A method and system for dynamically compensated laser power calibration

Abstract

This invention proposes a dynamic compensation laser power calibration method and system. The method includes: increasing the rated laser power in fixed steps according to the powder material type, and collecting infrared thermal radiation image data at each power point; extracting the effective area and calculating the thermal radiation intensity data; extracting the spatter area features of each infrared thermal radiation image and identifying and marking the spatter area; determining whether the spatter area coverage is less than a threshold, and if so, replacing the current spatter frame's data with the thermal radiation intensity data of the previous effective frame; obtaining standard continuous data of thermal radiation intensity corresponding to each powder material; and fitting the standard continuous data of thermal radiation intensity corresponding to each powder material with the laser power to obtain the optimal power point corresponding to the power value of each powder material. This method can calibrate the laser power of powders of different materials, avoid errors, and eliminate optical signal interference caused by powder spatter, thereby improving the effect of laser power calibration.

Description

Technical Field

[0001] This invention relates to the field of powder bed melting technology in 3D printing, and in particular to a dynamic compensation laser power calibration method and system. Background Technology

[0002] In 3D printing, laser beam power plays a crucial role, especially in additive manufacturing (AM) technologies such as selective laser sintering (SLS) and selective laser melting (SLM). The laser beam, as an energy source, directly affects the melting, solidification, and printing accuracy of the material. Laser beam power is a key control parameter in 3D printing, influencing multiple aspects such as material melting, printing speed, accuracy, and the final object's performance. Precise control of laser power can effectively optimize the printing process, ensuring the quality, strength, and detail of the printed object while improving production efficiency. Therefore, rationally adjusting laser power and coordinating it with other printing parameters is the core of high-quality additive manufacturing.

[0003] A laser device power automatic calibration system and method are disclosed in publication number CN117154537A. The automatic calibration system includes: a laser device and a power identification unit; the laser device is equipped with an automatic calibration function module; after the laser device outputs laser light, the power identification unit identifies the power value of the laser light output by the laser device and sends it to the laser device; after the laser device receives the power value fed back by the power identification unit, the automatic calibration function module calibrates the output power based on the relationship between the output power and the laser's operating temperature and output current.

[0004] Currently, laser equipment in the 3D printing industry undergoes power calibration at the factory. However, factors such as the laser equipment's use, storage, and installation / operation environment can cause a certain error between the power of the laser beam reaching the material surface and the actual set power. To reduce this error, the power of each output laser needs to be calibrated when the laser is installed on the printing equipment. However, traditional laser power calibration uses the same power calibration data for different powder materials, leading to errors and making it unsuitable for high-precision equipment. Furthermore, the light signal interference caused by powder splashing is not effectively eliminated, thus reducing the effectiveness of laser power calibration. Summary of the Invention

[0005] In view of this, the present invention proposes a dynamic compensation laser power calibration method and system, which can calibrate the laser power of powders of different materials, avoid errors, is suitable for high-precision equipment, and eliminates optical signal interference caused by powder splashing, thereby improving the effect of laser power calibration.

[0006] The technical solution of the present invention is implemented as follows: In a first aspect, the present invention provides a dynamically compensated laser power calibration method, the method comprising the following steps:

[0007] S1, based on the type of powder material, the rated power of the laser is increased in fixed steps, and infrared thermal radiation image data corresponding to each power point of each step are collected in sequence.

[0008] S2, preprocess each infrared thermal radiation image data, extract the effective area and calculate the thermal radiation intensity data;

[0009] S3 uses a splash detection algorithm to process each infrared thermal radiation image data, extract splash area features, and identify and mark splash areas;

[0010] S4, preset the splash area coverage threshold, and determine whether the splash area coverage is less than the threshold based on the infrared thermal radiation image data. If it is less than the threshold, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

[0011] S5. Fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point for each powder material based on the fitting results.

[0012] Based on the above technical solution, preferably, in step S1, the laser rated power is increased in fixed steps according to the type of powder material, and infrared thermal radiation image data corresponding to each power point are collected sequentially, including:

[0013] Individual laser power calibration is performed based on different types of powder materials;

[0014] The rated power of the laser is gradually increased from 10% to 100% of the rated power, and the increase is carried out in fixed increments.

[0015] At each step power point, irradiation is carried out for a fixed duration, and after an interval of waiting time, the scanning and acquisition of the next power point is performed.

[0016] During the duration of each power point step, an image acquisition device equipped with a corresponding infrared bandpass filter is used to acquire multiple frames of infrared thermal radiation image data.

[0017] Based on the above technical solutions, preferably, the preprocessing of each infrared thermal radiation image data, extraction of the effective area, and calculation of thermal radiation intensity data in step S2 includes the following sub-steps:

[0018] S21. Gaussian filtering is applied to each infrared thermal radiation image data, and a segmentation algorithm is used to binarize the filtered image, extract the maximum connected region, and expand outward along the boundary of the maximum connected region to obtain the effective region.

[0019] S22, acquire the infrared background image under no laser illumination condition, and calculate the thermal radiation intensity data of the effective area based on the infrared background image and infrared thermal radiation image data, expressed as:

[0020] I t =∑(x,y)∈ROI[I c (x,y)-I b [x,y)]×T(λ)

[0021] In the formula, I t The ROI represents the thermal radiation intensity within the effective region, and ∑(x,y)∈ROI represents summing over all pixels within the effective region. c (x,y) represents the intensity value at pixel (x,y) in the effective region of the current infrared thermal radiation image. b (x,y) represents the intensity value at pixel (x,y) in the infrared background image. c (x,y)-I b (x,y) represents the net thermal signal after removing background noise; T(λ) represents the wavelength correction coefficient.

[0022] Based on the above technical solutions, preferably, step S21, which involves using a segmentation algorithm to binarize the filtered image and extract the maximum connected region, includes the following sub-steps:

[0023] S211: Acquire the infrared thermal radiation image data after Gaussian filtering, and perform binarization processing on the filtered image to obtain a grayscale image;

[0024] S212, count the number of pixels at each gray level in the grayscale image, and calculate the probability distribution of each gray level, expressed as:

[0025] p(i)=g(i) / n

[0026] In the formula, p(i) represents the probability distribution value of each gray level, g(i) represents the number of pixels with gray level i, and n represents the total number of pixels;

[0027] S213, a preset segmentation threshold array is used to segment the grayscale image into a background class and a foreground class according to each segmentation threshold, and the pixel proportion and average grayscale value of the background class and the foreground class are calculated respectively.

[0028] The pixel percentage expression for the background class is:

[0029]

[0030] In the formula, t is the segmentation threshold;

[0031] The expression for the average grayscale value of the background class is:

[0032]

[0033] The expression for the pixel proportion of the foreground class is:

[0034] w1 = 1 - w0;

[0035] The expression for the average gray value of the foreground class is:

[0036]

[0037] S214, Calculate the inter-class variance σ based on the pixel proportion and average gray value of the background and foreground classes corresponding to each segmentation threshold, and obtain the maximum inter-class variance value σ. max Using this as the optimal segmentation threshold T, the grayscale image corresponding to each infrared thermal radiation image is segmented, and the expression for extracting the maximum connected region is:

[0038] σ 2 = w0 × w1 × (μ0 - μ1) 2 .

[0039] Based on the above technical solutions, preferably, step S3, which involves processing the infrared thermal radiation image data using a splash detection algorithm to extract splash area features and identify and mark the splash areas, includes the following sub-steps:

[0040] S31, Gaussian filtering is performed on each infrared thermal radiation image data to obtain a Gaussian filtered image;

[0041] S32, acquire two consecutive Gaussian filtered images corresponding to the same type of powder material, and perform pixel-by-pixel difference on the two consecutive Gaussian filtered images to obtain a difference image, the expression of which is:

[0042] D(x,y)=|I j+1 (x,y)-I j (x,y)|

[0043] In the formula, D(x,y) is the difference image, and I j (x,y) and I j+1 (x,y) is a continuously Gaussian filtered image;

[0044] S34. Threshold segmentation is performed on the difference image according to the optimal segmentation threshold to generate a binary image B(x,y). If D(x,y)>T, then B(x,y)=255, indicating that the pixel belongs to the motion region. If D(x,y)≤T, then B(x,y)=0, indicating that the pixel belongs to the background region.

[0045] S34, perform morphological operations on the binarized image to obtain a binarized enhanced image;

[0046] S35, extract connected regions in the binarized enhanced image, and filter and label splash regions based on features;

[0047] S36, select and highlight the splash area in the corresponding infrared thermal radiation image.

[0048] Based on the above technical solutions, preferably, step S35, which involves extracting connected regions from the binarized enhanced image and identifying splash regions based on features, includes the following sub-steps:

[0049] S351, traverse each pixel in the binarized image. If the pixel value is 255, check its 8 neighboring pixels. If there are no labeled pixels in the 8 neighboring pixels, mark the pixel as a new connected region. If there are labeled pixels in the 8 neighboring pixels, mark the pixel as a connected region with the same value as the smallest pixel in its 8 neighboring pixels, and obtain each connected region in the binarized enhanced image.

[0050] S352, For each connected region, calculate the total number of pixels inside it;

[0051] S353, preset the pixel threshold range of connected regions, compare the total number of pixels inside each connected region with the boundary value of the pixel threshold range of connected regions, obtain the region where the total number of pixels inside the connected region is within the pixel threshold range of connected regions, and obtain the first candidate splash region;

[0052] S354, For the first candidate splash area, calculate the width and height of its bounding rectangle to obtain the aspect ratio of each first candidate splash area;

[0053] S355, preset the aspect ratio threshold range of the connected region, compare the aspect ratio of each first candidate splash region with the boundary value of the aspect ratio threshold range of the connected region, obtain the region where the aspect ratio of the first candidate splash region is within the aspect ratio threshold range of the connected region, and obtain the second candidate splash region;

[0054] S356, for each second candidate splash region, calculate the average value of its internal difference image, and select a local window around the second candidate splash region to calculate the mean and standard deviation of the difference in its local region. The local dynamic threshold is then calculated based on the mean and standard deviation of the difference in the local region, expressed as:

[0055] T l =μ l +k l ×σ l

[0056] In the formula, Tl For local dynamic threshold, μ l Let σ1 be the mean difference of the local region, σ1 be the standard deviation of the difference of the local region, and k1 be an empirical coefficient.

[0057] S357, compare the average value of the internal difference image of the second candidate splash region with the local dynamic threshold, and obtain the region where the average value of the internal difference image of the second candidate splash region is greater than the local dynamic threshold, thus obtaining the final splash region.

[0058] Based on the above technical solution, preferably, the preset splash area coverage threshold mentioned in step S4 is used to determine whether the splash area coverage is less than the threshold according to each infrared thermal radiation image data. If it is less, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material. This includes the following sub-steps:

[0059] S41 uses a splash detection algorithm to detect the splash area in the infrared thermal radiation image data of each power point in real time, and records the time point and duration of the splash.

[0060] S42, when a splash area is detected, the area of ​​each splash area is calculated to obtain the total splash area area, and the splash coverage rate is calculated based on the ratio of the total splash area area to the total image area.

[0061] S43, preset the splash area coverage threshold, determine whether the splash area coverage is less than the splash coverage threshold. If it is less, during each frame of the splash duration, use the thermal radiation intensity data of the previous valid frame to replace the data of the current splash frame. Otherwise, trigger the laser to pause and re-spread powder for acquisition, and obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

[0062] Based on the above technical solution, preferably, step S5, which involves fitting the standard continuous data of thermal radiation intensity corresponding to each powder material with the laser power, and obtaining the optimal power point for each powder material based on the fitting result, includes the following sub-steps:

[0063] S51, based on the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power at the corresponding points, a thermal radiation intensity fitting model is obtained, the expression of which is:

[0064] I t =aP 2 +bP+c

[0065] In the formula, I tLet be the thermal radiation intensity value, a be the contribution intensity of the square of the laser power to the thermal accumulation, b be the contribution intensity of the linear term of the laser power to the thermal accumulation, P be the output power of the laser, and c be the basic thermal accumulation.

[0066] S52, based on the thermal radiation intensity fitting model, obtain the point where the rate of change of laser power with respect to the thermal radiation intensity value is the largest, and obtain the optimal power point for each powder material, expressed as:

[0067] P opt =ArgMaxP(dP / dI) t )

[0068] In the formula, P opt For the optimal power point, dP / dI t The rate of change of thermal radiation intensity;

[0069] Each powder material is associated with its corresponding optimal power point and stored in the powder power calibration database for laser power calibration.

[0070] Secondly, the present invention also provides a dynamically compensated laser power calibration system, implemented using a dynamically compensated laser power calibration method, the system comprising:

[0071] The acquisition module is used to increase the rated power of the laser in fixed steps according to the type of powder material, and to sequentially acquire the infrared thermal radiation image data corresponding to each power point.

[0072] The data processing and conversion module is used to preprocess the infrared thermal radiation image data, extract the effective area, and calculate the thermal radiation intensity data.

[0073] The splash region extraction module is used to process each infrared thermal radiation image data using a splash detection algorithm, extract splash region features, and identify and mark splash regions.

[0074] The data optimization module is used to preset the splash area coverage threshold. Based on the infrared thermal radiation image data, it determines whether the splash area coverage is less than the threshold. If it is less, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

[0075] The model building module is used to fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point for each powder material based on the fitting results.

[0076] Thirdly, the present invention also provides a computer-readable storage medium storing a dynamically compensated laser power calibration method program, wherein the dynamically compensated laser power calibration method program is executed to realize the dynamically compensated laser power calibration method.

[0077] The dynamic compensation laser power calibration method and system of the present invention have the following advantages over the prior art:

[0078] (1) By gradually increasing the rated power of the laser and collecting infrared thermal radiation image data at each step power point, the image data is preprocessed to extract the effective area and calculate the thermal radiation intensity. The splash detection algorithm is used to identify and mark the splash area. The splash area coverage threshold is preset to dynamically compensate for the impact of splash on the thermal radiation intensity data, ensuring that accurate and continuous thermal radiation intensity standard data is obtained. Finally, by fitting the relationship between thermal radiation intensity and laser power, the optimal power point corresponding to the power value of each powder material is obtained. This method is not only applicable to calibration of different powders, but also eliminates the optical signal interference caused by powder splash, thereby improving the accuracy and reliability of laser power calibration.

[0079] (2) Through differential detection, threshold segmentation and morphological operations, the splashing area can be accurately extracted. Feature selection further improves the accuracy of detection and reduces false detection and missed detection. Gaussian filtering and morphological operations can effectively suppress noise and improve the robustness of the algorithm. Adaptive thresholding and feature selection can adapt to different scenes and image features and improve the versatility of the algorithm.

[0080] (3) By using connected region labeling, area filtering, shape filtering and intensity filtering, the splash region can be accurately extracted. The multi-step filtering mechanism effectively reduces false detection and false negative detection, and improves the accuracy of detection. The use of local dynamic threshold and empirical coefficient improves the adaptability of the algorithm and can cope with different scenarios and image features. Attached Figure Description

[0081] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0082] Figure 1 A flowchart of the dynamic compensation laser power calibration method of the present invention;

[0083] Figure 2 This is a block diagram illustrating the principle of the synchronous control unit of the dynamic compensation laser power calibration method of the present invention.

[0084] Figure 3 Infrared thermal radiation intensity image of the dynamic compensation laser power calibration method of the present invention;

[0085] Figure 4 The Gaussian-filtered infrared thermal radiation grayscale image of the dynamic compensation laser power calibration method of the present invention.

[0086] Figure 5 This is a schematic diagram of laser output power calibration data for an embodiment of the dynamic compensation laser power calibration method of the present invention.

[0087] Figure 6 This is a schematic diagram of laser output power calibration data for Embodiment 2 of the dynamic compensation laser power calibration method of the present invention. Detailed Implementation

[0088] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0089] like Figure 1 As shown, in a first aspect, the present invention provides a dynamically compensated laser power calibration method, the method comprising the following steps:

[0090] S1, based on the type of powder material, increases the rated power of the laser in fixed steps, and sequentially collects infrared thermal radiation image data corresponding to each power point.

[0091] Step S1 includes:

[0092] Individual laser power calibration is performed based on different types of powder materials;

[0093] The rated power of the laser is gradually increased from 10% to 100% of the rated power, and the increase is carried out in fixed increments.

[0094] At each step power point, irradiation is carried out for a fixed duration, and after an interval of waiting time, the scanning and acquisition of the next power point is performed.

[0095] During the duration of each power point step, an image acquisition device equipped with a corresponding infrared bandpass filter is used to acquire multiple frames of infrared thermal radiation image data.

[0096] It should be noted that, based on the physical properties of the powder, it can be classified into metal powder and ceramic powder, etc. For each powder material, the laser power is calibrated independently to ensure the accuracy and uniqueness of the collected data. This laser equipment includes a powder box with N independent cavities, each containing different types of powder materials. The laser equipment also has a built-in stepper motor-driven powder pushing mechanism, which pushes the powder box to move on the guide rail and moves the corresponding independent cavity to the laser processing area, thereby allowing for the switching of different types of powder processing.

[0097] Furthermore, the initial power is set to 10% of the laser's rated power, and the laser power is gradually increased in fixed increments of 10% until 100% of the rated power is reached. At each power point, the laser is continuously irradiated for a fixed duration of 10ms to ensure that the powder is fully heated. After irradiation, a fixed interval of 50ms is waited for thermal relaxation to stabilize the system and prevent heat accumulation from affecting the data at the next power point.

[0098] like Figure 2 As shown, a high-sensitivity infrared thermal imaging camera with a CCD or CMOS sensor and a frame rate ≥1000fps is used. It is equipped with a detachable infrared bandpass filter, in which the center wavelength is matched with the laser wavelength, such as 1064nm, and the bandwidth is ±20nm, to filter infrared radiation of specific wavelengths and improve image quality. During the duration of each power point, 100 frames of infrared thermal radiation image data are continuously acquired to obtain stable thermal distribution information. The acquired image data is correlated with the corresponding laser power, powder material, acquisition time and other information for subsequent analysis.

[0099] Understandably, it also includes a synchronization control unit, which enables the laser, camera, and powder box to achieve millisecond-level timing synchronization via PLC or FPGA. This ensures that the power change of the laser is precisely synchronized with the start time of the image acquisition device, avoiding data deviation. It also allows experiments to be conducted in a dark room or under light-shielding measures to reduce the impact of ambient infrared radiation on the image and to avoid strong airflow in the experimental area, preventing powder movement or heat loss that could affect heat distribution.

[0100] S2 preprocesses each infrared thermal radiation image data, extracts the effective area, and calculates the thermal radiation intensity data.

[0101] like Figure 3 and Figure 4 As shown, step S2 includes the following sub-steps:

[0102] S21. Gaussian filtering is applied to each infrared thermal radiation image data, and a segmentation algorithm is used to binarize the filtered image, extract the maximum connected region, and expand outward along the boundary of the maximum connected region to obtain the effective region.

[0103] S22, acquire the infrared background image under no laser illumination condition, and calculate the thermal radiation intensity data of the effective area based on the infrared background image and infrared thermal radiation image data, expressed as:

[0104] I t =∑(x,y)∈ROI[I c (x,y)-I b [x,y)]×T(λ)

[0105] In the formula, I t The ROI represents the thermal radiation intensity within the effective region, and ∑(x,y)∈ROI represents summing over all pixels within the effective region. c (x,y) represents the intensity value at pixel (x,y) in the effective region of the current infrared thermal radiation image. b (x,y) represents the intensity value at pixel (x,y) in the infrared background image. c (x,y)-I b (x,y) represents the net thermal signal after removing background noise; T(λ) represents the wavelength correction coefficient.

[0106] It should be noted that a Gaussian filter is used to perform a convolution operation on each frame of the infrared thermal radiation image. An appropriate Gaussian kernel size and standard deviation are selected and adjusted according to the image resolution and noise level to convert the image into a binary image, which facilitates the extraction of connected regions. The largest connected region in the image is identified, which corresponds to the main area affected by the laser. The area of ​​each connected region is calculated, and the connected region with the largest area is selected as the target region. The region is then extended outward by 5px along the boundary of the largest connected region to avoid edge effects and obtain the effective region.

[0107] Step S21, which involves using a segmentation algorithm to binarize the filtered image and extract the largest connected region, includes the following sub-steps:

[0108] S211: Acquire the infrared thermal radiation image data after Gaussian filtering, and perform binarization processing on the filtered image to obtain a grayscale image;

[0109] S212, count the number of pixels at each gray level in the grayscale image, and calculate the probability distribution of each gray level, expressed as:

[0110] p(i)=g(i) / n

[0111] In the formula, p(i) represents the probability distribution value of each gray level, g(i) represents the number of pixels with gray level i, and n represents the total number of pixels;

[0112] S213, a preset segmentation threshold array is used to segment the grayscale image into a background class and a foreground class according to each segmentation threshold, and the pixel proportion and average grayscale value of the background class and the foreground class are calculated respectively.

[0113] The pixel percentage expression for the background class is:

[0114]

[0115] In the formula, t is the segmentation threshold;

[0116] The expression for the average grayscale value of the background class is:

[0117]

[0118] The expression for the pixel proportion of the foreground class is:

[0119] w1 = 1 - w0;

[0120] The expression for the average gray value of the foreground class is:

[0121]

[0122] S214, Calculate the inter-class variance σ based on the pixel proportion and average gray value of the background and foreground classes corresponding to each segmentation threshold, and obtain the maximum inter-class variance value σ. max Using this as the optimal segmentation threshold T, the grayscale image corresponding to each infrared thermal radiation image is segmented, and the maximum connected region is extracted. The expression is:

[0123] σ 2 = w0 × w1 × (μ0 - μ1) 2 .

[0124] It should be noted that the method can effectively extract the largest connected region from infrared thermal radiation images, providing an accurate effective region for subsequent thermal radiation intensity calculation. The application of the inter-class variance method improves the accuracy and robustness of binarization processing, ensuring the effectiveness and integrity of the extracted region.

[0125] S3 uses a splash detection algorithm to process each infrared thermal radiation image data, extract splash area features, and identify and mark splash areas;

[0126] Step S3 includes the following sub-steps:

[0127] S31, Gaussian filtering is performed on each infrared thermal radiation image data to obtain a Gaussian filtered image;

[0128] S32, acquire two consecutive Gaussian filtered images corresponding to the same type of powder material, and perform pixel-by-pixel difference on the two consecutive Gaussian filtered images to obtain a difference image, the expression of which is:

[0129] D(x,y)=|I j+1 (x,y)-I j (x,y)|

[0130] In the formula, D(x,y) is the difference image, and I j (x,y) and I j+1 (x,y) is a continuously Gaussian filtered image;

[0131] S34. Threshold segmentation is performed on the difference image according to the optimal segmentation threshold to generate a binary image B(x,y). If D(x,y)>T, then B(x,y)=255, indicating that the pixel belongs to the motion region. If D(x,y)≤T, then B(x,y)=0, indicating that the pixel belongs to the background region.

[0132] S34, perform morphological operations on the binarized image to obtain a binarized enhanced image;

[0133] S35, extract connected regions in the binarized enhanced image, and filter and label splash regions based on features;

[0134] S36, select and highlight the splash area in the corresponding infrared thermal radiation image.

[0135] It should be noted that Gaussian filtering is applied to each frame of the infrared thermal radiation image to obtain a Gaussian filtered image. The appropriate Gaussian kernel size and standard deviation are selected and adjusted according to image resolution and noise level. The filtered image is smoother, reducing the impact of noise on the difference results. Two consecutive Gaussian filtered images of the same type of powder material are obtained, and pixel-by-pixel difference is performed on these two consecutive images to obtain a difference image. The splashed area will show obvious pixel value changes between consecutive frames. In the difference image, the splashed area appears as a bright area, while the background area appears as a low-brightness or zero-value area. Thresholding is performed on the difference image according to the optimal segmentation threshold T to generate a binarized image. Thresholding can distinguish the splashed area from areas with minor changes or noise in the difference image. In the binarized image, the splashed area is white (255), and the background area is black (0), facilitating subsequent processing. Morphological operations are performed on the binarized image to eliminate noise and holes. The image undergoes opening and closing operations. The opening operation involves erosion followed by dilation to eliminate isolated noise points, while the closing operation involves dilation followed by erosion to fill small holes and connect adjacent regions. The binarized enhanced image after morphological operations is clearer, the splash area is more complete, and noise is effectively suppressed. A connected component labeling algorithm is used to label all connected components in the binarized image, removing those with excessively small areas, as these may be noise. Further shape feature filtering is performed based on the aspect ratio of the splash area. Considering the motion intensity of the splash area, motion feature filtering is performed by combining multi-frame information analysis. The extracted connected components more accurately reflect the splash area, reducing false positives and false negatives. The filtered splash areas are mapped back to the original infrared thermal radiation image. The splash areas are selected in the image and highlighted using different colors or brightness levels. In the original infrared thermal radiation image, the splash areas are clearly marked, facilitating observation and analysis of the splash phenomenon.

[0136] In this embodiment, the splash region can be accurately extracted through differential detection, threshold segmentation, and morphological operations. Feature selection further improves the detection accuracy and reduces false detections and false negatives. Gaussian filtering and morphological operations can effectively suppress noise and improve the robustness of the algorithm. Adaptive thresholding and feature selection can adapt to different scenarios and image features, improving the versatility of the algorithm.

[0137] Step S35, which involves extracting connected regions from the binarized enhanced image and filtering and labeling splash regions based on features, includes the following sub-steps:

[0138] S351, traverse each pixel in the binarized image. If the pixel value is 255, check its 8 neighboring pixels. If there are no labeled pixels in the 8 neighboring pixels, mark the pixel as a new connected region. If there are labeled pixels in the 8 neighboring pixels, mark the pixel as a connected region with the same value as the smallest pixel in its 8 neighboring pixels, and obtain each connected region in the binarized enhanced image.

[0139] S352, For each connected region, calculate the total number of pixels inside it;

[0140] S353, preset the pixel threshold range of connected regions, compare the total number of pixels inside each connected region with the boundary value of the pixel threshold range of connected regions, obtain the region where the total number of pixels inside the connected region is within the pixel threshold range of connected regions, and obtain the first candidate splash region;

[0141] S354, For the first candidate splash area, calculate the width and height of its bounding rectangle to obtain the aspect ratio of each first candidate splash area;

[0142] S355, preset the aspect ratio threshold range of the connected region, compare the aspect ratio of each first candidate splash region with the boundary value of the aspect ratio threshold range of the connected region, obtain the region where the aspect ratio of the first candidate splash region is within the aspect ratio threshold range of the connected region, and obtain the second candidate splash region;

[0143] S356, for each second candidate splash region, calculate the average value of its internal difference image, and select a local window around the second candidate splash region to calculate the mean and standard deviation of the difference in its local region. The local dynamic threshold is then calculated based on the mean and standard deviation of the difference in the local region, expressed as:

[0144] T l =μ l +k l ×σ l

[0145] In the formula, T l For local dynamic threshold, μ l Let σ1 be the mean difference of the local region, σ1 be the standard deviation of the difference of the local region, and k1 be an empirical coefficient.

[0146] S357, compare the average value of the internal difference image of the second candidate splash region with the local dynamic threshold, and obtain the region where the average value of the internal difference image of the second candidate splash region is greater than the local dynamic threshold, thus obtaining the final splash region.

[0147] It should be noted that by traversing pixels and checking neighborhoods, moving regions in the binarized image are marked as connected regions, ensuring that each connected region has a unique identifier for easy subsequent processing. Splashed regions usually have a certain area, and area filtering can remove noise regions. The preset pixel threshold range can adapt to different scenes and image features. Splashed regions usually have specific shape features, and aspect ratio filtering can further remove false detection regions. The preset aspect ratio threshold range can improve detection accuracy. Local dynamic thresholds can consider the differential features of local regions and improve the adaptability of threshold segmentation. The empirical coefficient k1 can be adjusted according to the actual situation to optimize the threshold segmentation effect. Splashed regions usually have high differential intensity, and intensity filtering can further remove false detection regions. Local dynamic thresholds ensure the accuracy of filtering.

[0148] Understandably, by using connected component labeling, area filtering, shape filtering, and intensity filtering, splash regions can be accurately extracted. The multi-step filtering mechanism effectively reduces false detections and false negatives, improving detection accuracy. The use of local dynamic thresholds and empirical coefficients enhances the algorithm's adaptability, enabling it to handle different scenarios and image features.

[0149] Specifically, in this embodiment, the pixel threshold range for connected regions is 10px. 2 Area < 1000px 2 The aspect ratio threshold range for connected regions is 1:3 < aspect ratio < 3:1.

[0150] S4: Preset a splash area coverage threshold. Based on each infrared thermal radiation image data, determine whether the splash area coverage is less than the threshold. If it is less, use the thermal radiation intensity data of the previous valid frame to replace the data of the current splash frame; otherwise, trigger laser pause and re-lay powder for acquisition to obtain standard continuous data of thermal radiation intensity corresponding to each powder material.

[0151] Step S4 includes the following sub-steps:

[0152] S41 uses a splash detection algorithm to detect the splash area in the infrared thermal radiation image data of each power point in real time, and records the time point and duration of the splash.

[0153] S42, when a splash area is detected, the area of ​​each splash area is calculated to obtain the total splash area area, and the splash coverage rate is calculated based on the ratio of the total splash area area to the total image area.

[0154] S43, preset the splash area coverage threshold, determine whether the splash area coverage is less than the splash coverage threshold. If it is less, during each frame of the splash duration, use the thermal radiation intensity data of the previous valid frame to replace the data of the current splash frame. Otherwise, trigger the laser to pause and re-spread powder for acquisition, and obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

[0155] It should be noted that by using threshold judgment and data correction, the impact of splash on thermal radiation intensity data is reduced, and the accuracy of the data is improved. When the splash area coverage is small, data from the previous valid frame is used as a substitute to maintain the continuity of the data, which helps to ensure the consistency and coherence of the data in subsequent analysis. The laser pause and re-spreading mechanism ensures the stability and reliability of the system when splashing is severe.

[0156] S5. Fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point for each powder material based on the fitting results.

[0157] Step S5 includes the following sub-steps:

[0158] S51, based on the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power at the corresponding points, a thermal radiation intensity fitting model is obtained, the expression of which is:

[0159] I t =aP 2 +bP+c

[0160] In the formula, I t Let be the thermal radiation intensity value, a be the contribution intensity of the square of the laser power to the thermal accumulation, b be the contribution intensity of the linear term of the laser power to the thermal accumulation, P be the output power of the laser, and c be the basic thermal accumulation.

[0161] S52, based on the thermal radiation intensity fitting model, obtain the point where the rate of change of laser power with respect to the thermal radiation intensity value is the largest, and obtain the optimal power point for each powder material, expressed as:

[0162] P opt =ArgMaxP(dP / dI) t )

[0163] In the formula, P opt For the optimal power point, dP / dI t The rate of change of thermal radiation intensity;

[0164] Each powder material is associated with its corresponding optimal power point and stored in the powder power calibration database for laser power calibration.

[0165] like Figure 5 This is the output power calibration data for the 316L powder laser of the SLM250 device in Example 1; Figure 6 This is the calibration data for the Ti6Al4V powder laser output power of the SLM250 device in Example 2.

[0166] It should be noted that the optimal power point corresponds to the point where the rate of change of laser power with respect to the thermal radiation intensity is the largest. That is, at this point, a small change in laser power will lead to a significant change in thermal radiation intensity. By solving for the point with the largest rate of change, the optimal power point can be obtained, providing a basis for laser power calibration. By establishing a mathematical relationship between thermal radiation intensity and laser power through polynomial fitting, a thermal radiation intensity fitting model is obtained. The fitting model can well describe the nonlinear relationship between thermal radiation intensity and laser power. The optimal power point corresponds to the point where the rate of change of laser power with respect to the thermal radiation intensity is the largest. By solving for the point with the largest rate of change, the optimal power point can be obtained, improving the accuracy and reliability of laser power calibration.

[0167] Secondly, the present invention also provides a dynamically compensated laser power calibration system, implemented using a dynamically compensated laser power calibration method, the system comprising:

[0168] The acquisition module is used to increase the rated power of the laser in fixed steps according to the type of powder material, and to sequentially acquire the infrared thermal radiation image data corresponding to each power point.

[0169] The data processing and conversion module is used to preprocess the infrared thermal radiation image data, extract the effective area, and calculate the thermal radiation intensity data.

[0170] The splash region extraction module is used to process each infrared thermal radiation image data using a splash detection algorithm, extract splash region features, and identify and mark splash regions.

[0171] The data optimization module is used to preset the splash area coverage threshold. Based on the infrared thermal radiation image data, it determines whether the splash area coverage is less than the threshold. If it is less, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

[0172] The model building module is used to fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point for each powder material based on the fitting results.

[0173] It should be noted that this system is a system corresponding to the above-mentioned dynamic compensation laser power calibration method. All implementation methods in the above method embodiments are applicable to the embodiments of this system and can achieve the same technical effect.

[0174] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0175] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0176] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0177] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0178] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0179] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0180] Furthermore, it should be noted that in the system and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0181] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing system. The computing system can be a known general-purpose system. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0182] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamically compensated laser power calibration, characterized in that: The method includes the following steps: S1, based on the type of powder material, the rated power of the laser is increased in fixed steps, and infrared thermal radiation image data corresponding to each power point of each step are collected in sequence. S2, preprocess each infrared thermal radiation image data, extract the effective area and calculate the thermal radiation intensity data; S3 uses a splash detection algorithm to process each infrared thermal radiation image data, extract splash area features, and identify and mark splash areas; Step S3 includes the following sub-steps: S31, Gaussian filtering is performed on each infrared thermal radiation image data to obtain a Gaussian filtered image; S32, acquire two consecutive Gaussian filtered images corresponding to the same type of powder material, and perform pixel-by-pixel difference on the two consecutive Gaussian filtered images to obtain a difference image, the expression of which is: D(x,y)=|I j+1 (x,y)-I j (x,y)| In the formula, D(x,y) is the difference image, and I j (x,y) and I j+1 (x,y) is a continuously Gaussian filtered image; S34. Threshold segmentation is performed on the difference image according to the optimal segmentation threshold to generate a binary image B(x,y). If D(x,y)>T, then B(x,y)=255, indicating that the pixel belongs to the motion region. If D(x,y)≤T, then B(x,y)=0, indicating that the pixel belongs to the background region. S34, perform morphological operations on the binarized image to obtain a binarized enhanced image; S35, extract connected regions in the binarized enhanced image, and filter and label splash regions based on features; S36, select and highlight the splash area in the corresponding infrared thermal radiation image; S4, preset the splash area coverage threshold, and determine whether the splash area coverage is less than the threshold based on the infrared thermal radiation image data. If it is less than the threshold, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material. S5. Fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point corresponding to the power value of each powder material based on the fitting results.

2. The laser power calibration method with dynamic compensation as described in claim 1, characterized in that, Step S1 involves increasing the laser's rated power in fixed increments according to the powder material type, and sequentially acquiring infrared thermal radiation image data corresponding to each power level, including: Individual laser power calibration is performed based on different types of powder materials; The rated power of the laser is gradually increased from 10% to 100% of the rated power, and the increase is carried out in fixed increments. At each step power point, irradiation is carried out for a fixed duration, and the next power point is scanned and acquired after an interval of waiting time. During the duration of each step power point, an image acquisition device equipped with a corresponding infrared bandpass filter is used to acquire multiple frames of infrared thermal radiation image data.

3. The dynamic compensation laser power calibration method as described in claim 1, characterized in that, Step S2, which involves preprocessing the infrared thermal radiation image data, extracting the effective region, and calculating the thermal radiation intensity data, includes the following sub-steps: S21. Gaussian filtering is applied to each infrared thermal radiation image data, and a segmentation algorithm is used to binarize the filtered image, extract the maximum connected region, and expand outward along the boundary of the maximum connected region to obtain the effective region. S22, acquire the infrared background image under no laser illumination condition, and calculate the thermal radiation intensity data of the effective area based on the infrared background image and infrared thermal radiation image data, expressed as: I t =∑(x,y)∈ROI[I c (x,y)-I b (x,y)]×T(λ) In the formula, I t The ROI represents the thermal radiation intensity within the effective region, and ∑(x,y)∈ROI represents summing over all pixels within the effective region. c (x,y) represents the intensity value at pixel (x,y) in the effective region of the current infrared thermal radiation image. b (x,y) represents the intensity value at pixel (x,y) in the infrared background image. c (x,y)-I b (x,y) represents the net thermal signal after removing background noise; T(λ) represents the wavelength correction coefficient.

4. The dynamic compensation laser power calibration method as described in claim 3, characterized in that, Step S21, which involves using a segmentation algorithm to binarize the filtered image and extract the largest connected region, includes the following sub-steps: S211: Acquire the infrared thermal radiation image data after Gaussian filtering, and perform binarization processing on the filtered image to obtain a grayscale image; S212, count the number of pixels at each gray level in the grayscale image, and calculate the probability distribution of each gray level, expressed as: p(i)=g(i) / n In the formula, p(i) represents the probability distribution value of each gray level, g(i) represents the number of pixels with gray level i, and n represents the total number of pixels; S213, a preset segmentation threshold array is used to segment the grayscale image into a background class and a foreground class according to each segmentation threshold, and the pixel proportion and average grayscale value of the background class and the foreground class are calculated respectively. The pixel percentage expression for the background class is: In the formula, t is the segmentation threshold; The expression for the average grayscale value of the background class is: The expression for the pixel proportion of the foreground class is: w1 = 1 - w0; The expression for the average gray value of the foreground class is: S214, Calculate the inter-class variance σ based on the pixel proportion and average gray value of the background and foreground classes corresponding to each segmentation threshold, and obtain the maximum inter-class variance value σ. max Using this as the optimal segmentation threshold T, the grayscale image corresponding to each infrared thermal radiation image is segmented, and the maximum connected region is extracted. The expression is: s 2 =w0×w1×(μ0-μ1) 2 。 5. The laser power calibration method with dynamic compensation as described in claim 1, characterized in that, Step S35, which involves extracting connected regions from the binarized enhanced image and filtering and labeling splash regions based on features, includes the following sub-steps: S351, traverse each pixel in the binarized image. If the pixel value is 255, check its 8 neighboring pixels. If there are no labeled pixels in the 8 neighboring pixels, mark the pixel as a new connected region. If there are labeled pixels in the 8 neighboring pixels, mark the pixel as a connected region with the same value as the smallest pixel in its 8 neighboring pixels, and obtain each connected region in the binarized enhanced image. S352, For each connected region, calculate the total number of pixels inside it; S353, preset the pixel threshold range of connected regions, compare the total number of pixels inside each connected region with the boundary value of the pixel threshold range of connected regions, obtain the region where the total number of pixels inside the connected region is within the pixel threshold range of connected regions, and obtain the first candidate splash region; S354, For the first candidate splash area, calculate the width and height of its bounding rectangle to obtain the aspect ratio of each first candidate splash area; S355, preset the aspect ratio threshold range of the connected region, compare the aspect ratio of each first candidate splash region with the boundary value of the aspect ratio threshold range of the connected region, obtain the region where the aspect ratio of the first candidate splash region is within the aspect ratio threshold range of the connected region, and obtain the second candidate splash region; S356, for each second candidate splash region, calculate the average value of its internal difference image, and select a local window around the second candidate splash region to calculate the mean and standard deviation of the difference in its local region. The local dynamic threshold is then calculated based on the mean and standard deviation of the difference in the local region, expressed as: T l =μ l +k l ×s l In the formula, T l For local dynamic threshold, μ l Let σ1 be the mean difference of the local region, σ1 be the standard deviation of the difference of the local region, and k1 be an empirical coefficient. S357, compare the average value of the internal difference image of the second candidate splash region with the local dynamic threshold, and obtain the region where the average value of the internal difference image of the second candidate splash region is greater than the local dynamic threshold, thus obtaining the final splash region.

6. The laser power calibration method with dynamic compensation as described in claim 1, characterized in that, The preset splash area coverage threshold mentioned in step S4 is used to determine whether the splash area coverage is less than the threshold based on the infrared thermal radiation image data. If it is less, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material. This includes the following sub-steps: S41 uses a splash detection algorithm to detect the splash area in the infrared thermal radiation image data of each power point in real time, and records the time point and duration of the splash. S42, when a splash area is detected, the area of ​​each splash area is calculated to obtain the total splash area area, and the splash coverage rate is calculated based on the ratio of the total splash area area to the total image area. S43, preset the splash area coverage threshold, determine whether the splash area coverage is less than the splash coverage threshold. If it is less, during each frame of the splash duration, use the thermal radiation intensity data of the previous valid frame to replace the data of the current splash frame. Otherwise, trigger the laser to pause and re-spread powder for acquisition, and obtain the standard continuous data of thermal radiation intensity corresponding to each powder material.

7. The laser power calibration method with dynamic compensation as described in claim 6, characterized in that, Step S5, which involves fitting the standard continuous data of thermal radiation intensity corresponding to each powder material with the laser power, and obtaining the optimal power point for each powder material based on the fitting results, includes the following sub-steps: S51, based on the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power at the corresponding points, a thermal radiation intensity fitting model is obtained, the expression of which is: I t =aP 2 +bP+c In the formula, I t Let be the thermal radiation intensity value, a be the contribution intensity of the square of the laser power to the thermal accumulation, b be the contribution intensity of the linear term of the laser power to the thermal accumulation, P be the output power of the laser, and c be the basic thermal accumulation. S52, based on the thermal radiation intensity fitting model, obtain the point where the rate of change of laser power with respect to the thermal radiation intensity value is the largest, and obtain the optimal power point for each powder material, expressed as: P opt =ArgMaxP(dP / dI t ) In the formula, P opt For the optimal power point, dP / dI t The rate of change of thermal radiation intensity; Each powder material is associated with its corresponding optimal power point and stored in the powder power calibration database for laser power calibration.

8. A dynamically compensated laser power calibration system, implemented using the dynamically compensated laser power calibration method as described in any one of claims 1 to 7, characterized in that, The system includes: The acquisition module is used to increase the rated power of the laser in fixed steps according to the type of powder material, and to sequentially acquire the infrared thermal radiation image data corresponding to each power point. The data processing and conversion module is used to preprocess the infrared thermal radiation image data, extract the effective area, and calculate the thermal radiation intensity data. The splash region extraction module is used to process each infrared thermal radiation image data using a splash detection algorithm, extract splash region features, and identify and mark splash regions. The data optimization module is used to preset the splash area coverage threshold. Based on the infrared thermal radiation image data, it determines whether the splash area coverage is less than the threshold. If it is less, the thermal radiation intensity data of the previous valid frame is used to replace the data of the current splash frame; otherwise, the laser is paused and the powder is re-spread and collected to obtain the standard continuous data of thermal radiation intensity corresponding to each powder material. The model building module is used to fit the standard continuous data of thermal radiation intensity corresponding to each powder material and the laser power, and obtain the optimal power point for each powder material based on the fitting results.

9. A computer-readable storage medium, characterized in that, The storage medium stores a dynamic compensation laser power calibration method program, which, when executed, implements the dynamic compensation laser power calibration method as described in any one of claims 1 to 7.

Images