Laser cladding head light powder coaxial detection device and method based on image processing

Abstract

The invention relates to a laser cladding head light powder coaxial detection device and method based on image processing, and belongs to the technical field of laser additive manufacturing and remanufacturing. The device comprises a lifting support, a stand column, a manual gear, a rack, an industrial camera, a camera rotating disc, a spring positioning pin, an angle positioning hole and the like; height adjustment of the industrial camera is achieved through meshing of a manual gear and a rack, and rapid positioning of the angle of the camera is achieved through cooperation of a spring positioning pin and an angle positioning hole. The detection method comprises the steps of adjusting the position of the camera, shooting a powder feeding image, preprocessing the image, segmenting the image, extracting a powder flow center line and a light beam center line, and calculating the spatial deviation of a powder flow focus and the light beam center line, quantitative detection and accurate adjustment of the light powder coaxial state are achieved, and the repeatability and the forming quality of the cladding technology are improved.

Description

Technical Field

[0001] This invention relates to an image processing-based laser cladding head photo-powder coaxial detection device and method, belonging to the field of laser additive manufacturing and remanufacturing technology. It is used to realize the photo-powder coaxial status detection and precise adjustment of a three-channel coaxial powder feeding cladding head, ensuring cladding repeatability. Background Technology

[0002] In laser additive manufacturing and remanufacturing processes, the coaxiality of the photo-powder in a three-channel coaxial powder feeding cladding head directly affects the uniformity, forming accuracy, and process stability of the cladding layer. Currently, photo-powder coaxiality adjustment mainly relies on operator experience and is manually adjusted by visual observation. This method has the following shortcomings: a. Human visual perception is highly subjective and cannot accurately quantify the magnitude of the deviation; b. Low adjustment accuracy and poor consistency affect the repeatability of the cladding process; c. The industry lacks mature and efficient dedicated testing equipment and quantitative methods.

[0003] Therefore, developing a device and method that can achieve rapid and accurate detection and adjustment of the coaxial state of photosensitive powder is of great engineering significance. Summary of the Invention

[0004] This invention overcomes the shortcomings of existing technologies and provides a laser cladding head photosensitive powder coaxial detection device and method based on image processing. It realizes quantitative detection and precise adjustment of photosensitive powder coaxial status, and solves the problems of reliance on manual operation, inaccurate deviation quantification, and difficulty in ensuring coaxial accuracy in three-channel coaxial photosensitive powder feeding cladding head photosensitive powder coaxial detection.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: a coaxial detection device for laser cladding head photosensitive powder based on image processing, comprising a lifting bracket, a column, a manual gear, a rack, an industrial camera, a camera rotating disk, a spring positioning pin, and an angle positioning hole. The rack is vertically fixed on the column. The lifting bracket is a structure in which multiple horizontal legs are distributed around the outer circumference of a hollow disc. The ends of the legs of the lifting bracket are fixed with manual gears. The manual gears mesh with the corresponding racks. The lifting bracket moves up and down along the column through the meshing of the manual gears and the racks. The industrial camera is fixed on a camera rotating disk; the camera rotating disk is movably inserted into the disc of the lifting bracket, and a spring positioning pin is provided on the outer side of the disc of the lifting bracket. Multiple angle positioning holes are provided on the circumference of the camera rotating disk. By rotating the lifting bracket, the spring positioning pin can be matched and inserted into the angle positioning hole to position and fix the camera rotating disk. The camera angle can be manually positioned by the spring positioning pin and the angle positioning hole.

[0006] Furthermore, a matte black backlight panel is provided on the camera rotating disk opposite the camera rotating disk.

[0007] Furthermore, the bottoms of the multiple circularly arrayed columns are fixedly mounted on the base plate, and the horizontal legs of the lifting bracket correspond one-to-one with the columns.

[0008] Furthermore, a powder collection tray is provided on the base plate, and the powder collection tray is located directly below the lifting bracket.

[0009] Furthermore, the spring positioning pin adopts a ball spring structure, and the positioning holes are distributed at 90° intervals to achieve rapid positioning at four orthogonal angles.

[0010] The present invention provides a method for coaxial detection of photosensitive materials using the aforementioned device, comprising the following steps: S1. Manually rotate the manual gear to adjust the industrial camera to the viewing height, manually rotate the camera rotary disk and use the spring positioning pin and the angle positioning hole to position the industrial camera to the preset height; S2. An infrared beam is used to simulate a laser beam. In the powder feeding state, an industrial camera is used to capture images including the powder feeding nozzle, the beam, and the powder flow. S3. Preprocess the captured images to enhance the contrast between the powder flow and the background; S4. Using image segmentation technology, extract the powder feeding nozzle outline, beam outline, and powder flow region from the image respectively; S5. Based on the extracted powder flow region, identify and fit the centerline of the powder flow in each channel, and calculate their intersection point as the powder flow focus; S6. Identify the beam profile and determine its centerline; S7. Calculate the pixel distance between the powder flow focal point and the beam centerline, and calculate the actual spatial deviation between the two by using the pre-calibrated pixel and actual size ratio.

[0011] Furthermore, the image segmentation technique in step S4 employs a method of first separating and then fusing, specifically including the following steps: S41. Use a clustering algorithm to segment the original image into three regions: the powder nozzle, the beam, and the background. S42. Binarize the clustering results to generate the first binary image, in which the powder feeding nozzle and beam area are white and the background area is black; S43. Using the first binary image as a mask, remove the powder feeding nozzle and beam from the original image to obtain a sub-image containing only powder and background; S44. Perform threshold segmentation on the sub-image to generate a second binary image, where the powder area is white and the background area is black; S45. Fuse the first binary image and the second binary image to obtain the final binary image, in which the powder feeding nozzle, beam and powder area are all white and the background is black; S46. Perform edge detection on the final binary image to extract the powder feeding nozzle outline, beam outline, and powder flow region.

[0012] Further, the method for fitting the powder flow centerline in step S5 includes: for each powder flow channel, drawing its circumscribed rectangle based on the extracted powder flow region contour; using the centerline of the major axis of the circumscribed rectangle as the centerline of the powder flow in that channel; and then calculating the intersection point of the powder flow centerlines of each channel as the powder flow focus.

[0013] Furthermore, the pre-calibration process is as follows: a standard grid plate of known actual size is placed in the shooting field of view, a grid image is captured by the camera, the actual size represented by a pixel in the image is calculated, and the conversion relationship between pixel distance and actual distance is established.

[0014] Furthermore, after calculating the actual spatial deviation, an inspection report containing the deviation direction and value is generated. Based on this report, the operator is guided to make corresponding manual adjustments to the cladding head. After adjustment, the above inspection steps are repeated until the deviation is less than the allowable tolerance range.

[0015] The advantages of this invention compared to existing technologies are as follows: This invention achieves quantitative detection of the coaxial state of the photoparticles, improving adjustment accuracy and efficiency; it adopts a modular structure, facilitating installation, adjustment, and maintenance; it combines image processing and calibration technologies to achieve non-contact, high-precision detection; and it significantly improves the repeatability and forming quality of the laser cladding process. This invention adjusts the coaxiality of the photoparticles before cladding, which helps improve the repeatability of laser cladding forming and the coating morphology. Attached Figure Description

[0016] The present invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a schematic diagram of the detection device of the present invention (with cladding head).

[0018] Figure 2 This is a top view of the detection device of the present invention (with cladding head).

[0019] Figure 3 This is a schematic diagram of the lifting support structure in this invention.

[0020] Figure 4 This is a top view of the lifting support structure in this invention.

[0021] Figure 5 This is a schematic diagram showing the projected position of the three-channel coaxial powder feeding and cladding head in this invention.

[0022] Figure 6 This is a schematic diagram illustrating how an infrared beam is used instead of a laser beam to capture images of the powder being fed during the coaxial detection process of the photosensitive powder in this invention.

[0023] Figure 7 This is a schematic diagram of the powder delivery image processing flow acquired in this invention.

[0024] Figure 8 This is a schematic diagram of the coaxial detection results of photosensitive powder in this invention.

[0025] In the diagram: 1 is the lifting bracket, 2 is the column, 3 is the manual gear, 4 is the rack, 5 is the industrial camera, 6 is the camera rotary table, 7 is the spring positioning pin, 8 is the angle positioning hole, 9 is the matte black backlight panel, 10 is the base plate, and 11 is the powder collection tray. Detailed Implementation

[0026] The present invention will be further described below with reference to specific embodiments.

[0027] like Figures 1 to 4 As shown, the detection device of the present invention includes a lifting bracket 1, a column 2, a manual gear 3, a rack 4, an industrial camera 5, a camera rotating disk 6, a spring positioning pin 7, an angle positioning hole 8, a matte black backlight panel 9, and a powder collection tray 11. Multiple columns 2 arranged in a circumferential array are fixedly mounted on a base plate 10 at their bottoms. The horizontal legs of the lifting bracket 1 correspond one-to-one with the columns 2. The lifting bracket 1 moves up and down along the columns 2 via the engagement of the manual gear 3 and the rack 4, used to adjust the height of the industrial camera 5. The industrial camera 5 is fixed to the camera rotating disk 6, which is movably inserted into the disk of the lifting bracket 1 and can be manually rotated to adjust the angle of the industrial camera 5. The spring positioning pin 7 is located on the disk of the lifting bracket 1, and the angle positioning hole 8 is located on the camera rotating disk 6. During rotation, the spring positioning pin 7 pushes a ball into the angle positioning hole 8 to position the industrial camera 5. Continued rotation causes the spring positioning pin 7 to push the ball away from the angle positioning hole 8, thus adjusting the angle of the industrial camera 5. The spring positioning pin 7 adopts a ball spring structure, and the positioning holes 8 are distributed at 90° intervals to achieve rapid positioning at four orthogonal angles. The matte black backlight panel 9 is fixed on the camera rotating disk 6, opposite to the industrial camera 5, to enhance the contrast of the powder image. The powder collection tray 11 is located on the base plate 10, directly below the lifting bracket 1, to collect the powder sprayed during the detection process.

[0028] The coaxial photosensitive material testing method includes the following steps: First, manually rotate the manual gear 3 to adjust the industrial camera 5 to the viewing height, then manually rotate the camera rotary disk 6 and use the spring positioning pin 7 and the angle positioning hole 8 to position the industrial camera 5 at the preset height.

[0029] Next, an infrared beam is used to simulate a laser beam. During powder feeding, an industrial camera 5 captures images including the powder feeding nozzle, the beam, and the powder flow. Figure 6 As shown.

[0030] Then, the captured images are preprocessed to enhance the contrast between the powder flow and the background. The image processing workflow is as follows: Figure 7 As shown, the process includes using K-Means clustering to divide the image into powder delivery nozzle, beam, and background; binarization to separate the background and target regions; using the binary image as a mask to remove the powder delivery nozzle and beam while retaining powder information; using threshold segmentation to extract the powder region; and merging the powder, nozzle, and beam images to generate a binary image with a black background and a white target.

[0031] Image segmentation technology is used to extract the contours of the powder delivery nozzle, the beam, and the powder flow region from the image. This image segmentation technique is based on grayscale thresholding; by setting different grayscale threshold ranges, the powder delivery nozzle, beam, and powder flow are separated from the background.

[0032] Based on the extracted powder flow regions, the centerlines of the powder flows in each channel are identified and fitted, and their intersection points are calculated as the powder flow foci. The method for fitting the powder flow centerlines is as follows: for each powder flow channel, the edge point set of its contour is extracted, and then the least squares method is used to fit a straight line to this point set. The resulting straight line is the centerline of the powder flow in that channel.

[0033] Identify the beam profile and determine its centerline.

[0034] The pixel distance between the powder flow focal point and the beam centerline is calculated, and the actual spatial deviation between the two is calculated using a pre-calibrated pixel-to-actual-size ratio. The pre-calibration process involves placing a standard grid plate of known actual size in the shooting field of view, capturing a grid image with the camera, calculating the actual size represented by one pixel in the image, and establishing the conversion relationship between pixel distance and actual distance.

[0035] After calculating the actual spatial deviation, an inspection report is generated, including the deviation direction and value. This report guides the operator in manually adjusting the cladding head accordingly. The inspection process is repeated until the deviation is less than the allowable tolerance. The inspection results are as follows: Figure 8 As shown, the relative position of the powder flow focal point and the beam centerline can be displayed intuitively, guiding the coaxial adjustment of the photopowder.

[0036] In this invention, during device adjustment and image acquisition, the operator manually rotates the manual gear 3 based on experience or procedures, adjusting the industrial camera 5 to a suitable height by observing the height scale. Then, the camera rotary disk 6 is manually rotated. When the ball bearing of the spring positioning pin 7 engages with the angle positioning hole 8, a distinct "click" is felt, indicating the camera is positioned at a preset orthogonal angle. For minor angle adjustments, the fine-tuning screw can be operated. With the infrared light source and powder feeder activated, the industrial camera 5 captures a clear image of the powder flow against a matte black backlight panel 9.

[0037] In image processing and feature extraction, the original image obtained from preprocessing and image segmentation may contain noise, which can be addressed using conventional methods such as median filtering. Subsequently, image segmentation based on grayscale thresholding is performed. This is because the powder nozzle, infrared beam, and metal powder have different grayscale values ​​under backlighting conditions. By experimentally determining a suitable threshold range, the three elements can be separated from the black background, yielding their respective binary images.

[0038] In contour extraction and centerline fitting, the contours of the powder nozzle, beam, and each powder stream are extracted from the segmented binary image using a standard contour tracking algorithm. For each powder stream, a series of edge points on its contour are taken, and a line is fitted using the least squares method to find the line that best represents the distribution trend of these points, i.e., the powder stream centerline. The intersection of all powder stream centerlines is the powder stream focus. Similarly, the centerline of the beam region can be determined.

[0039] Calibration and deviation calculation are routine operations in the field of metrology. A calibration plate with a known-spaced grid is placed within the camera's field of view, and an image is captured. By calculating the number of pixels between grid points in the image, the actual size represented by each pixel can be determined. This scaling factor is used to convert pixel distances to actual distances.

[0040] Deviation calculation: In the image coordinate system, measure the vertical distance (number of pixels) from the focal point of the powder flow to the center line of the beam, and multiply it by the calibrated scaling factor to obtain the actual spatial deviation value between the two.

[0041] Feedback and Adjustment: The system generates an intuitive inspection report, indicating the direction and specific value of the deviation. Based on this report, operators manually adjust the relevant mechanical components of the cladding head. After adjustment, inspection is performed again, forming a closed loop of "inspection-adjustment-re-inspection" until the deviation meets the process requirements.

[0042] Example 1: Device Adjustment Process The operator first loosens the locking mechanism, manually rotates the manual gear 3 handle, and observes the height scale to adjust the industrial camera 5 to the predetermined position. Then, rotate the camera rotary table 6; when the spring positioning pin 7 is engaged, the preset angle has been reached. For fine adjustments, the adjusting screw on the side of the industrial camera 5 can be rotated to achieve angle optimization with a precision of 0.5°. Finally, tighten all mechanisms to ensure stability during the inspection process.

[0043] Example 2: Implementation of the detection method After image acquisition, the system automatically performs the following processing: First, Gaussian filtering is applied for noise reduction, followed by region segmentation using an improved K-Means++ algorithm. During feature extraction, sub-pixel-level edge information is obtained through multi-scale analysis. The calibration process uses bilinear interpolation to convert pixel coordinates to world coordinates. Finally, a detection report containing a 3D deviation vector is generated to guide operators in making precise adjustments.

[0044] Example 3: Accuracy Verification The accuracy of the device was verified through repeatability testing. The standard deviation of the positional deviation was less than 0.05 mm and the standard deviation of the angle deviation was less than 0.3° after 10 repeated measurements, which meets the requirements for use in industrial settings.

[0045] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A laser cladding head photosensitive powder coaxial detection device based on image processing, characterized in that, The system includes a lifting bracket (1), a column (2), a manual gear (3), a rack (4), an industrial camera (5), a camera rotating disk (6), a spring positioning pin (7), and an angle positioning hole (8). The rack (4) is vertically fixed on the column (2). The lifting bracket (1) is a structure with multiple horizontal legs distributed around the outer circumference of a hollow disc. The ends of the legs of the lifting bracket (1) are fixed with a manual gear (3). The manual gear (3) meshes with the corresponding rack (4). The lifting bracket (1) moves up and down along the column (2) through the meshing of the manual gear (3) and the rack (4). The industrial camera (5) is fixed on the camera rotating disk (6); the camera rotating disk (6) is movably inserted into the disc of the lifting bracket (1), and a spring positioning pin (7) is provided on the outer side of the disc of the lifting bracket (1). Multiple angle positioning holes (8) are provided on the circumference of the rotating disk (6). By rotating the lifting bracket (1), the spring positioning pin (7) can be matched and movably inserted into the angle positioning hole (8) to position and fix the camera rotating disk (6). The camera angle is manually positioned by the spring positioning pin (7) and the angle positioning hole (8).

2. The laser cladding head photosensitive powder coaxial detection device based on image processing according to claim 1, characterized in that, A matte black backlight panel (9) is provided on the camera rotating disk (6) opposite to the camera rotating disk (6).

3. The laser cladding head photosensitive powder coaxial detection device based on image processing according to claim 1, characterized in that, Multiple columns (2) arranged in a circular array are fixedly mounted on the base plate (10) at the bottom, and the horizontal legs of the lifting bracket (1) correspond one-to-one with the columns (2).

4. The laser cladding head photosensitive powder coaxial detection device based on image processing according to claim 3, characterized in that, A powder collection tray (11) is provided on the base plate (10), and the powder collection tray (11) is located directly below the lifting bracket (1).

5. The laser cladding head photosensitive powder coaxial detection device based on image processing according to claim 1, characterized in that, The spring positioning pin (7) adopts a ball spring structure, and the positioning holes (8) are distributed at 90° intervals to achieve rapid positioning at four orthogonal angles.

6. A method for coaxial detection of photosensitive materials using the apparatus described in any one of claims 1-5, characterized in that, Includes the following steps: S1. Manually rotate the manual gear (3) to adjust the industrial camera (5) to the observation height, manually rotate the camera rotary disk (6) and use the spring positioning pin (7) and the angle positioning hole (8) to position the industrial camera (5) to the preset height; S2. Using an infrared beam to simulate a laser beam, in the powder feeding state, an industrial camera (5) is used to capture an image containing the powder feeding nozzle, the beam and the powder flow; S3. Preprocess the captured images to enhance the contrast between the powder flow and the background; S4. Using image segmentation technology, extract the powder feeding nozzle outline, beam outline, and powder flow region from the image respectively; S5. Based on the extracted powder flow region, identify and fit the centerline of the powder flow in each channel, and calculate their intersection point as the powder flow focus; S6. Identify the beam profile and determine its centerline; S7. Calculate the pixel distance between the powder flow focal point and the beam centerline, and calculate the actual spatial deviation between the two by using the pre-calibrated pixel and actual size ratio.

7. The method for coaxial detection of photosensitive materials according to claim 6, characterized in that, The image segmentation technique in step S4 employs a method of first separating and then fusing, specifically including the following steps: S41. Use a clustering algorithm to segment the original image into three regions: the powder nozzle, the beam, and the background. S42. Binarize the clustering results to generate the first binary image, in which the powder feeding nozzle and beam area are white and the background area is black; S43. Using the first binary image as a mask, remove the powder feeding nozzle and beam from the original image to obtain a sub-image containing only powder and background; S44. Perform threshold segmentation on the sub-image to generate a second binary image, where the powder area is white and the background area is black; S45. Fuse the first binary image and the second binary image to obtain the final binary image, in which the powder feeding nozzle, beam and powder area are all white and the background is black; S46. Perform edge detection on the final binary image to extract the powder feeding nozzle outline, beam outline, and powder flow region.

8. The method for coaxial detection of photosensitive materials according to claim 6, characterized in that, The method for fitting the powder flow centerline in step S5 includes: for each powder flow channel, drawing its circumscribed rectangle based on the extracted powder flow region contour; using the centerline of the major axis of the circumscribed rectangle as the centerline of the powder flow in that channel; and then calculating the intersection point of the powder flow centerlines of each channel as the powder flow focus.

9. The method for coaxial detection of photosensitive materials according to claim 6, characterized in that, The pre-calibration process is as follows: a standard grid plate of known actual size is placed in the shooting field of view, the grid image is captured by the camera, the actual size represented by a pixel in the image is calculated, and the conversion relationship between pixel distance and actual distance is established.

10. The method for coaxial detection of photosensitive materials according to claim 6, characterized in that, After calculating the actual spatial deviation, an inspection report containing the deviation direction and value is generated. Based on this report, the operator is guided to make corresponding manual adjustments to the cladding head. After adjustment, the above inspection steps are repeated until the deviation is less than the allowable tolerance range.

Images