Device for irregular steel surface detection and operation method thereof

Abstract

The invention provides a device for irregular steel surface detection and an operation method thereof, belongs to the technical field of steel surface detection, and solves the problems of optical interference, poor complex surface adaptability, difficulty in balancing first-level detection precision and efficiency and incapability of meeting the requirements of an efficient production line in an existing steel detection mode. The detection device comprises two sliding rails, a moving structure and a 3D imaging system are both placed in a darkroom, the two sliding rails are arranged in parallel, the moving structure is perpendicular to the sliding rails, and the 3D imaging system is fixedly connected with the moving structure. The detection process is carried out in the darkroom, detection errors caused by the fact that the reflectivity of the surface of a high-reflectivity object to a light source is too high can be avoided, observation is carried out in an automatic mode, instability of manual observation is avoided, workpiece types and workpiece styles are not limited, and products of various models can be detected.

Description

Technical Field

[0001] The present invention relates to a device for detecting the surface of irregular steel and its operation method, belonging to the technical field of steel surface detection. Background Art

[0002] For steel with a high refractive index and an irregular surface, traditional detection systems face significant challenges in accuracy and stability. On the surface of materials with a high refractive index, especially the smooth or complex curved surfaces of steel, strong reflection and multiple refractions of light are easily induced. Traditional optical detection equipment relies on the linear propagation and stable reflection of light to obtain surface information. However, on the surface of high-refractive-index materials, light is strongly reflected, scattered, and even undergoes multiple refractions and offsets in complex surface structures, resulting in the distortion and attenuation of optical signals. As a result, it is difficult for the detection equipment to accurately capture the microscopic defects or subtle unevenness on the steel surface, and even in high-brightness reflection areas, effective visual information may be completely lost.

[0003] Currently, the methods for steel detection include laser detection, visual detection, and ultrasonic detection. Laser detection is widely used in high-precision surface measurement. However, due to the reflection and scattering of the laser beam by high-refractive-index materials, the reflected signal is weakened or distorted, affecting the accuracy and reliability of the measurement. Traditional visual detection systems rely on surface light reflection to obtain images. On high-refractive-index steel, due to the interference of reflected light on the image quality, it is difficult for the visual system to capture clear surface information. Especially in complex geometric regions, ultrasonic waves can penetrate the surface, but when detecting the steel surface, its resolution is relatively low. Especially when the detected surface has a complex geometric shape or a high refractive index, ultrasonic waves are easily affected by the surface structure, resulting in detection blind spots. The above three detection methods have the following problems: Optical interference: The surface of high-refractive-index materials seriously interferes with laser and visible light signals, resulting in the inability of traditional laser and visual detection systems to obtain accurate surface data, especially when detecting micro-defects, the effect is not good. Poor adaptability to complex surfaces: The irregular steel surface increases the complexity of the detection equipment, and existing systems are difficult to adapt to changing geometric shapes, resulting in the equipment being unable to detect in some areas or having blind spots. The balance problem between detection accuracy and efficiency: Existing high-precision detection methods often require longer processing times, reducing the detection efficiency and making it difficult to meet the requirements of high-speed production lines.

[0004] In addition, high-refractive-index materials also exacerbate the specular effect on the surface, making it difficult for optical sensors and sensors to distinguish between reflected light and actual defect areas, especially on irregular geometries or curved surfaces. This phenomenon often leads to the system being unable to accurately capture the true shape of the surface, especially in the detection of micro-cracks, pores, or roughness, increasing the misjudgment rate. Summary of the Invention

[0005] In order to solve the problems existing in the existing steel detection methods, such as optical interference, poor adaptability to complex surfaces, and difficulty in balancing detection accuracy and efficiency, which cannot meet the requirements of high-efficiency production lines, the present invention further proposes a device for detecting irregular steel surfaces and its operation method. The detection device includes:

[0006] Two slide rails, a moving structure, and a 3D imaging system are all placed in a darkroom. The two slide rails are arranged in parallel, the moving structure is arranged perpendicular to the slide rails, and the 3D imaging system is fixedly connected to the moving structure.

[0007] Preferably, sliders are arranged on the slide rails, positioning bolts are provided at both ends of the slide rails, and the moving structure is fixedly connected to the sliders.

[0008] Preferably, the 3D imaging system includes sensor 1, sensor 2, sensor 3, and sensor 4. One end of sensor 1 is connected to the moving mechanism, and the other end is connected to sensor 2. The angle between sensor 1 and the horizontal plane is 45°. Sensor 2 is connected to sensor 3, sensor 2 and sensor 3 are placed horizontally, the other end of sensor 3 is connected to sensor 4, the other end of sensor 4 is connected to the moving structure, and the angle between sensor 4 and the horizontal plane is 45°.

[0009] An operation method for a device for detecting irregular steel surfaces includes:

[0010] Step 1: Place a customized calibration fixture between the two slide rails. Drive the 3D imaging system to move within the slide rails through the moving structure to perform off-line calibration on the calibration fixture, and obtain the calibration parameters of the detection device. Among them, the size of the calibration fixture is fixed, and the surface is provided with protrusions of determined sizes.

[0011] Step 2: Place a high-reflectivity steel sample between the two slide rails. After confirming that the darkroom works normally, drive the 3D imaging system to move through the moving structure to collect data for automatic detection, output the 3D reconstruction effect diagram, and mark the defect positions in the reconstructed effect diagram.

[0012] Preferably, the steps of obtaining the calibration parameters of the detection device in Step 1 include:

[0013] Step 1.1: Reconstruct the protrusions of the calibration fixture based on a single sensor, and compare with the actual size of the protrusions to complete high-precision calibration, and output the detection data of the calibration fixture.

[0014] Step 1.2: Restore the detection data of all sensors to the same coordinate system through calibration. Among them, the calibration method uses a customized calibration plate, on which markers of fixed sizes are engraved, and the position distribution of the markers is known. Scan and reconstruct the entire calibration plate through the movement of the moving structure, and compare with the actual physical size to calculate and obtain the first calibration parameter.

[0015] Step 1.3: Calibrate the 3D imaging system and the moving structure. The moving structure adopts encoded movement control. The sensor is scanned under the traction of the moving structure. The size of the object after scanning and reconstruction, the actual movement distance, and the physical size of the detected object are calculated to obtain the second calibration parameter.

[0016] Step 1.4: Integrate the detection data, the first calibration parameter, and the second calibration parameter to obtain the calibration parameter of the detection device.

[0017] Preferably, the steps of outputting the 3D reconstruction effect diagram and marking the defect position in the reconstructed effect diagram in Step 2 include:

[0018] Step 2.1: After confirming that the darkroom works normally and the high-reflectivity steel sample is in place, the detection device automatically runs the detection process, the moving structure starts, and drives the 3D imaging system to start collecting data.

[0019] Step 2.2: The online scanning is triggered by the moving mechanism. The trigger of the moving mechanism is encoded according to the movement distance. Sensors 1, 2, 3, and 4 are triggered by the same signal source, and the data sampling intervals are the same.

[0020] Step 2.3: Perform online calculation on the collected data.

[0021] Step 2.4: Read the cad drawing and model the design size of the high-reflectivity steel sample.

[0022] Step 2.5: Perform an online comparison between the calculation result in Step 2.3 and the reading result in Step 2.4.

[0023] Step 2.6: Output the 3D reconstruction effect diagram according to the comparison result, and mark the defect position in the reconstructed effect diagram.

[0024] Preferably, the steps of online calculation in Step 2.3 include:

[0025] Step 2.3.1: Preprocess the point cloud data collected by the 3D imaging system to remove the noise in the point cloud data.

[0026] Step 2.3.2: Normalize the preprocessed point cloud data.

[0027] Step 2.3.3: Obtain the detection result of a single sensor according to the normalized point cloud data, and perform interpolation calculation on the detection result through sub-pixel processing.

[0028] Step 2.3.4: Perform fusion processing on the detection results of Sensor 1, Sensor 2, Sensor 3, and Sensor 4, and reconstruct the scanned spatial profile data in the same coordinate system;

[0029] Step 2.3.5: Perform spatial reconstruction based on each triggered scan result and the position of the encoder to obtain a three-dimensional model of the high-reflectivity steel sample.

[0030] The beneficial effects of the present invention are as follows:

[0031] 1. The present invention can avoid detection errors caused by excessive reflectivity of the light source on the surface of high-reflectivity objects;

[0032] 2. The present invention can perform observations in an automated manner to avoid the instability of manual observations;

[0033] 3. There are no restrictions on the types and styles of workpieces, and various models of products can be detected. Description of the Drawings

[0034] Figure 1 is a schematic structural diagram of a device for detecting the surface of irregular steel provided by the present invention;

[0035] Figure 2 is a schematic diagram of the arrangement of sensors of the 3D imaging system provided by the present invention;

[0036] Figure 3 is a flowchart of an operation method of a device for detecting the surface of irregular steel provided by the present invention. Detailed Embodiments

[0037] Detailed Embodiment 1: Combine Figure 1 and Figure 2 to illustrate this embodiment. As Figure 1 shown, the structure of a device for detecting the surface of irregular steel described in this embodiment includes: two slide rails, a moving structure, and a 3D imaging system; sliders are arranged on the slide rails, positioning bolts are provided at both ends of the slide rails, the moving structure is fixedly connected to the sliders, the detection area is a darkroom and is not affected by ambient light. Among them, the 3D imaging system is a line laser and is arranged in the manner of Figure 2 . During the detection process, the moving mechanism drives the 3D imaging system to scan the surface of the high-reflectivity steel and calculate based on the scan results.

[0038] As Figure 2As shown in the figure, the 3D imaging system includes sensor 1, sensor 2, sensor 3, and sensor 4. One end of sensor 1 is connected to the moving mechanism, and the other end is connected to sensor 2. The angle between sensor 1 and the horizontal plane is 45°. Sensor 2 is connected to sensor 3, and sensor 2 and sensor 3 are horizontally placed. The other end of sensor 3 is connected to sensor 4, and the other end of sensor 4 is connected to the moving structure. The angle between sensor 4 and the horizontal plane is 45°. In this embodiment, the scanning parameters of sensor 1, sensor 2, sensor 3, and sensor 4 include: the highest scanning frequency is 10 kHz, the resolution is 7 μm in the x-axis and 1 μm in the z-axis, the laser wavelength is 660 nm, and the number of scanning points is 2048.

[0039] The detection process of this embodiment is carried out in a dark room, which can avoid detection errors caused by too high reflectivity of the surface of high-reflectivity objects to the light source.

[0040] Specific Embodiment 2: Combined Figure 3 To illustrate this embodiment, the steps of the operation method of a device for detecting the surface of irregular steel are as follows:

[0041] a: Offline calibration

[0042] The detection system provided in this embodiment is a vision-based high-precision detection system. Therefore, the calibration of the 3D imaging system and the moving mechanism can ensure the stability and detection accuracy of the detection. The calibration work of the detection system is completed by the following multiple steps.

[0043] a1: Single sensor calibration

[0044] The calibration of a single sensor uses a customized calibration fixture. The size of the calibration fixture is fixed, and there are raised objects with determined sizes on the surface. The sensor will reconstruct the raised objects and compare them with the actual sizes of the raised objects to complete high-precision calibration.

[0045] a2: Multi-sensor calibration

[0046] Due to the need to ensure the detection accuracy of the system and the adaptability to workpieces of different sizes, in the present invention, data is collected simultaneously by multiple sensors. The detection data of multiple sensors need to be restored to the same coordinate system through calibration. The calibration uses a customized calibration board with markers of fixed sizes engraved on it, and the position distribution of the markers is known. The entire calibration board is scanned and reconstructed by moving the sliding table and compared with the actual physical sizes to obtain calibration parameters.

[0047] a3: Calibration of the imaging system and the moving mechanism

[0048] Due to certain involved errors and manual installation errors in the physical installation of the imaging system and the moving mechanism, it is necessary to calibrate the imaging system and the motion system. The moving mechanism adopts encoded motion control. The sensor is scanned under the traction of the moving mechanism, and the calibrated parameters are obtained by calculating the size of the object after scanning and reconstruction, the actual motion distance, and the physical size of the detected object.

[0049] a4: Offline calibration parameters

[0050] Integrate the calibration parameters in a1, a2, and a3 to obtain the calibration parameters of the entire system for subsequent calculations.

[0051] b: Online calculation

[0052] b1: Start detection

[0053] After the high-reflectivity steel sample is in place, confirm that the darkroom is working properly and there is no ambient light interference, the detection device automatically runs the detection process, the moving mechanism starts, and the 3D imaging system starts to collect data.

[0054] b2: Online scanning

[0055] Online scanning is triggered by the moving mechanism. The trigger of the moving mechanism is encoded according to the motion distance to ensure the uniformity and stability of the sampling interval. Multiple sensors are triggered by the same signal source.

[0056] b3: Online calculation

[0057] b3.1: Point cloud preprocessing

[0058] The point cloud data collected by the 3D imaging system has noise. Point cloud preprocessing will process this noise and perform operations such as meshing on the data.

[0059] b3.2: Data normalization

[0060] The sampling of the point cloud data varies according to the contour of the tool surface, so the scale range is also different. Therefore, normalizing the data can improve the detection accuracy.

[0061] b3.3: Sub-pixel processing

[0062] After data normalization, for inspection workpieces of different scales, the inspection data will have a certain step. Through sub-pixel processing, interpolation calculation is performed on the inspection results to further improve the accuracy of the inspection.

[0063] b3.4: Multi-sensor data fusion

[0064] According to the calculation results in a4, fuse the data collected by multiple 3D sensors and reconstruct the scanned spatial contour data in the same coordinate system.

[0065] b3.5: Spatial reconstruction

[0066] Based on the scanning results triggered each time and the position of the encoder, spatial reconstruction is carried out to obtain a three-dimensional model of the high-reflectivity steel sample.

[0067] b4: Reading the drawing

[0068] Read the cad drawing and model the design dimensions of the workpiece to be detected.

[0069] b5: Online comparison

[0070] Compare the results calculated in b3 with the results read in b4.

[0071] b6: Outputting the result

[0072] Output the effect diagram of the 3D reconstruction, and mark the defect positions in the reconstructed effect diagram.

[0073] In summary, this embodiment can perform observations in an automated manner, avoiding the instability of manual observations, and there are no restrictions on the types and styles of workpieces, and various models of products can be detected.

[0074] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the equivalent embodiments with equivalent changes by using the technical content disclosed above within the scope of the technical solution of the present invention. However, as long as it does not depart from the content of the technical solution of the present invention, according to the technical essence of the present invention, any simple modification, equivalent replacement and improvement of the above embodiments within the spirit and principle of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A device for detecting irregular steel surfaces, characterized in that: The structure of the device for irregular steel surface detection includes: two slide rails, a moving structure and a 3D imaging system; The two slide rails, the mobile structure and the 3D imaging system are all placed in a dark room, the two slide rails are arranged in parallel, the mobile structure is arranged perpendicular to the slide rails, and the 3D imaging system is fixedly connected to the mobile structure. The slide rail is composed of a toothed belt electric cylinder slide module, the 3D imaging system is fixed on the moving mechanism by mechanical fixings, and the moving mechanism is fixed on the moving slide of the toothed belt electric cylinder slide module by mechanical fixings.

2. The device for detecting irregular steel surfaces according to claim 1, characterized in that: A sliding block is arranged on the sliding rail, positioning bolts are arranged at both ends of the sliding rail, and the moving structure and the sliding block are fixedly connected.

3. The device for detecting irregular steel surfaces according to claim 1, characterized in that: The 3D imaging system includes sensor 1, sensor 2, sensor 3 and sensor 4. One end of sensor 1 is connected to the mobile mechanism, and the other end is connected to sensor 2. The angle between sensor 1 and the horizontal plane is 45°. Sensor 2 and sensor 3 are connected. Sensor 2 and sensor 3 are placed horizontally. The other end of sensor 3 is connected to sensor 4. The other end of sensor 4 is connected to the mobile structure. The angle between sensor 4 and the horizontal plane is 45°.

4. An operating method of a device for detecting irregular steel surfaces, applied to a device for detecting irregular steel surfaces as claimed in any one of claims 1 to 3, characterized in that: include: Step 1: Place the customized calibration jig between two slide rails, drive the 3D imaging system to move in the slide rails through the mobile structure, perform offline calibration on the calibration jig, and obtain the calibration parameters of the detection device, wherein the calibration jig has a fixed size and has protrusions of a certain size on its surface; Step 2: Place the high-reflectivity steel sample between the two slide rails. After confirming that the darkroom is working properly, use the mobile structure to drive the 3D imaging system to collect data, perform automatic detection, output the 3D reconstructed renderings, and mark the defect locations in the reconstructed renderings.

5. The operating method of the device for detecting irregular steel surface according to claim 4, characterized in that: The step of obtaining the calibration parameters of the detection device in step 1 includes: Step 1.1: Reconstruct the protrusions of the calibration fixture based on a single sensor and compare them with the actual size of the protrusions to complete high-precision calibration and output the detection data of the calibration fixture; Step 1.2: The detection data of all sensors are restored to the same coordinate system by calibration. The calibration method uses a customized calibration plate, on which markers of fixed size are engraved. At the same time, the position distribution of the markers is known. The entire calibration plate is scanned and reconstructed by moving the mobile structure, and compared and calculated with the actual physical size to obtain the first calibration parameter; Step 1.3: Calibrate the 3D imaging system and the mobile structure. The mobile structure adopts coded mobile control. The sensor scans under the traction of the mobile structure. The size of the object reconstructed by the scan and the actual movement distance and the physical size of the detected object are calculated to obtain the second calibration parameter; Step 1.4: Integrate the detection data, the first calibration parameter and the second calibration parameter to obtain the calibration parameter of the detection device.

6. The method for operating the device for detecting irregular steel surfaces according to claim 4, characterized in that: The steps of outputting the 3D reconstructed rendering in step 2 and marking the defect location in the reconstructed rendering include: Step 2.1: After confirming that the darkroom is working properly and the high-reflectivity steel sample is in place, the detection device automatically runs the detection process, the mobile structure starts, and the 3D imaging system starts to collect data; Step 2.2: The online scan is triggered by the moving mechanism, and the triggering of the moving mechanism is encoded according to the movement distance. Sensor 1, sensor 2, sensor 3 and sensor 4 are triggered by the same signal source, and the data sampling interval is consistent; Step 2.3: Perform online calculation on the collected data; Step 2.4: Read the CAD drawing and model the design dimensions of the high reflectivity steel sample; Step 2.5: Compare the calculation result in step 2.3 and the reading result in step 2.4 online; Step 2.6: Output a 3D reconstructed rendering according to the comparison result, and mark the defect location in the reconstructed rendering.

7. The method for operating the device for detecting irregular steel surfaces according to claim 6, characterized in that: The steps of online calculation in step 2.3 include: Step 2.3.1: Preprocess the point cloud data collected by the 3D imaging system to remove noise from the point cloud data; Step 2.3.2: Normalize the preprocessed point cloud data; Step 2.3.3: Obtain the detection results of a single sensor based on the normalized point cloud data, and perform interpolation calculation on the detection results through sub-pixel processing; Step 2.3.4: Perform fusion processing on the detection results of sensor 1, sensor 2, sensor 3 and sensor 4, and reconstruct the scanned spatial contour data into the same coordinate system; Step 2.3.5: Perform spatial reconstruction based on the scanning results of each trigger and the position of the encoder to obtain a three-dimensional model of the high-reflectivity steel sample.

Images