Laser directional energy deposition coaxial powder feeding light powder concentric calibration method and system

Abstract

The invention discloses a laser directional energy deposition coaxial powder feeding light powder concentricity calibration method, which belongs to the technical field of laser cladding, and comprises the following steps: S1, ensuring that an imaging optical axis is parallel to or coincides with a theoretical central axis of a cladding head; s2, the base plate is placed below a cladding head; s3, collecting a powder center image; s4, performing image processing and feature extraction to obtain a powder center; s5, collecting a laser spot image, starting a laser, and capturing the spot image of the laser on a focal plane by a camera; s6, performing image processing and feature extraction to obtain a light center; step S7, carrying out coaxiality calculation and calibration; and S8, repeating the steps S3 to S7 until the offset is smaller than a set threshold value, namely considering that calibration is completed, according to the method, a coaxial imaging method is utilized, the powder disc image and the laser spot image are obtained, the offset information of the powder disc center and the laser spot center is obtained through machine vision algorithm processing, calibration operation is guided, and the method is simple and easy to implement, low in cost and good in practicability.

Description

Technical Field

[0001] This invention relates to the field of laser cladding technology, and in particular to a method and system for calibrating coaxial powder feeding and photoconcentricity of laser directional energy deposition. Background Technology

[0002] Directed Energy Deposition (DED) is an advanced additive manufacturing (3D printing) technology specifically designed for manufacturing or repairing high-value metal parts. During DED, multiple powder beams need to converge uniformly and stably onto the cladding work plane, forming a powder spot with a Gaussian normal distribution. A high-energy heat source (usually a laser) is precisely focused onto a small area (the molten pool), forming a spot. Both the laser and the powder beams ultimately reach the surface of the metal substrate through the cladding head. The concentricity of the laser and the powder spot has a significant impact on the cladding quality. If the laser and powder are not concentric, it can lead to reduced material utilization and powder contamination, loss of geometric accuracy, deterioration of internal part quality, process instability, and spattering.

[0003] In existing technologies, the main method has been ablation, which involves using a laser to create a dot on the substrate without feeding powder. Then, powder is fed without activating the laser, allowing the powder to accumulate on the substrate for a short period. The center of the accumulated powder is then observed to ensure it coincides with the center of the laser ablation point. The ablation method has the following drawbacks:

[0004] Highly subjective: Calibration personnel need to visually inspect whether the center of the powder accumulation point coincides with the center of the laser ablation point. Different people may have slight differences in their standards for judging the "center," which introduces human error.

[0005] Low resolution: The human eye has limited ability to resolve offsets at the sub-millimeter level, especially tens of micrometers. High-precision DED machining often requires even higher alignment accuracy. Even when seemingly "aligned," there may still be minute deviations affecting the process.

[0006] Unquantifiable: This method can usually only give an "yes or no" alignment judgment, and cannot accurately measure the magnitude and direction of the offset, which is not conducive to data recording and process standardization.

[0007] A Chinese patent application (application number 202510319187.X) discloses a method for calibrating the coaxiality of the powder feeding head and laser in laser cladding. Utilizing an established mapping relationship, based on the offset angle θ and direction φp of the printed annular contour, the offset distance l and direction φl of the laser system 11 can be determined. This allows for the reverse adjustment of the adjustment thread 13 of the laser system 11 to compensate for coaxiality deviations. While this method can quickly and accurately calibrate deviations when the powder coaxiality is inconsistent, and can be applied to calibrating powder coaxiality for different cladding heads by adjusting the laser beam position, it has the following drawback:

[0008] High cost: This method relies on high-precision measuring equipment (such as laser profilometers). The equipment itself is expensive, resulting in high initial investment costs.

[0009] High requirements for technical personnel: Operators need to have knowledge of laser cladding process, precision measurement skills and certain data analysis ability (such as understanding regression models), which is a high technical threshold.

[0010] Therefore, to address the above issues, it is necessary for this invention to provide a laser-guided energy deposition coaxial powder feeding and photoconcentric calibration method and system based on the principle of machine vision. Summary of the Invention

[0011] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method and system for calibrating the concentricity of photoconductor in a coaxial photoconductor feeding system for laser directional energy deposition (EDD). This method calibrates the concentricity of photoconductor in a laser directional energy deposition (EDD) device in a more accurate and cost-effective manner.

[0012] To achieve the above objectives, the present invention provides the following technical solution:

[0013] This invention discloses a method for coaxial powder feeding and concentric calibration of laser-directed energy deposition (EDD), which includes the following steps:

[0014] Step S1: Ensure that the imaging optical axis is parallel to or coincides with the theoretical central axis of the cladding head;

[0015] Step S2: Place the substrate under the cladding head;

[0016] Step S3: Acquire an image of the powder center;

[0017] Step S4: Image processing and feature extraction to obtain the powder center;

[0018] Step S5: Acquire laser spot image, turn on the laser, and have the camera capture the laser spot image on the focal plane;

[0019] Step S6: Image processing and feature extraction to obtain the light center;

[0020] Step S7: Coaxiality calculation and calibration;

[0021] Step S8: Repeat steps S3 to S7 until the offset is less than the set threshold, at which point the calibration is considered complete.

[0022] Furthermore, in step S3, the center of the powder tray is the powder center, and no powder needs to be dispensed during the entire calibration process.

[0023] Furthermore, step S4 includes:

[0024] Step S41: Image edge contour extraction. Perform edge contour extraction on the input image to be processed to obtain an initial set of edge contours.

[0025] Step S42: Contour length filtering, based on the physical length of the contour, the initial set of edge contours is filtered;

[0026] Step S43: Contour merging based on cocircularity constraints;

[0027] Step S44: Algebraic fitting of circular parameters. The merged contour obtained in step S43 is fitted with a circular equation using an algebraic fitting algorithm.

[0028] Furthermore, in step S4.3, the contours in the candidate contour set are intelligently merged based on their geometric characteristics. The merging criterion is based on the following cocircularity geometric constraint:

[0029] Curvature consistency: The contour segments to be merged should have similar radii of curvature;

[0030] Circle center proximity: The instantaneous circle centers fitted by each contour segment should converge within a narrow spatial range;

[0031] Contour endpoint proximity: Contour segments belonging to the same circle should have their breakpoints close to each other in space.

[0032] Furthermore, in step S4.4, the geometric center coordinates of the laser spot are C_powder(X_p,Y_p);

[0033] Let the equation of the circle be:

[0034] (xa) 2 +(yb)2=r 2

[0035] Where (a,b) is the center of the circle and r is the radius.

[0036] Furthermore, step S6 includes:

[0037] Step S61: Adaptive extraction of region of interest for laser spot. For the acquired substrate surface image, the high contrast between the laser spot and the background is used to perform preliminary segmentation through a low global threshold to obtain the connected region containing the laser spot. Based on the bounding rectangle of the connected region, a sub-image is dynamically extracted as the region of interest for the laser spot.

[0038] Step S62: Gaussian filtering for noise reduction of the spot image. Gaussian low-pass filtering is applied to the region of interest image of the spot to obtain a smoothed spot image.

[0039] Step S63: Calculate the weighted centroid based on the gray values ​​of the entire image. Use the gray value of each pixel in the smoothed spot image as its weight to calculate the gray-weighted centroid of the entire spot region of interest.

[0040] 7. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 6, characterized in that:

[0041] In step 6.3, the coordinates of the centroid are the desired coordinates of the light spot center C_laser(X_l,Y_l);

[0042]

[0043] in:

[0044] W and H are the width and height of the region of interest of the light spot, respectively;

[0045] I(x,y) is the gray value of the pixel with coordinates (x,y) after Gaussian filtering;

[0046] X_l and Y_l are the calculated coordinates of the light spot center.

[0047] Furthermore, step S7 includes:

[0048] Step S71: Calculate the offset vector:

[0049] Offset ΔX = X_p - X_l

[0050] Offset ΔY = Y_p - Y_l

[0051] Total offset Offset = sqrt(ΔX) 2 +ΔY 2 )

[0052] The offset direction θ = atan2(ΔY,ΔX);

[0053] Step S72: Perform calibration. Based on the calculated (ΔX, ΔY), make a fine adjustment in the opposite direction by adjusting the pitch screw on the laser module or powder feeding nozzle inside the cladding head.

[0054] The calibration system for implementing the above-described laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method includes:

[0055] A coaxial CCD camera, blue light illumination, and a data processing system. The data processing system is connected to the coaxial CCD camera and includes an image processing unit, a coaxiality calculation unit, and an output information unit.

[0056] A coaxial CCD is used to acquire images of the powder center and the laser spot.

[0057] Blue light illumination is used to provide lighting conditions for the substrate surface;

[0058] The image processing unit is used for image processing and feature extraction;

[0059] The coaxiality calculation unit is used to calculate the lateral offset ΔX, longitudinal offset ΔY, total offset Offset, and offset direction θ;

[0060] The output information unit is used to display the lateral offset ΔX and the longitudinal offset ΔY, as well as information on whether the calibration is complete.

[0061] In the above technical solution, the present invention provides a method and system for calibrating the concentricity of photopowder in laser directional energy deposition (DED) equipment by coaxial powder feeding. Compared with the ablation method, pixel-level measurement based on machine vision is more accurate than that by human eye. Compared with existing methods for calibrating the coaxiality of the powder feeding head and laser in laser cladding, CCD industrial cameras are cheaper than laser profilometers, the entire measurement process is simpler, and the requirements for technical personnel are lower.

[0062] Secondly, using coaxial imaging, images of the powder tray and the laser spot are acquired. After processing by machine vision algorithms, the offset information of the powder tray center and the laser spot center is obtained to guide the calibration operation. This method is simple, easy to implement, and low in cost, and has good practicality. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0064] Figure 1 This is a schematic diagram of the cladding head disclosed in this invention;

[0065] Figure 2 This invention discloses a coaxial CCD for capturing powder disc images.

[0066] Figure 3 This invention discloses a coaxial CCD image of a laser ablation zone.

[0067] Figure 4 This is a schematic diagram of the structure of a laser-directed energy deposition coaxial powder delivery and photoconcentric calibration method disclosed in this invention;

[0068] Figure 5 This is a schematic diagram of the laser-directed energy deposition coaxial powder feeding and photoconcentric calibration system disclosed in this invention.

[0069] Explanation of reference numerals in the attached figures:

[0070] 1. Calibration substrate; 2. Laser powder feeding and cladding head; 3. Blue light illumination; 4. Blue light illumination controller; 5. Coaxial CCD camera; 6. Laser body; 7. Data processing system; 8. Display. Detailed Implementation

[0071] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0072] See Figure 4 As shown;

[0073] A method for coaxial powder feeding and concentric calibration of laser-directed energy deposition is invented, comprising the following steps:

[0074] Step S1: Ensure that the imaging optical axis of the device is approximately parallel to or coincides with the theoretical central axis of the cladding head;

[0075] Step S2: Place the special substrate under the cladding head;

[0076] Step S3: Acquire an image of the powder center. Due to the structure of the cladding head, we assume that the center of the powder tray is the powder center, so no powder needs to be dispensed during the entire calibration process.

[0077] Step S4: Image processing and feature extraction to obtain the powder center;

[0078] Step S5: Acquire an image of the laser spot. Turn on the laser (usually at low power), but do not feed toner. The camera captures an image of the laser spot on the focal plane.

[0079] Step S6: Image processing and feature extraction to obtain the light center;

[0080] Step S7: Coaxiality calculation and calibration;

[0081] Step S8: Repeat steps S3 to S7 until the offset is less than the set threshold, at which point the calibration is considered complete.

[0082] To better implement the present invention, preferably, step S4 includes the following steps:

[0083] Step S41: Image edge contour extraction:

[0084] The input image to be processed is subjected to edge contour extraction to obtain an initial set of edge contours. Specifically, this step converts grayscale transition regions in the image into contour vector data, providing a basis for subsequent geometric analysis. The contour extraction method can be the Canny operator, the Sobel operator, or any other algorithm that can output XLD contours.

[0085] Step S42: Contour length filtering:

[0086] Based on the physical length of the contours, the initial set of edge contours is filtered. Specifically, contours that are too short (potentially noise) or too long (potentially non-target interference contours) are discarded, and contour segments with lengths within a preset effective range are retained to form a candidate contour set. This step effectively filters out noise and interference from irrelevant large-sized structures by setting length thresholds L_min (e.g., 500 pixels) and L_max (e.g., 10000 pixels), focusing on contour segments that may belong to the target circle.

[0087] Step S43: Contour merging based on cocircularity constraints:

[0088] For the candidate contour set, contours are intelligently merged based on their geometric properties. It determines whether multiple discrete contour segments might originate from the same actual circle; if so, they are merged into a single complete contour object. The merging criterion is based on the following cocircularity geometric constraint:

[0089] a) Curvature consistency: The contour segments to be merged should have similar radii of curvature;

[0090] b) Proximity of center: The instantaneous centers of the circles fitted by each contour segment should converge within a narrow spatial range;

[0091] c) Proximity of contour endpoints: Contour segments belonging to the same circle should have their breakpoints close to each other in space.

[0092] In a preferred embodiment, this is achieved by calling the UnionCocircularContoursXld operator, which encapsulates the judgment logic for the aforementioned geometric constraints.

[0093] Step S44: Algebraic fitting of circular parameters:

[0094] The merged contour obtained in step S43 is fitted with a circular equation using an algebraic fitting algorithm, and the geometric center coordinates of the laser spot are C_powder(X_p,Y_p).

[0095] The goal of this algebraic fitting method is to find a circle that minimizes the sum of the squares of the algebraic distances from all points on the contour to the circle.

[0096] Let the equation of the circle be:

[0097] (xa) 2 +(yb)2=r 2

[0098] Where (a, b) is the center of the circle and r is the radius. The fitting process involves using mathematical optimization methods such as least squares to find the optimal solutions for parameters a, b, and r, which best represent the geometric distribution of the contour points. The fitting method is robust to noise and local outliers.

[0099] To better implement the present invention, preferably, step S6 includes the following steps:

[0100] Step S61: Adaptive extraction of the region of interest for the light spot:

[0101] For the acquired substrate surface image, the high contrast between the laser spot and the background is utilized to perform preliminary segmentation using a low global threshold to obtain the connected region containing the laser spot. Based on the bounding rectangle of this connected region, a sub-image is dynamically extracted as the region of interest for the laser spot.

[0102] This step significantly reduces the amount of data required for subsequent processing, improving efficiency. On the other hand, it isolates potential noise interference from other areas of the image, providing a clean data source for subsequent accurate calculations.

[0103] Step S62: Gaussian filtering for noise reduction of the light spot image:

[0104] Gaussian low-pass filtering is applied to the region of interest image of the light spot to obtain a smoothed light spot image.

[0105] Specifically, the calibration method of this invention does not aim to preserve sharp edges, because edges are where burrs are located. The core purpose of Gaussian filtering is to smooth out high-frequency burr noise while preserving the overall grayscale distribution of the light spot. The size of the filter kernel can be adaptively set according to the physical size of the light spot and the image resolution;

[0106] Step S63: Calculate the weighted centroid based on the grayscale values ​​of the entire image:

[0107] The gray value of each pixel in the smoothed spot image is used as its weight to calculate the gray-weighted centroid of the entire spot region of interest. The coordinates of this centroid are the desired spot center coordinates C_laser(X_l,Y_l).

[0108]

[0109] in:

[0110] W and H are the width and height of the region of interest of the light spot, respectively.

[0111] I(x,y) is the gray value of the pixel at coordinates (x,y) after Gaussian filtering.

[0112] X_l and Y_l are the calculated coordinates of the light spot center.

[0113] To better implement the present invention, preferably, step S7 includes the following steps:

[0114] Step S71: Calculate the offset vector:

[0115] Offset ΔX = X_p - X_l

[0116] Offset ΔY = Y_p - Y_l

[0117] Total offset Offset = sqrt(ΔX) 2 +ΔY 2 )

[0118] The offset direction θ = atan2(ΔY,ΔX);

[0119] Step S72: Perform calibration. Based on the calculated (ΔX, ΔY), make a fine adjustment in the opposite direction by adjusting the pitch screw on the laser module or powder feeding nozzle inside the cladding head.

[0120] The invention relates to a calibration system for a laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method, comprising: a coaxial CCD camera, blue light illumination, and a data processing system. The data processing system is connected to the coaxial CCD camera and includes an image processing unit, a coaxiality calculation unit, and an output information unit.

[0121] A coaxial CCD is used to acquire images of the powder center and the laser spot.

[0122] Blue light illumination is used to provide illumination conditions for the substrate surface;

[0123] An image processing unit is used for image processing and feature extraction.

[0124] The coaxiality calculation unit is used to calculate the lateral offset ΔX, longitudinal offset ΔY, total offset Offset, and offset direction θ.

[0125] The output information unit is used to display the lateral offset ΔX and the longitudinal offset ΔY, as well as information on whether the calibration is complete.

[0126] Example 1:

[0127] The coaxial powder feeding and concentric calibration method for laser-directed energy deposition involves fitting the edge of the powder tray into a circle under non-light and non-powder conditions to obtain the geometric center coordinates C_powder(X_p,Y_p); during brief light output, an image processing algorithm is used to measure the ablation area to obtain the geometric center coordinates C_laser(X_l,Y_l) of the laser spot, and the offsets ΔX and ΔY are calculated. Fine adjustments are then made in the opposite direction by adjusting the adjustment screw on the cladding head's feeding nozzle.

[0128] Step S1: Using the coaxial powder feeding and cladding head, place the calibration substrate 1 directly beneath it (e.g., Figure 1 (as shown);

[0129] Step S2: Laser external multi-channel toner feeding process. To obtain a clear toner tray image, it is necessary to first turn on the blue light illumination controller 4 and adjust the blue light source 3 to make the contrast between the toner tray edge and the inner and outer parts of the image obvious (e.g., ...). Figure 2 (as shown);

[0130] Step S3: Use the coaxial CCD camera 5 to view the current image (e.g., Figure 3 The data system 7 uses image processing technology to fit a circle to the edge of the powder tray, and then obtains the geometric center coordinates of the circle, C_powder(X_p,Y_p).

[0131] Step S4: The laser body 6 emits a laser beam. When the laser beam is stable, the coaxial CCD camera 5 captures an image.

[0132] Step S5: Data system 7 uses the image processing algorithm grayscale centroid method to accurately calculate the geometric center coordinates C_laser(X_l,Y_l) of the laser spot.

[0133] Step S6: Data system 7 calculates the offset vector:

[0134] Offset ΔX = X_p - X_l

[0135] Offset ΔY = Y_p - Y_l

[0136] Total offset Offset = sqrt(ΔX) 2 +ΔY 2 )

[0137] Offset direction θ = atan2(ΔY, ΔX)

[0138] Step S7: Compare ΔOffset with the threshold 4pixel. If it is greater than ΔOffset, fine-tune it in the opposite direction by adjusting the pitch screw on the powder feeding nozzle of the cladding head. If it is less than ΔOffset, output a calibration success message to the display 8.

[0139] Example 2:

[0140] A method for coaxial powder feeding and concentric calibration of laser-directed energy deposition (EDD) includes the following steps:

[0141] Step S1: The coaxial CCD camera 5 acquires an image of the powder tray, and the data processing system 7 processes the image to obtain the coordinates of the powder center C_powder(X_p,Y_p)=(610,330).

[0142] Step S2: Turn on the laser and use a coaxial CCD camera to capture the ablation image. The data processing system 7 processes the image to obtain the geometric center of the laser spot C_laser(X_l,Y_l)=(640,360).

[0143] Step S3: Calculate the offset vector:

[0144] ΔX = 610 - 640 = -30 pixels

[0145] ΔY = 330 - 360 = -30 pixels

[0146] Offset = sqrt((-30)) 2 +(-30) 2 = sqrt(1800) ≈ 42.4 pixels

[0147] θ = atan2(-30,-30) = atan2(1) ≈ -135° or 225° (rotated 225 degrees counterclockwise from the positive X-axis, or in the lower left quadrant).

[0148] Step S4: Perform calibration by fine-tuning in the opposite direction by adjusting the pitch screw on the powder feeding nozzle of the cladding head.

[0149] Step S5: Measurement data, as shown in Table 1:

[0150] C_powder(X_p,Y_p) (625,350) C_laser(X_l,Y_l) (640,360) ΔX -15 pixels ΔY -10 pixels Offset 18.03 pixels θ 213.69°

[0151] Step S6: Compare Offset > threshold 4 pixels, then repeat step 4.

[0152] Step S7: Measurement data, as shown in Table 2:

[0153] C_powder(X_p,Y_p) (638,357) C_laser(X_l,Y_l) (640,360) ΔX -2 pixels ΔY -3 pixels Offset 3.61 pixels θ 236.31°

[0154] Step S8: If Offset < threshold 4 pixels, display 8 outputs a calibration success message.

[0155] like Figure 5 As shown, Example 3:

[0156] A laser-directed energy deposition coaxial powder feeding and concentric calibration system includes: a calibration substrate 1, a laser powder feeding and cladding head 2, a blue light illumination 3, a blue light illumination controller 4, a coaxial CCD camera 5, a laser body 6, a data processing system 7, and a display 8. The blue light illumination 3 is mainly used to illuminate the calibration substrate 1, the coaxial CCD camera 5 mainly acquires images of the powder tray and the laser spot, and the cladding head 4 is the channel between the laser and the powder.

[0157] The data processing system 7 includes an image processing unit, a coaxiality calculation unit, and an output information unit;

[0158] The image processing unit is used to process images captured by the coaxial CCD camera 5, and to obtain the center of the powder disc C_powder(X_p,Y_p) and the center of the laser spot C_laser(X_l,Y_l).

[0159] The coaxiality calculation units are ΔX, ΔY, Offset, and θ. When Offset > the threshold of 4 pixels, further calibration is required.

[0160] The output information unit is used to output offset information and calibration success information.

[0161] In the above technical solution, the laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method and system provided by the present invention utilizes coaxial imaging to acquire images of the powder tray and the laser spot. After processing by a machine vision algorithm, the offset information between the center of the powder tray and the center of the laser spot is obtained to guide the calibration operation. This method is simple, easy to implement, and low in cost, possessing good practicality.

[0162] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for coaxial powder feeding and concentric calibration of laser-directed energy deposition (EDD), characterized in that, The method includes the following steps: Step S1: Ensure that the imaging optical axis is parallel to or coincides with the theoretical central axis of the cladding head; Step S2: Place the substrate under the cladding head; Step S3: Acquire an image of the powder center; Step S4: Image processing and feature extraction to obtain the powder center; Step S5: Acquire laser spot image, turn on the laser, and have the camera capture the laser spot image on the focal plane; Step S6: Image processing and feature extraction to obtain the light center; Step S7: Coaxiality calculation and calibration; Step S8: Repeat steps S3 to S7 until the offset is less than the set threshold, at which point the calibration is considered complete.

2. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 1, characterized in that: In step S3, the center of the powder tray is the powder center, and no powder needs to be dispensed during the entire calibration process.

3. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 1, characterized in that: Step S41: Image edge contour extraction. Perform edge contour extraction on the input image to be processed to obtain an initial set of edge contours. Step S42: Contour length filtering, based on the physical length of the contour, the initial set of edge contours is filtered; Step S43: Contour merging based on cocircularity constraints; Step S44: Algebraic fitting of circular parameters. The merged contour obtained in step S43 is fitted with a circular equation using an algebraic fitting algorithm.

4. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 3, characterized in that: In step S4.3, the contours in the candidate contour set are intelligently merged based on their geometric characteristics. The merging criterion is based on the following cocircularity geometric constraint: Curvature consistency: The contour segments to be merged should have similar radii of curvature; Circle center proximity: The instantaneous circle centers fitted by each contour segment should converge within a narrow spatial range; Contour endpoint proximity: Contour segments belonging to the same circle should have their breakpoints close to each other in space.

5. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 3, characterized in that: In step S4.4, the geometric center coordinates of the laser spot are C_powder(X_p,Y_p); Let the equation of the circle be: (x-a) 2 +(y-b)2=r 2 Where (a,b) is the center of the circle and r is the radius.

6. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 1, characterized in that: Step S61: Adaptive extraction of region of interest for laser spot. For the acquired substrate surface image, the high contrast between the laser spot and the background is used to perform preliminary segmentation through a low global threshold to obtain the connected region containing the laser spot. Based on the bounding rectangle of the connected region, a sub-image is dynamically extracted as the region of interest for the laser spot. Step S62: Gaussian filtering for noise reduction of the spot image. Gaussian low-pass filtering is applied to the region of interest image of the spot to obtain a smoothed spot image. Step S63: Calculate the weighted centroid based on the gray values ​​of the entire image. Use the gray value of each pixel in the smoothed spot image as its weight to calculate the gray-weighted centroid of the entire spot region of interest.

7. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 6, characterized in that: In step 6.3, the coordinates of the centroid are the desired coordinates of the light spot center C_laser(X_l,Y_l); in: W and H are the width and height of the region of interest of the light spot, respectively; I(x,y) is the gray value of the pixel with coordinates (x,y) after Gaussian filtering; X_l and Y_l are the calculated coordinates of the light spot center.

8. The laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to claim 6, characterized in that: Step S71: Calculate the offset vector: Offset ΔX = X_p - X_l Offset ΔY = Y_p - Y_l Total offset Offset = sqrt(ΔX) 2 +ΔY 2 ) The offset direction θ = atan2(ΔY,ΔX); Step S72: Perform calibration. Based on the calculated (ΔX, ΔY), make a fine adjustment in the opposite direction by adjusting the pitch screw on the laser module or powder feeding nozzle inside the cladding head.

9. A calibration system for implementing the laser-directed energy deposition coaxial powder feeding and photoconcentric calibration method according to any one of claims 1-8, characterized in that, include: A coaxial CCD camera, blue light illumination, and a data processing system. The data processing system is connected to the coaxial CCD camera and includes an image processing unit, a coaxiality calculation unit, and an output information unit. A coaxial CCD is used to acquire images of the powder center and the laser spot. Blue light illumination is used to provide lighting conditions for the substrate surface; The image processing unit is used for image processing and feature extraction; The coaxiality calculation unit is used to calculate the lateral offset ΔX, longitudinal offset ΔY, total offset Offset, and offset direction θ; The output information unit is used to display the lateral offset ΔX and the longitudinal offset ΔY, as well as information on whether the calibration is complete.

Images