defect detection device as described in claim 1, characterized in that, The mobile platform includes a turntable on which the part to be tested is carried. The turntable drives the part to be tested to rotate at a constant speed along the central axis of the part to be tested. And / or, the frame rate of the line scan camera is matched with the moving speed of the mobile platform.

10. The surface defect detection device according to any one of claims 1-9, characterized in that, The surface defect detection device also includes a display; The display is electrically connected to the processor and is used to display the received image data and / or surface defect detection results.

Images