Part layer height dimension detection method and device for laser wire feeding additive manufacturing

Abstract

The present invention relates to the technical field of linear dimension measurement, and in particular to a method and device for detecting layer height dimensions of parts manufactured by laser wire feeding additive manufacturing, wherein the method comprises: obtaining an image to be detected of a part manufactured by laser wire feeding additive manufacturing, denoising the image to be detected to obtain a smooth image, and processing the smooth image to obtain a binary image, obtaining a seed contour located inside the edge of the image of the part manufactured by laser wire feeding additive manufacturing according to the binary image, solving a pre-constructed image edge segmentation model, and using the seed contour as an initial evolution curve input to perform image edge segmentation of the part manufactured by laser wire feeding additive manufacturing to calculate the actual layer height dimension corresponding to the image edge. Thus, the problem in the related art that the part edge positioning method based on the traditional edge detection operator has low detection accuracy and weak robustness, which makes it difficult to combine it with the visual ranging solution is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of linear dimension measurement, and in particular to a method and device for detecting layer height dimensions of parts used in laser wire feeding additive manufacturing. Background Art

[0002] In related technologies, the measurement of layer height dimensions of parts manufactured by laser wire feeding additive manufacturing can be carried out by establishing a part positioning model based on deep learning methods, predicting the part position based on this model, and calculating the layer height dimensions of the part in combination with a camera calibration algorithm.

[0003] However, in the related art, the deep learning model established by the above method lacks explicit analytical mathematical expression and only predicts the approximate area where the part is located. In addition, the actual laser wire feeding additive manufacturing environment is complex and changes make it difficult for the model prediction accuracy to stabilize, which ultimately leads to insufficient prediction accuracy of the part position. At the same time, the part edge positioning method based on the traditional edge detection operator has problems such as low detection accuracy and weak robustness, which makes it difficult to combine with the visual ranging solution and needs to be improved urgently. Summary of the invention

[0004] The present invention provides a method and device for detecting layer height dimensions of parts for laser wire feeding additive manufacturing, so as to solve the problem in the related art that the part edge positioning method based on the traditional edge detection operator has low detection accuracy and weak robustness, which makes it difficult to combine it with the visual ranging solution.

[0005] The first aspect of the present invention provides a method for detecting layer height dimensions of parts manufactured by laser wire feeding additive manufacturing, comprising the following steps: obtaining an image to be detected of a part manufactured by laser wire feeding additive manufacturing; denoising the image to be detected to obtain a smooth image, and processing the smooth image to obtain a binary image; obtaining a seed contour located inside the edge of the image of the part manufactured by laser wire feeding additive manufacturing based on the binary image; solving a pre-constructed image edge segmentation model, and using the seed contour as an initial evolution curve input to perform image edge segmentation of the part manufactured by laser wire feeding additive manufacturing, so as to calculate the actual layer height dimension corresponding to the image edge.

[0006] Through the above technical solution, the embodiment of the present invention can perform denoising on the image to be detected to obtain a smooth image, remove interference factors, provide reliable basic data for subsequent detection, process the smooth image into a binary image, simplify information, and facilitate the acquisition of seed contours. In terms of part contour detection, obtaining the seed contour based on the binary image can accurately locate the key area, improve the detection targeting, and use the pre-constructed image edge segmentation model and the seed contour as the initial evolution curve to perform image edge segmentation to calculate the layer height size, which can not only adapt to complex shape structures to improve calculation accuracy, but also has good versatility.

[0007] Optionally, in one embodiment of the present invention, before solving the pre-constructed image edge segmentation model, it also includes: introducing an iterative penalty term and a variable-scale convolution kernel in the energy functional modeling process of the target Chan-Vese algorithm to establish the image edge segmentation model.

[0008] Through the above technical solution, the embodiment of the present invention can introduce iterative penalty terms and variable-scale convolution kernels in the energy functional modeling process of the target Chan-Vese algorithm to establish an image edge segmentation model. The introduction of iterative penalty terms helps to constrain the model solution process and avoid unreasonable solutions or overfitting in the iterative calculation process, thereby improving the accuracy and stability of the model solution. The variable-scale convolution kernel can adapt to the feature extraction requirements of part images at different scales. For images of laser wire additive manufacturing parts that may have different scale features, it can capture image edge information related to layer height size more comprehensively and accurately, thereby improving the accuracy of the entire part layer height size detection, which is more advantageous when processing part images with complex structures and multi-scale features.

[0009] Optionally, in one embodiment of the present invention, the introduction of the iterative penalty term and the variable-scale convolution kernel includes: when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration, the iterative penalty term is introduced based on the similarity between the pixel value of each pixel point and the target value.

[0010] Through the above technical solution, the embodiment of the present invention can introduce an iterative penalty term based on the similarity between the pixel point and the target value in each iterative calculation, and can accurately adjust the calculation process according to the actual situation of the pixel. In this way, the calculation results can be effectively constrained to avoid calculation deviations caused by fluctuations or deviations in pixel values, and the accuracy of fitting the grayscale values ​​of pixels inside and outside the evolution curve can be improved, thereby enhancing the reliability of the entire image edge segmentation model and improving the accuracy of part layer height size detection.

[0011] Optionally, in one embodiment of the present invention, the introduction of the iterative penalty term and the variable-scale convolution kernel includes: when calculating the grayscale value fitting values ​​of the internal pixel points and the grayscale value fitting values ​​of the external pixel points of the evolution curve in each iteration, introducing a two-dimensional Gaussian convolution kernel whose convolution kernel size can vary with the size of the iterative penalty term.

[0012] Through the above technical scheme, the embodiment of the present invention can introduce a two-dimensional Gaussian convolution kernel whose convolution kernel size varies with the size of the iterative penalty term, and can dynamically adjust the convolution kernel size according to the iterative penalty term, so that the convolution kernel can better adapt to the feature extraction requirements of the image at different iterative stages, and more accurately process the grayscale value fitting of pixel points inside and outside the evolution curve, thereby improving the accuracy of image edge segmentation, and further improving the accuracy of part layer height size detection.

[0013] Optionally, in one embodiment of the present invention, the calculation formula of the actual layer height is:

[0014] ,

[0015] in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

[0016] Through the above technical scheme, the embodiment of the present invention can take multiple factors into consideration. By incorporating many parameters such as image distance, physical length of pixels on the photosensitive chip, angle of the photosensitive chip side, center coordinates of the camera photosensitive chip, and rotation matrix and offset vector between the camera and world coordinate systems, the two-dimensional image pixel points can be accurately associated with the three-dimensional points in the objective world to calculate the actual layer height size, which helps to accurately obtain the layer height size under different hardware conditions and spatial relationships, and improve the accuracy and applicability of the measurement.

[0017] The second aspect of the present invention provides a device for detecting layer height dimensions of parts for laser wire feeding additive manufacturing, comprising: an acquisition module for acquiring an image to be detected of a part manufactured by laser wire feeding additive manufacturing; a first processing module for denoising the image to be detected to obtain a smooth image, and processing the smooth image to obtain a binary image; a second processing module for acquiring a seed contour located inside the edge of the image of the part manufactured by laser wire feeding additive manufacturing based on the binary image; a detection module for solving a pre-constructed image edge segmentation model, and using the seed contour as an initial evolution curve input to perform image edge segmentation of the part manufactured by laser wire feeding additive manufacturing, so as to calculate the actual layer height dimension corresponding to the image edge.

[0018] Through the above technical solution, the embodiment of the present invention can perform denoising on the image to be detected to obtain a smooth image, remove interference factors, provide reliable basic data for subsequent detection, process the smooth image into a binary image, simplify information, and facilitate the acquisition of seed contours. In terms of part contour detection, obtaining the seed contour based on the binary image can accurately locate the key area, improve the detection targeting, and use the pre-constructed image edge segmentation model and the seed contour as the initial evolution curve to perform image edge segmentation to calculate the layer height size, which can not only adapt to complex shape structures to improve calculation accuracy, but also has good versatility.

[0019] Optionally, in one embodiment of the present invention, the detection module includes: a modeling unit, used to introduce an iterative penalty term and a variable-scale convolution kernel in the energy functional modeling process of the target Chan-Vese algorithm to establish the image edge segmentation model.

[0020] Through the above technical solution, the embodiment of the present invention can introduce iterative penalty terms and variable-scale convolution kernels in the energy functional modeling process of the target Chan-Vese algorithm to establish an image edge segmentation model. The introduction of iterative penalty terms helps to constrain the model solution process and avoid unreasonable solutions or overfitting in the iterative calculation process, thereby improving the accuracy and stability of the model solution. The variable-scale convolution kernel can adapt to the feature extraction requirements of part images at different scales. For images of laser wire additive manufacturing parts that may have different scale features, it can capture image edge information related to layer height size more comprehensively and accurately, thereby improving the accuracy of the entire part layer height size detection, which is more advantageous when processing part images with complex structures and multi-scale features.

[0021] Optionally, in one embodiment of the present invention, the detection module includes: a first introduction unit, used to introduce the iterative penalty term based on the degree of similarity between the pixel value of each pixel and the target value when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration.

[0022] Through the above technical solution, the embodiment of the present invention can introduce an iterative penalty term based on the similarity between the pixel point and the target value in each iterative calculation, and can accurately adjust the calculation process according to the actual situation of the pixel. In this way, the calculation results can be effectively constrained to avoid calculation deviations caused by fluctuations or deviations in pixel values, and the accuracy of fitting the grayscale values ​​of pixels inside and outside the evolution curve can be improved, thereby enhancing the reliability of the entire image edge segmentation model and improving the accuracy of part layer height size detection.

[0023] Optionally, in one embodiment of the present invention, the detection module includes: a second introduction unit, used to introduce a two-dimensional Gaussian convolution kernel whose convolution kernel size can vary with the size of the iterative penalty term when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration.

[0024] Through the above technical scheme, the embodiment of the present invention can introduce a two-dimensional Gaussian convolution kernel whose convolution kernel size varies with the size of the iterative penalty term, and can dynamically adjust the convolution kernel size according to the iterative penalty term, so that the convolution kernel can better adapt to the feature extraction requirements of the image at different iterative stages, and more accurately process the grayscale value fitting of pixel points inside and outside the evolution curve, thereby improving the accuracy of image edge segmentation, and further improving the accuracy of part layer height size detection.

[0025] Optionally, in one embodiment of the present invention, the calculation formula of the actual layer height is:

[0026] ,

[0027] in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, Z is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, f is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

[0028] Through the above technical scheme, the embodiment of the present invention can take multiple factors into consideration. By incorporating many parameters such as image distance, physical length of pixels on the photosensitive chip, angle of the photosensitive chip side, center coordinates of the camera photosensitive chip, and rotation matrix and offset vector between the camera and world coordinate systems, the two-dimensional image pixel points can be accurately associated with the three-dimensional points in the objective world to calculate the actual layer height size, which helps to accurately obtain the layer height size under different hardware conditions and spatial relationships, and improve the accuracy and applicability of the measurement.

[0029] A third aspect of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the part layer height dimension detection method for laser wire feeding additive manufacturing as described in the above embodiment.

[0030] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned method for detecting layer height dimensions of parts used in laser wire additive manufacturing.

[0031] A fifth aspect of the present invention provides a computer program product, including a computer program, which, when executed, is used to implement the above-mentioned method for detecting layer height dimensions of parts used in laser wire additive manufacturing.

[0032] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0034] Figure 1 A flowchart of a method for detecting layer height dimensions of a part for laser wire feeding additive manufacturing according to an embodiment of the present invention;

[0035] Figure 2 It is a flow chart of a method for detecting layer height dimensions of parts manufactured by laser wire feeding additive manufacturing of 316 stainless steel wire according to a specific embodiment of the present invention;

[0036] Figure 3 A schematic structural diagram of a device for detecting layer height dimensions of parts for laser wire feeding additive manufacturing according to an embodiment of the present invention;

[0037] Figure 4The figure is a structural example diagram of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0039] The following describes a method and device for detecting the layer height size of parts for laser wire feeding additive manufacturing according to an embodiment of the present invention with reference to the accompanying drawings. In view of the problem that the part edge positioning method based on the traditional edge detection operator in the related art mentioned in the above background technology has low detection accuracy and weak robustness, which makes it difficult to combine it with the visual ranging solution, the present invention provides a method for detecting the layer height size of parts for laser wire feeding additive manufacturing, in which the image to be detected of the laser wire feeding additive manufacturing part can be obtained by a welding camera, the image to be detected is denoised by a median filter to obtain a smooth image, the smooth image is processed by a threshold segmentation algorithm to obtain a binary image, the binary image is combined with the image morphology processing method, the seed contour located inside the edge of the image of the laser wire feeding additive manufacturing part is obtained, an image edge segmentation model based on the improved Chan-Vese algorithm is established, the image edge segmentation model is solved based on the finite difference method, and the seed contour is used as the initial evolution curve input to realize the image edge segmentation of the laser wire feeding additive manufacturing part, and the actual layer height size corresponding to the image edge is calculated based on the camera calibration algorithm, thereby improving the layer height size detection accuracy of the laser wire feeding additive manufacturing part. This solves the problem in the related art that the part edge positioning method based on the traditional edge detection operator has low detection accuracy and weak robustness, which makes it difficult to combine it with the visual ranging solution.

[0040] Specifically, Figure 1 A schematic flow chart of a method for detecting layer height dimensions of parts used in laser wire additive manufacturing provided in an embodiment of the present invention.

[0041] like Figure 1 As shown, the part layer height size detection method for laser wire feeding additive manufacturing includes the following steps:

[0042] In step S101, an image to be inspected of a laser wire feeding additively manufactured part is obtained.

[0043] It is understandable that the image to be tested refers to the image obtained for the purpose of testing the relevant performance or characteristics of the parts manufactured by laser wire feeding additive manufacturing. It can usually be obtained through image acquisition devices such as industrial cameras. These image acquisition devices need to reasonably set acquisition parameters such as resolution and acquisition angle according to the size, shape, and detection requirements of the parts to ensure that the acquired images can accurately reflect the relevant feature information of the parts and provide a reliable data source for subsequent detection work.

[0044] As a possible implementation method, an embodiment of the present invention uses an active imaging welding camera to complete the collection of images related to laser wire feeding additive manufacturing parts. Specifically, the active imaging welding camera projects a specific light source onto the surface of the laser wire feeding additive manufacturing part. The light source here has certain characteristics, such as a specific wavelength, intensity, and irradiation angle, etc. These characteristics are determined comprehensively based on multiple factors such as the material, surface roughness, shape, and subsequent inspection requirements of the laser wire feeding additive manufacturing part. When light is projected onto the surface of the laser wire feeding additive manufacturing part, reflection will occur. Due to the differences in the microstructure and macroscopic shape of the surface of the laser wire feeding additive manufacturing part, the direction and intensity of the light reflection will also be different. The welding camera is equipped with a corresponding optical acquisition device, which can accurately capture the light reflected back from the surface of the laser wire feeding additive manufacturing part.

[0045] Furthermore, the photoelectric conversion device inside the welding camera converts the received light signal into a digital signal. This conversion process involves complex photoelectric conversion principles. Through specific sensors and circuit designs, the intensity, color and other information of the light can be accurately converted into the numerical value corresponding to the digital signal. Finally, these images in the form of digital signals are uploaded to the computer image processing system. The computer image processing system will further analyze, process and store these digital signal images, such as but not limited to image enhancement, feature extraction, dimensional measurement, etc., in order to perform relevant inspection and evaluation work on laser wire feeding additive manufacturing parts.

[0046] The embodiment of the present invention can collect images of laser wire feeding additively manufactured parts through a welding camera, and convert them into digital signal images and upload them to a computer image processing system, thereby providing a reliable data source for subsequent detection work.

[0047] In step S102, the image to be detected is subjected to denoising to obtain a smoothed image, and the smoothed image is processed to obtain a binary image.

[0048] It is understandable that denoising can reduce or eliminate the noise in the image. The denoising process can improve the quality of the image, making the useful information in the image (such as but not limited to the real contour and features of the part) clearer, which is convenient for further processing and analysis. The smoothed image is the image obtained after the image to be tested is denoised. Its characteristics are that compared with the original image to be tested, the noise is effectively suppressed, the image is visually smoother, and the pixel values ​​in the image are more evenly distributed, which is conducive to highlighting the main features in the image. Binary image is a type of image whose pixel values ​​have only two possible values ​​(usually 0 and 1, or black and white). Binary image can simplify the representation of the image, highlight the difference between the target object in the image (such as but not limited to the contour of the laser wire feeding additive manufacturing part in the image) and the background, and facilitate the subsequent measurement and recognition of specific features of the part (such as shape, size, etc.).

[0049] In the actual implementation process, a median filter is used to perform a smoothing operation on the pixel matrix of the image to be detected to reduce the noise or abnormal pixels in the image to be detected. The calculation method of the median filter satisfies the following formula:

[0050]

[0051] in, is a positive integer greater than or equal to 1, The target pixel and all the pixels in its neighborhood are combined into a region of length Then, sort the pixel values ​​of all pixels in the sequence from small to large, and take the value as the updated pixel value of the target pixel.

[0052] Further, a threshold segmentation algorithm is used to process the smoothed image to obtain a binary image, and a global threshold segmentation algorithm is used, such as but not limited to a histogram technique, an entropy algorithm, and a maximum inter-class variance method. The algorithm and its parameters can be set by those skilled in the art according to actual conditions, and are not specifically limited here.

[0053] The embodiment of the present invention can process the image to be detected by a median filter with a specific calculation method, reduce noise and abnormal pixels and retain important features, improve image quality, and convert the smooth image into a binary image through a global threshold segmentation algorithm. The algorithm and parameters can be flexibly set to increase adaptability. The binary image is convenient for accurate measurement and identification of part features, thereby improving the accuracy and efficiency of detection and evaluation.

[0054] In step S103, a seed contour located inside the edge of the laser wire feeding additive manufacturing part image is obtained according to the binary image.

[0055] It can be understood that the seed contour is a contour structure located inside the edge of the laser wire feeding additive manufacturing part image. It is calculated based on the binary image through a specific algorithm and may represent a structural contour with specific meaning inside the part, such as the contour starting form of a key component or a specific structure inside the part. By obtaining the seed contour, the structural details inside the part can be further studied, providing an important basis for part quality inspection, etc.

[0056] In some embodiments, the binary image is combined with the image morphology processing method, and the obtained binary image is expanded and eroded with a larger scale structure element, so as to obtain the seed contour located inside the edge of the laser wire feeding additive manufacturing part image, and use it to initialize the evolution curve. The specific structure element scale and expansion and erosion algorithm parameters can be set by those skilled in the art according to actual conditions, and are not specifically limited here.

[0057] In the embodiment of the present invention, the structural element scale and the expansion and corrosion algorithm parameters can be set as needed, and can adapt to different laser wire feeding additive manufacturing part image features. The embodiment of the present invention can effectively obtain the seed contour, based on the binary image that highlights the part contour features, and use the expansion and corrosion operation to accurately locate the seed contour inside the edge of the part image, and the obtained seed contour can be used to initialize the evolution curve, which lays a good foundation for in-depth research on the internal structure and shape characteristics of the part, and helps to improve the accuracy of detection and analysis.

[0058] In step S104, the pre-constructed image edge segmentation model is solved, and the seed contour is used as the initial evolution curve input to perform image edge segmentation of the laser wire feeding additive manufacturing part to calculate the actual layer height size corresponding to the image edge.

[0059] It can be understood that the image edge segmentation model is a pre-built mathematical model, which is constructed to accurately segment the edge of the image of the laser wire feeding additive manufacturing part, so as to further obtain important information such as the actual layer height size corresponding to the edge of the part image. The construction of the model may be based on a large amount of sample data, theoretical knowledge of image processing, and a deep understanding of the image characteristics of the laser wire feeding additive manufacturing part.

[0060] Optionally, in one embodiment of the present invention, before solving the pre-constructed image edge segmentation model, it also includes: introducing an iterative penalty term and a variable-scale convolution kernel in the energy functional modeling process of the target Chan-Vese algorithm to establish an image edge segmentation model.

[0061] Specifically, the calculation formula of the energy functional of the Chan-Vese algorithm is as follows:

[0062]

[0063] in, is the evolution curve, The evolution curve The energy functional of The evolution curve Within the Unicom area, The evolution curve Outside the Unicom area, is the side length coefficient, The evolution curve The side length of is the area coefficient, The evolution curve The area inside the connected area, is the internal energy coefficient, is the external energy coefficient, For the input image The gray value of the pixel at The evolution curve Gray value fitting value of internal pixels, The evolution curve Gray value fitting value of external pixels.

[0064] The calculation formula for the gray value fitting value of the internal pixel is:

[0065]

[0066] The calculation formula for the gray value fitting value of the external pixel point is:

[0067]

[0068] in, is the level set function, is the Hyades function.

[0069] The calculation formula of the level set function is:

[0070]

[0071] in, is the current evolution curve, The current evolution curve The connected area enclosed, The current evolution curve Areas outside the enclosed interconnected area, The evolution curve from the pixel point in each area to the current The shortest distance, if point Located on the current evolution curve Internally, the corresponding is a positive value, if the point Located on the current evolution curve Above, the corresponding is 0, if point Located on the current evolution curve Externally, the corresponding Is a negative value.

[0072] The calculation formula of Hy's function is:

[0073]

[0074]

[0075] in, is the Hyatt function, is the first-order derivative of the Hyatt function, is an approximately constant coefficient.

[0076] Optionally, in one embodiment of the present invention, an iterative penalty term and a variable-scale convolution kernel are introduced, including: when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration, an iterative penalty term is introduced based on the similarity between the pixel value of each pixel point and the target value.

[0077] In the actual execution process, each time the grayscale value fitting value of the internal pixel point of the evolution curve and the grayscale value fitting value of the external pixel point are calculated iteratively, the pixel value of each pixel point is considered. and or The similarity between them introduces an iterative penalty term and :

[0078]

[0079]

[0080] in, is the number of iterations, The evolution curve The internal penalty term, The evolution curve The external penalty term, , For the Evolution curve during iterative calculation Gray value fitting value of internal pixels, For the Evolution curve during iterative calculation The gray value fitting value of the external pixel point, at this time and The calculation formula can be expressed as:

[0081]

[0082]

[0083] Optionally, in one embodiment of the present invention, an iterative penalty term and a variable-scale convolution kernel are introduced, including: when calculating the grayscale value fitting values ​​of the internal pixel points and the grayscale value fitting values ​​of the external pixel points of the evolution curve in each iteration, a two-dimensional Gaussian convolution kernel whose convolution kernel size can vary with the size of the iterative penalty term is introduced.

[0084] Specifically, when calculating the grayscale value fitting value of the internal pixel point and the grayscale value fitting value of the external pixel point of the evolution curve in each iteration, the convolution kernel size is introduced to adjust with the iteration penalty term 2D Gaussian convolution kernel with varying size K :

[0085]

[0086]

[0087] in, is the standard deviation of the two-dimensional Gaussian kernel function, is the width of the two-dimensional Gaussian convolution kernel, is the height dimension of the two-dimensional Gaussian convolution kernel. and The calculation formula can be expressed as:

[0088]

[0089]

[0090] Furthermore, the image edge segmentation model is solved based on the finite difference method, and the seed contour is used as the initial evolution curve input to achieve the image edge segmentation of laser wire feeding additive manufacturing parts. To express it in discretization, let is the spatial step length, is the time step, is the number of time steps that have passed, let is the level set function The discrete approximate expression of , the equally divided grids satisfy the following difference format:

[0091]

[0092] in, , is a point in an equally divided grid, and They represent the discrete grids in Forward and backward differences in direction, and They represent the discrete grids in Forward and backward differences in direction.

[0093] Specifically, the process of solving the image edge segmentation model based on the finite difference method is as follows: According to the current level set function value , first calculate the evolution curve at this time Gray value fitting value of internal pixels and evolution curve Gray value fitting value of external pixels , and then iterate according to the following formula until It tends to be stable, and finally realizes the image edge segmentation of laser wire feeding additive manufacturing parts:

[0094]

[0095] in, is the time step, is the first-order derivative of the Hyatt function, is the side length coefficient, is the spatial step length, and They represent the discrete grids in Forward and backward differences in direction, and They represent the discrete grids in Forward and backward differences in direction, is the area coefficient, is the internal energy coefficient, is the external energy coefficient, For the input image The gray value of the pixel at .

[0096] Furthermore, the average pixel distance of the image edge of the laser wire feeding additive manufacturing part in the part height direction is obtained, and the mapping matrix between the image pixel coordinate system and the world coordinate system is constructed based on the camera calibration algorithm. The actual layer height size corresponding to the average pixel distance of the image edge in the part height direction is calculated:

[0097] ,

[0098] in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

[0099] The embodiment of the present invention can build an image edge segmentation model based on a deep understanding of the image features of laser wire additive manufacturing parts, and combine a large amount of sample data and image processing theory knowledge to enable the model to accurately segment the edge of the part image, thereby providing the possibility of obtaining important information such as the actual layer height size corresponding to the edge of the part image. In this way, the actual height of the part in the vertical direction can be more accurately grasped, which is helpful for effectively evaluating the dimensional accuracy of part manufacturing.

[0100] The following specific embodiment will describe in detail the part layer height dimension detection method for laser wire feeding additive manufacturing of the present invention.

[0101] like Figure 2 As shown, the layer height dimension detection method of the 316 stainless steel wire laser wire feeding additive manufacturing part includes the following steps:

[0102] Step S201: Acquire an image to be inspected of a part manufactured by laser wire feeding additively with 316 stainless steel wire.

[0103] An active imaging welding camera is used to project light onto the surface of 316 stainless steel wire additively manufactured parts, collect light signals generated after reflection from the part surface, and convert them into digital signal images through the welding camera photosensitive chip.

[0104] Step S202: filtering and denoising the image to be detected to obtain a smoothed image, using Otsu's Method (maximum inter-class variance method) threshold segmentation algorithm to process the smoothed image to obtain a binary image, and dilating and corroding the binary image to obtain a seed contour.

[0105] For the image to be inspected of a 316 stainless steel wire additively manufactured part in a specific embodiment of the present invention, the smoothing and denoising process can be directly completed by the mask operation of the median filter convolution kernel, and the size of the median filter convolution kernel is 5×5; the binarization process of the smoothed image can be completed by the Otsu threshold segmentation algorithm.

[0106] Furthermore, the obtained binary image can be expanded and eroded by using a larger-scale structure element to obtain a seed contour located inside the edge of the laser wire feeding additive manufacturing part image, and used to initialize the evolution curve. Specifically, in a specific embodiment of the present invention, rectangular structure elements of different scales are used to successively expand and erode the binary image to obtain a seed contour, the size of the expanded convolution kernel is 15×15, and the size of the eroded convolution kernel is 30×30.

[0107] Step S203: Using the seed contour as the initial evolution curve, input it into the image edge segmentation model based on the Chan-Vese algorithm, and solve the image edge segmentation model based on the finite difference method to achieve the image edge segmentation of the laser wire feeding additive manufacturing parts of 316 stainless steel wire.

[0108] A specific embodiment of the present invention can introduce an iterative penalty term and a variable-scale convolution kernel into the energy functional modeling process of the traditional Chan-Vese algorithm. The calculation formula of the energy functional of the traditional Chan-Vese algorithm is as follows:

[0109]

[0110] in, is the evolution curve, The evolution curve The energy functional of The evolution curve Within the Unicom area, The evolution curve Outside the Unicom area, is the side length coefficient, The evolution curve The side length of is the area coefficient, The evolution curve The area inside the connected area, is the internal energy coefficient, is the external energy coefficient, For the input image The gray value of the pixel at The evolution curve Gray value fitting value of internal pixels, The evolution curve The gray value fitting value of the external pixel point. Specifically, in a specific embodiment of the present invention, .

[0111] The calculation formula of the gray value fitting value of the internal pixel can be expressed as:

[0112]

[0113] The calculation formula of the gray value fitting value of the external pixel point can be expressed as:

[0114]

[0115] in, is the level set function, is the Hyades function.

[0116] The calculation formula of the level set function can be expressed as:

[0117]

[0118] in, is the current evolution curve, The current evolution curve The connected area enclosed, The current evolution curve Areas outside the enclosed interconnected area, The evolution curve from the pixel point in each area to the current The shortest distance, if point Located on the current evolution curve Internally, the corresponding is a positive value, if the point Located on the current evolution curve Above, the corresponding is 0, if point Located on the current evolution curve Externally, the corresponding Is a negative value.

[0119] The calculation formula of the Hyatt function can be expressed as:

[0120]

[0121]

[0122] in, is the Hyatt function, is the first-order derivative of the Hyatt function, is an approximate constant coefficient. Specifically, in one embodiment of the present invention, .

[0123] In a specific embodiment of the present invention, an iterative penalty term is introduced in the energy functional modeling process of the traditional Chan-Vese algorithm. When calculating the gray value fitting value of the internal pixel point and the gray value fitting value of the external pixel point of the evolution curve in each iteration, the pixel value of each pixel point is considered. and or The similarity between them introduces an iterative penalty term and :

[0124]

[0125]

[0126] in, is the number of iterations, The evolution curve The internal penalty term, The evolution curve The external penalty term, , For the Evolution curve during iterative calculation Gray value fitting value of internal pixels, For the Evolution curve during iterative calculation The gray value fitting value of the external pixel point, at this time and The calculation formula can be expressed as:

[0127]

[0128]

[0129] Specifically, a specific embodiment of the present invention introduces a variable-scale convolution kernel in the energy functional modeling process of the traditional Chan-Vese algorithm. When calculating the grayscale value fitting value of the internal pixel point and the grayscale value fitting value of the external pixel point of the evolution curve in each iteration, the size of the convolution kernel introduced can be adjusted according to the iteration penalty term. 2D Gaussian convolution kernel with varying size K :

[0130]

[0131]

[0132] in, is the standard deviation of the two-dimensional Gaussian kernel function, is the width of the two-dimensional Gaussian convolution kernel, is the height dimension of the two-dimensional Gaussian convolution kernel, then and The calculation formula can be expressed as:

[0133]

[0134]

[0135] Furthermore, a specific embodiment of the present invention can use an equally divided grid and a finite difference method to calculate the level set function. To express it in discretization, let is the spatial step length, is the time step, is the number of time steps that have passed, let is the level set function The discrete approximate expression of , the equally divided grids satisfy the following difference format:

[0136]

[0137] in, , is a point in an equally divided grid, and They represent the discrete grids in Forward and backward differences in direction, and They represent the discrete grids in Forward and backward differences in direction.

[0138] The process of solving the image edge segmentation model based on the finite difference method is as follows: According to the current level set function value , first calculate the evolution curve at this time Gray value fitting value of internal pixels and evolution curve Gray value fitting value of external pixels , and then iterate according to the following formula until It tends to be stable, and finally realizes the image edge segmentation of laser wire feeding additive manufacturing parts:

[0139]

[0140] in, is the time step, is the first-order derivative of the Hyatt function, is the side length coefficient, is the spatial step length, and They represent the discrete grids in Forward and backward differences in direction, and They represent the discrete grids in Forward and backward differences in direction, is the area coefficient, is the internal energy coefficient, is the external energy coefficient, For the input image The gray value of the pixel at .

[0141] Step S204: Based on the Zhang Zhengyou camera calibration algorithm, the calibrated internal and external parameter matrix of the imaging system is obtained, and the actual layer height size corresponding to the image edge is calculated.

[0142] First, the average pixel distance of the image edge of the laser wire feeding additive manufacturing part in the part height direction is obtained. Then, based on the camera calibration algorithm, a mapping matrix between the image pixel coordinate system and the world coordinate system is constructed to calculate the actual layer height size corresponding to the average pixel distance of the image edge in the part height direction:

[0143] ,

[0144] in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

[0145] According to the part layer height size detection method for laser wire feeding additive manufacturing proposed in an embodiment of the present invention, the image to be detected of the laser wire feeding additive manufacturing part can be obtained through a welding camera, the image to be detected is denoised by a median filter to obtain a smooth image, the smooth image is processed by a threshold segmentation algorithm to obtain a binary image, the binary image is combined with the image morphology processing method to obtain a seed contour located inside the edge of the laser wire feeding additive manufacturing part image, an image edge segmentation model based on an improved Chan-Vese algorithm is established, the image edge segmentation model is solved based on the finite difference method, and the seed contour is used as an initial evolution curve input to realize the image edge segmentation of the laser wire feeding additive manufacturing part, and the actual layer height size corresponding to the image edge is calculated based on the camera calibration algorithm, thereby improving the layer height size detection accuracy of the laser wire feeding additive manufacturing part.

[0146] Next, a device for detecting layer height dimensions of parts for laser wire feeding additive manufacturing according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0147] Figure 3 It is a block diagram of a device for detecting layer height dimensions of parts used in laser wire feeding additive manufacturing according to an embodiment of the present invention.

[0148] like Figure 3 As shown, the part layer height size detection device 10 for laser wire feeding additive manufacturing includes: an acquisition module 100, a first processing module 200, a second processing module 300 and a detection module 400.

[0149] Specifically, the acquisition module 100 is used to acquire the image to be inspected of the laser wire feeding additively manufactured part.

[0150] The first processing module 200 is used to perform denoising on the image to be detected to obtain a smoothed image, and to process the smoothed image to obtain a binary image.

[0151] The second processing module 300 is used to obtain a seed contour located inside the edge of the laser wire feeding additive manufacturing part image according to the binary image.

[0152] The detection module 400 is used to solve the pre-constructed image edge segmentation model and use the seed contour as the initial evolution curve input to perform image edge segmentation of the laser wire feeding additive manufacturing part to calculate the actual layer height size corresponding to the image edge.

[0153] Optionally, in one embodiment of the present invention, the detection module 400 includes: a modeling unit, which is used to introduce an iterative penalty term and a variable-scale convolution kernel in the energy functional modeling process of the target Chan-Vese algorithm to establish an image edge segmentation model.

[0154] Optionally, in one embodiment of the present invention, the detection module 400 includes: a first introduction unit, used to introduce an iterative penalty term based on the degree of similarity between the pixel value of each pixel and the target value when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration.

[0155] Optionally, in one embodiment of the present invention, the detection module 400 includes: a second introduction unit, used to introduce a two-dimensional Gaussian convolution kernel whose convolution kernel size can vary with the size of the iterative penalty term when calculating the grayscale value fitting value of the internal pixel points and the grayscale value fitting value of the external pixel points of the evolution curve in each iteration.

[0156] Optionally, in one embodiment of the present invention, the calculation formula for the actual layer height is:

[0157] ,

[0158] in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

[0159] It should be noted that the aforementioned explanation of the embodiment of the method for detecting the layer height size of parts used for laser wire feeding additive manufacturing is also applicable to the device for detecting the layer height size of parts used for laser wire feeding additive manufacturing of this embodiment, and will not be repeated here.

[0160] According to the part layer height size detection device for laser wire feeding additive manufacturing proposed in an embodiment of the present invention, the image to be detected of the laser wire feeding additive manufacturing part can be obtained through a welding camera, the image to be detected is denoised by a median filter to obtain a smooth image, the smooth image is processed by a threshold segmentation algorithm to obtain a binary image, and the binary image is combined with the image morphology processing method to obtain a seed contour located inside the edge of the laser wire feeding additive manufacturing part image, an image edge segmentation model based on the improved Chan-Vese algorithm is established, the image edge segmentation model is solved based on the finite difference method, and the seed contour is used as the initial evolution curve input to realize the image edge segmentation of the laser wire feeding additive manufacturing part, and the actual layer height size corresponding to the image edge is calculated based on the camera calibration algorithm, thereby improving the layer height size detection accuracy of the laser wire feeding additive manufacturing part.

[0161] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device may include:

[0162] Memory 401 , processor 402 , and a computer program stored in the memory 401 and executable on the processor 402 .

[0163] When the processor 402 executes the program, the part layer height size detection method for laser wire feeding additive manufacturing provided in the above embodiment is implemented.

[0164] Furthermore, the electronic device further comprises:

[0165] The communication interface 403 is used for communication between the memory 401 and the processor 402 .

[0166] The memory 401 is used to store computer programs that can be executed on the processor 402 .

[0167] The memory 401 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0168] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0169] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.

[0170] The processor 402 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0171] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for detecting layer height dimensions of parts used in laser wire feeding additive manufacturing.

[0172] An embodiment of the present invention further provides a computer program product including a computer program, which, when executed, is used to implement the above-mentioned part layer height dimension detection method for laser wire feeding additive manufacturing.

[0173] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0174] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0175] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

[0176] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or N wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0177] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one or a combination of multiple of the following technologies known in the art: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0178] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0179] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0180] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for detecting layer height dimensions of parts used in laser wire feeding additive manufacturing, characterized in that: The following steps are involved: Acquire the image to be inspected of the laser wire feeding additive manufacturing part; Performing denoising on the image to be detected to obtain a smoothed image, and processing the smoothed image to obtain a binary image; Acquire a seed contour located inside the edge of the laser wire feeding additive manufacturing part image according to the binary image; Solving a pre-built image edge segmentation model, and using the seed profile as an initial evolution curve input to perform image edge segmentation of a laser wire feeding additive manufacturing part, so as to calculate an actual layer height size corresponding to the image edge; Before solving the pre-built image edge segmentation model, the method further includes: In the energy functional modeling process of the target Chan-Vese algorithm, an iterative penalty term and a variable-scale convolution kernel are introduced to establish the image edge segmentation model; The introduction of iterative penalty terms and variable-scale convolution kernels includes: When calculating the grayscale value fitting value of the internal pixel point and the grayscale value fitting value of the external pixel point of the evolution curve in each iteration, the iterative penalty term is introduced based on the similarity between the pixel value of each pixel point and the target value; The calculation formulas for the grayscale value fitting value of the internal pixel point and the grayscale value fitting value of the external pixel point can be expressed as: Wherein, φ is the level set function, H(·) is the Hypogean function, and the calculation formulas of the level set function and the Hypogean function can be expressed as follows: in, C ( t ) is the current evolution curve, C ( t ) in The current evolution curve C ( t ) surrounded by the interconnected area, C ( t ) out The current evolution curve C ( t ) outside the connected area, d ( x , y , C ( t )) is the evolution curve from the pixel point in each area to the current C ( t ), if point ( x , y ) is located on the current evolution curve C ( t ) inside, then its corresponding φ ( x , y , t ) is a positive value, if the point ( x , y ) is located on the current evolution curve C ( t ), then its corresponding φ ( x , y , t ) is 0, if point ( x , y ) is located on the current evolution curve C ( t ) outside, then the corresponding φ ( x , y , t ) is a negative value, H ( z ) is the Hyatt function, δ( z ) is the first-order derivative of the Hyatt function, ε is an approximate constant coefficient; Introducing the iterative penalty term and : in, i is the number of iterations, The evolution curve C The internal penalty term, The evolution curve C The external penalty term, .

2. The part layer height dimension detection method for laser wire feeding additive manufacturing according to claim 1 is characterized in that: The introduction of iterative penalty terms and variable-scale convolution kernels includes: When calculating the grayscale value fitting values ​​of the internal pixels and the external pixels of the evolution curve in each iteration, a two-dimensional Gaussian convolution kernel whose size can vary with the size of the iterative penalty term is introduced.

3. The method according to claim 1, characterized in that The calculation formula for the actual layer height is: ， in, is a pixel point on a two-dimensional image. is the corresponding three-dimensional point in the objective world, is the scale factor, is the physical length of a pixel in the column direction on the camera photosensitive chip. is the physical length of a pixel in the row direction on the camera’s photosensitive chip. is the angle between the horizontal and vertical sides of the photosensitive chip, is the coordinate of the center of the camera's photosensitive chip in the pixel coordinate system, is the image distance, To achieve the rotation matrix required for the rigid transformation between the camera coordinate system and the world coordinate system, The offset vector required to achieve the stiff transformation between the camera coordinate system and the world coordinate system.

4. A device for detecting layer height dimensions of parts used in laser wire feeding additive manufacturing, characterized in that: The device is used to implement the part layer height size detection method for laser wire feeding additive manufacturing according to any one of claims 1 to 3, and the device includes: An acquisition module, used for acquiring an image to be inspected of a laser wire feeding additively manufactured part; A first processing module is used to perform denoising on the image to be detected to obtain a smoothed image, and to process the smoothed image to obtain a binary image; A second processing module is used to obtain a seed contour located inside the edge of the laser wire feeding additive manufacturing part image according to the binary image; The detection module is used to solve the pre-constructed image edge segmentation model and use the seed contour as the initial evolution curve input to perform image edge segmentation of the laser wire feeding additive manufacturing part to calculate the actual layer height size corresponding to the image edge.

5. The part layer height dimension detection device for laser wire feeding additive manufacturing according to claim 4 is characterized in that: The detection module comprises: The modeling unit is used to introduce an iterative penalty term and a variable-scale convolution kernel in the energy functional modeling process of the target Chan-Vese algorithm to establish the image edge segmentation model.

6. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the part layer height dimension detection method for laser wire feeding additive manufacturing as described in any one of claims 1 to 3.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the part layer height size detection method for laser wire feeding additive manufacturing as described in any one of claims 1 to 3.

8. A computer program product, comprising a computer program, characterized in that The computer program is executed to implement the part layer height size detection method for laser wire feeding additive manufacturing as described in any one of claims 1 to 3.

Images