Method and system for detecting yarn density of three-dimensional woven preform

Abstract

The application discloses a three-dimensional woven preform yarn density detection method and system, and the method comprises the following steps: acquiring a three-dimensional woven preform image; sequentially performing rotation correction, adaptive gamma enhancement and three-dimensional block matching cooperative filtering to obtain a clear image; realizing yarn segmentation through dynamic threshold segmentation and multi-direction morphological optimization; adopting an improved gray integral projection algorithm, generating an adaptive threshold based on global and local gray projection integral mean, binarizing a projection sequence into a square wave signal; extracting yarn quantity and position information by detecting waveform transition points, and calculating yarn density in combination with a space scale coefficient. The method effectively overcomes image tilt, noise and yarn interweaving interference, converts a traditional projection curve with dramatic fluctuations into a stable square wave signal, significantly improves the robustness and accuracy of yarn identification and positioning, and realizes automatic and high-precision detection of high-density three-dimensional woven preform yarn density.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a method and system for detecting the yarn density of a three-dimensional woven preform. Background Technology

[0002] Three-dimensional woven preforms have attracted widespread attention due to their advantages such as 3D integral structure, multi-directional fiber reinforcement, near-net-shape forming and strong process designability. In particular, the use of this technology to manufacture key components such as engine blades in the aerospace industry plays a crucial role in improving engine performance and ensuring the forming of composite material components.

[0003] However, the existing traditional grayscale integral projection method generates yarn grayscale projection images with large fluctuations, and the yarn information extraction is not accurate enough, thus it cannot accurately measure the yarn density of the preform. Summary of the Invention

[0004] This application provides a method and system for detecting the yarn density of three-dimensional woven preforms, which can effectively solve the problem that traditional methods cannot accurately extract the yarn of preforms. It can significantly improve the quality assessment capability in the industrial production process of three-dimensional woven preforms and meet the actual needs of high-standard industrial production.

[0005] A first aspect of this application provides a method for detecting the yarn density of a three-dimensional woven preform, the method comprising:

[0006] Acquire three-dimensional images of woven prefabricated structures;

[0007] Image enhancement is performed on the three-dimensional woven preform image to obtain a first three-dimensional woven preform image;

[0008] The first three-dimensional woven preform image is subjected to yarn segmentation processing to obtain the second three-dimensional woven preform image;

[0009] An improved gray-scale integral projection algorithm is used to extract information from the second three-dimensional woven preform image to obtain a binarized target direction gray-scale integral projection sequence;

[0010] Based on the binarized target direction grayscale integral projection sequence, determine the quantity information and position information of the woven preform yarns;

[0011] The yarn density of the prefabricated body is calculated based on the yarn quantity information and yarn position information of the woven prefabricated body.

[0012] In one possible implementation, the step of image enhancement of the three-dimensional woven preform image to obtain a first three-dimensional woven preform image includes:

[0013] The Radon transform based on edge detection is used to perform rotation correction on the three-dimensional woven preform image to obtain the rotation correction result;

[0014] Adaptive gamma correction is used to enhance the image of the rotation correction result, resulting in an enhanced image.

[0015] The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain the first three-dimensional woven prefabricated image.

[0016] In one possible implementation, the adaptive gamma correction is expressed as follows:

[0017]

[0018] In the formula, The corrected output brightness value. The brightness value of the input image; These are the Gamma transformation coefficients; For normalization Brightness value; for the median; Gamma correction factor This represents the average value of the target brightness.

[0019] In one possible implementation, the step of denoising the image enhancement result using a three-dimensional block filtering algorithm to obtain a first three-dimensional woven prefabricated image includes:

[0020] The image enhancement result is divided into blocks. For each reference target image block, highly similar image blocks are searched for matching and combined to obtain a three-dimensional array.

[0021] The three-dimensional array is subjected to three-dimensional transformation processing. After reducing image noise by a collaborative hard thresholding filtering method, a three-dimensional inverse transformation is performed to obtain the estimated value of the two-dimensional similar block.

[0022] The weighted average of multiple estimates for each similar block is taken, and the filtering results of each image block are aggregated to obtain the basic estimate of the image;

[0023] In the basic estimation, the positions of similar blocks are found using block matching. Using the positions of the similar blocks, a first three-dimensional array and a second three-dimensional array are obtained from the noisy image and the basic estimation image, respectively.

[0024] The first three-dimensional array and the second three-dimensional array are subjected to three-dimensional transformation processing. The three-dimensional array in the basic estimate is used as the energy spectrum of the real signal. The noisy image is processed by collaborative Wiener filtering using the energy spectrum. Then, a three-dimensional inverse transformation is performed and the image is returned to the original position of the similar block to obtain the final estimated value, which is used as the first three-dimensional woven prefabricated image.

[0025] In one possible implementation, the step of performing yarn segmentation processing on the first three-dimensional woven preform image to obtain a second three-dimensional woven preform image includes:

[0026] The first three-dimensional woven prefabricated image is processed using a dynamic threshold segmentation method to obtain an initial binary image;

[0027] Calculate the pixel gradient weight image based on the first three-dimensional woven preform image;

[0028] Based on the pixel gradient weight image, the initial binary image is subjected to morphological optimization processing to obtain the second three-dimensional woven preform image.

[0029] In one possible implementation, the step of performing morphological optimization processing on the initial binary image based on the pixel gradient weight image to obtain a second three-dimensional woven prefabricated image includes:

[0030] Construct n structuring elements in different directions;

[0031] Based on the n structural elements in different directions and the pixel gradient weight image, the initial binary image is processed using an opening operation method that first performs erosion and then dilation to obtain the second three-dimensional woven prefabricated image.

[0032] In one possible implementation, the step of extracting information from the second three-dimensional woven prefabricated image using an improved gray-scale integral projection algorithm to obtain a binarized target-direction gray-scale integral projection sequence includes:

[0033] Obtain the grayscale integral projection value sequence of the second three-dimensional woven preform image in the target direction;

[0034] The mean of gray-scale integral projection is calculated based on the sequence of gray-scale integral projection values ​​in the target direction to obtain the global mean of gray-scale integral projection.

[0035] For the gray-level integral projection value sequence in the target direction, the local gray-level projection integral mean of each sampling point in the gray-level integral projection value sequence in the target direction is calculated using the sliding window method;

[0036] A dynamic adaptive threshold is generated based on the global grayscale projection integral mean and the local grayscale projection integral mean;

[0037] The projection values ​​of each point in the gray-scale integral projection value sequence in the target direction are compared with the dynamic adaptive threshold to obtain the binarized gray-scale integral projection sequence in the target direction.

[0038] In one possible implementation, a dynamic adaptive threshold is generated based on the global grayscale projection integral mean and the local grayscale projection integral mean, and the calculation formula for the dynamic adaptive threshold is as follows:

[0039]

[0040] In the formula, For dynamic adaptive threshold, The mean of the global grayscale projection integral. The mean of the local grayscale projection integral. These are the global-local weighting coefficients.

[0041] In one possible implementation, determining the quantity information and position information of the woven prefabricated yarns based on the binarized target direction grayscale integral projection sequence includes:

[0042] Waveform transition point detection is performed on the binarized target direction grayscale integral projection sequence to obtain the number of waveform transition points;

[0043] Based on the waveform transition point count information, determine the yarn count information for the woven preform;

[0044] The region between each pair of adjacent rising edge transition points and falling edge transition points is determined as the candidate interval for a single yarn.

[0045] In the second three-dimensional woven preform image, an image region corresponding to each of the candidate intervals is determined;

[0046] Calculate the average grayscale distribution along the target direction in each image region, and determine the index coordinates of the pixel with the largest average grayscale value corresponding to the average grayscale distribution as the position information of the corresponding yarn.

[0047] This example provides a method for detecting yarn density in a three-dimensional woven preform. First, the three-dimensional woven preform image is sequentially subjected to Radon transform rotation correction based on edge detection, adaptive gamma correction enhancement, and three-dimensional block matching collaborative filtering for noise reduction. Then, dynamic threshold segmentation combined with morphological optimization using multi-directional structural elements and pixel gradient weights is employed to achieve accurate separation of the yarn from the background, resulting in a second three-dimensional woven preform image. An improved gray-level integral projection algorithm is used to dynamically generate an adaptive threshold based on the global and local gray-level projection integral mean, binarizing the gray-level integral projection sequence into a projection sequence with significant square wave characteristics. Finally, waveform transitions are applied to the binarized sequence. Point detection and screening accurately extract yarn quantity and location information, and calculate yarn density based on spatial scaling coefficients. Through the above multi-level preprocessing and feature enhancement, the difficulties in yarn extraction caused by image tilt, uneven lighting, and noise interference can be effectively overcome. By introducing a gray-scale integral projection method with dynamic adaptive threshold, the drastically fluctuating projection curve in traditional methods is converted into a stable binary square wave signal, significantly improving the robustness and accuracy of yarn identification and positioning. Through the overall automated detection process, rapid and accurate measurement of yarn density in high-density, non-uniformly arranged three-dimensional woven prefabricated bodies is achieved, providing a reliable technical means for quality assessment and process optimization of aerospace composite material components.

[0048] A second aspect of this application provides a three-dimensional woven preform yarn density detection system, the system comprising:

[0049] The acquisition unit is used to acquire three-dimensional woven prefabricated images;

[0050] The image enhancement unit is used to enhance the image of the three-dimensional woven preform to obtain a first three-dimensional woven preform image;

[0051] The image segmentation unit is used to perform yarn segmentation processing on the first three-dimensional woven preform image to obtain the second three-dimensional woven preform image;

[0052] The image information filtering unit is used to extract information from the second three-dimensional woven preform image using an improved gray-scale integral projection algorithm to obtain a binarized target direction gray-scale integral projection sequence.

[0053] The image information extraction unit is used to determine the quantity information and position information of the woven prefabricated yarns based on the binarized target direction grayscale integral projection sequence.

[0054] The density calculation unit is used to calculate the yarn density of the prefabricated body based on the yarn quantity information and yarn position information of the woven prefabricated body.

[0055] A third aspect of this application provides a terminal including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, and the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the step instructions as described in the three-dimensional woven prefabricated yarn density detection method of the first aspect of this application.

[0056] A fourth aspect of this application provides a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the three-dimensional woven preform yarn density detection method of the first aspect of this application.

[0057] A fifth aspect of this application provides a computer program product, comprising a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the three-dimensional woven prefabricated yarn density detection method of the first aspect of this application. The computer program product may be a software installation package. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this application 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This application provides a schematic diagram of the overall process for a method of detecting the yarn density of a three-dimensional woven preform.

[0060] Figure 2 This application provides a schematic diagram of the overall structure of a three-dimensional woven prefabricated yarn density detection system.

[0061] Figure 3 This application provides a schematic diagram of the structure of a terminal.

[0062] Figure 4 This application provides a schematic flowchart for image enhancement of a three-dimensional woven prefabricated image.

[0063] Figure 5 This application provides a schematic diagram of Radon transform based on edge detection and a schematic diagram of the rotation correction process for embodiments of the present application;

[0064] Figure 6 This application provides a schematic diagram illustrating the effect of an enhanced three-dimensional woven prefabricated image for an embodiment of the present application;

[0065] Figure 7 A flowchart of the BM3D filtering algorithm is provided for embodiments of this application;

[0066] Figure 8 This application provides a schematic diagram of the image effect after image denoising;

[0067] Figure 9 A flowchart illustrating the yarn splitting process is provided for embodiments of this application;

[0068] Figure 10 A schematic diagram showing the comparison effect before and after the improvement of grayscale integral projection is provided for the embodiments of this application;

[0069] Figure 11 This application provides schematic diagrams illustrating the effects of four types of preform yarn extraction in its embodiments;

[0070] Figure label:

[0071] Image acquisition unit-1, image enhancement unit-2, image segmentation unit-3, image information filtering unit-4, image information extraction unit-5, density calculation unit-6. Detailed Implementation

[0072] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0073] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0074] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0075] The three-dimensional yarn density detection method for woven preforms is applied to the three-dimensional yarn density detection method system for woven preforms. Figure 1 A schematic diagram of the overall process for detecting the yarn density of a three-dimensional woven prefabricated structure is shown. Figure 1 As shown, it includes:

[0076] S1. Obtain a three-dimensional woven prefabricated image.

[0077] S2. Perform image enhancement on the three-dimensional woven preform image to obtain a first three-dimensional woven preform image.

[0078] Figure 4 A schematic diagram illustrating an image enhancement process for a 3D woven prefabricated image is shown. Please refer to [link / reference]. Figure 4 As shown, it can be understood that step S2 includes the following sub-steps:

[0079] S201. The Radon transform based on edge detection is used to perform rotation correction on the three-dimensional woven prefabricated image to obtain the rotation correction result.

[0080] Figure 5 A schematic diagram of the Radon transform based on edge detection and a schematic diagram of the rotation correction process are shown below. Figure 5 In this example, the Radon transform based on edge detection is used to correct the tilt angle. The Canny operator formula is as follows:

[0081] (1)

[0082] (2)

[0083] In the formula, For the image in gradient components in the direction, For the image in The gradient component in the direction.

[0084] The gradient magnitude value for each pixel is expressed as follows:

[0085] (3)

[0086] In the formula, This represents the gradient magnitude value.

[0087] Furthermore, based on edge Canny operator detection, the grayscale contour information is linearly integrated in a specified direction, and when the grayscale image... At angle The projection formula generated above is shown in Formula 4. When the maximum projection value is reached, the angle corresponding to the projection distance is... This refers to the tilt angle of the prefabricated image; specifically, see the diagram. At angle The projection formula generated above is shown below:

[0088] (4)

[0089] in, .

[0090] In the formula, grayscale image At angle The projection generated above is based on The axis coordinate parameter is in rows, in angles. A two-dimensional matrix with columns, where the values ​​in the matrix represent the projection results of the image.

[0091] S202. Adaptive gamma correction is used to enhance the image of the rotation correction result to obtain the image enhancement result.

[0092] The formula for adaptive gamma correction is shown below:

[0093] (5)

[0094] In the formula, The corrected output brightness value. The brightness value of the input image; These are the Gamma transformation coefficients; For normalization The brightness value; for the median; Gamma correction factor This represents the average value of the target brightness.

[0095] Furthermore, The calculation formula is as follows:

[0096] (6)

[0097] In the formula, yes The average value.

[0098] Furthermore, The calculation formula is as follows:

[0099] (7)

[0100] In the formula, For pixels in the input image brightness value, It is the coefficient that determines the exponential function.

[0101] The definition is shown in Equation 8:

[0102] (8)

[0103] In the formula, This is the preset upper limit value for the gamma correction coefficient.

[0104] The lower the image intensity, The smaller, The larger the value, the greater the increase in brightness; the higher the image intensity, The larger, The smaller the value, the smaller the increase in brightness.

[0105] S203. The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain the first three-dimensional woven prefabricated image. Figure 6 This diagram illustrates the effect of image enhancement on the 3D woven prefabricated structure. Please refer to [link / reference]. Figure 6 Step S203 includes the following sub-steps:

[0106] S2031. The image enhancement result is divided into blocks. For each reference target image block, image blocks with high similarity are matched and combined to obtain a three-dimensional array.

[0107] S2032. Perform three-dimensional transformation processing on the three-dimensional array, reduce image noise by using a collaborative hard thresholding filtering method, and then perform three-dimensional inverse transformation to obtain the estimated value of the two-dimensional similar block.

[0108] In this example, the BM3D filtering algorithm is used for filtering. Please refer to [link / reference needed]. Figure 7 , Figure 7 A flowchart of the BM3D filtering algorithm is shown.

[0109] S2033. Take a weighted average of the multiple estimates for each similar block, and aggregate the filtering results of each image block to obtain the basic estimate of the image.

[0110] S2034. In the basic estimation, the positions of similar blocks are found using the block matching method. Using the positions of the similar blocks, a first three-dimensional array and a second three-dimensional array are obtained from the noisy image and the basic estimation image, respectively.

[0111] S2035. Perform three-dimensional transformation processing on the first three-dimensional array and the second three-dimensional array. Use the three-dimensional array in the basic estimate as the energy spectrum of the real signal. After using the energy spectrum to perform cooperative Wiener filtering on the noisy image, perform a three-dimensional inverse transformation and return to the original position of the similar block to obtain the final estimated value, which is used as the first three-dimensional woven prefab image. Figure 8 A schematic diagram of the image effect after noise reduction is shown. Please refer to [link / reference]. Figure 8 .

[0112] S3. Perform yarn segmentation processing on the first three-dimensional woven preform image to obtain the second three-dimensional woven preform image.

[0113] Figure 9 A flowchart illustrating the yarn splitting process is shown below. Please refer to [link / reference]. Figure 9 .

[0114] Step S3 includes the following sub-steps:

[0115] S301. The first three-dimensional woven prefabricated image is processed using a dynamic threshold segmentation method to obtain an initial binary image.

[0116] In order to achieve accurate segmentation of the preformed yarn, based on a fixed threshold, the following is referenced: By dynamically adjusting the threshold, the yarn segmentation of the prefabricated structure can be accurately and effectively achieved. The calculation formula for dynamic threshold segmentation is as follows:

[0117] (9)

[0118] In the formula, The input image grayscale value, For the threshold; For Centered on Local mean of the domain; The local grayscale standard deviation; For adjustment coefficients, This is used to balance the influence of local mean and variance on the threshold.

[0119] Furthermore, The calculation formula is as follows:

[0120] (10)

[0121] In the formula, In pixels The mean local gray level is centered on the value; the k-neighborhood radius parameter determines the size of the local window. The size of a local window (e.g.) At that time window); Input image at position The grayscale value at that location.

[0122] Furthermore, The calculation formula is as follows:

[0123] (11)

[0124] In the formula, In pixels The local grayscale standard deviation centered on the value.

[0125] S302. Calculate the pixel gradient weight image based on the first three-dimensional woven preform image.

[0126] Wherein, the pixel gradient weight image The calculation process is as follows:

[0127] First, calculate the first three-dimensional woven preform image. The gradient magnitude can be calculated using operators such as Sobel and Prewitt. Taking the Sobel operator as an example, the gradient magnitude of the image in the horizontal direction is calculated. and vertical direction gradient components and .

[0128] Each pixel gradient magnitude According to the formula Calculation. The larger the gradient magnitude, the more drastic the grayscale change is between the yarn and the background boundary.

[0129] gradient magnitude Mapped to pixel gradient weights In high gradient regions (such as yarn edges) The value should be relatively small; in low gradient regions (such as inside the yarn or against a uniform background). The value should be relatively large to achieve adaptive constraints on the intensity of morphological operations. The specific formula for the mapping is as follows:

[0130] (12)

[0131] In the formula, and These are the minimum and maximum values ​​of the gradient magnitude in the entire image, respectively. Thus, The range of values ​​is It is also negatively correlated with the gradient magnitude.

[0132] S303. Based on the pixel gradient weight image, perform morphological optimization processing on the initial binary image to obtain the second three-dimensional woven preform image.

[0133] Step S303 includes the following sub-steps:

[0134] S3031. Construct n structural elements with different directions.

[0135] To further optimize the binarization segmentation effect, this example uses morphological methods to process the prefabricated fabric image to extract component information that plays a key role in the identification of warp and weft yarns.

[0136] The corrosion process effectively eliminates boundary points in the target area by scanning structural elements, thereby shrinking the boundary inward and filtering out fine noise.

[0137] Dilation processing fills in the holes inside the target region by expanding the boundary outwards. The input image is defined as... The structural element is The conventional formula uses a single structuring element. This example uses a combination of multiple structuring elements and pixel gradient constraints to construct n structuring elements in different directions. .

[0138] S3032. Based on the n structural elements in different directions and the pixel gradient weight image, the initial binary image is processed using an opening operation method that first performs erosion and then dilation to obtain the second three-dimensional woven prefabricated image.

[0139] The formulas related to the increase in the adaptability of yarns in different directions, expansion, and corrosion are shown below:

[0140] (13)

[0141] (14)

[0142] In the formula, Structural elements in different directions (such as linear structural elements at 0°, 45°, etc.); Define pixel step weights. It is used to weaken the expansion of high gradient regions and avoid over-expansion of the blurred boundaries; For image Domain; For the first structural elements Domain; weight Used to weaken the corrosion intensity in high gradient regions.

[0143] In this example, after image segmentation processing of the prefabricated yarn based on the threshold segmentation algorithm and the effect of morphological dilation processing, it can be observed from the image that the horizontal bright stripes on the warp yarns of the prefabricated fabric are clearly separated from the background, which can better prepare for the next step of extracting the warp and weft yarn information of the fabric.

[0144] S4. An improved gray-scale integral projection algorithm is used to extract information from the second three-dimensional woven preform image to obtain a binarized target direction gray-scale integral projection sequence.

[0145] Figure 10 A schematic diagram showing the comparison of grayscale integral projection before and after improvement is provided. Please refer to [link / reference]. Figure 10 This example improves upon the traditional gray-scale integral projection algorithm, addressing issues such as severe fluctuations in the projection curve and difficulties in feature quantization when processing prefabricated yarns.

[0146] Step S4 includes the following sub-steps:

[0147] S401. Obtain the grayscale integral projection value sequence of the second three-dimensional woven preform image in the target direction.

[0148] The second three-dimensional woven preform image is defined as a grayscale image. ,in For row index, For column indexes, and These represent the total number of rows and columns of the image, respectively.

[0149] To extract warp information, the target direction is vertical, and the vertical grayscale integral projection sequence is calculated. :

[0150] (15)

[0151] To extract weft information, the target direction is horizontal, and the horizontal grayscale integral projection sequence is calculated. :

[0152] (16)

[0153] S402. Calculate the mean of gray-scale integral projection based on the gray-scale integral projection value sequence in the target direction to obtain the global mean of gray-scale integral projection.

[0154] The global grayscale projection integral mean The calculation formula is as follows:

[0155] (17)

[0156] In the formula, For projection sequence Length (for vertical projection) For horizontal projection ).

[0157] S403. For the gray-level integral projection value sequence in the target direction, the local gray-level projection integral mean of each sampling point in the gray-level integral projection value sequence in the target direction is calculated using the sliding window method.

[0158] For each position in the sequence Using it as the center, take a size of sliding window Calculate the average value of the projected values ​​within the window, and use it as the local mean for that point:

[0159] (18)

[0160] In the formula, This indicates the number of valid sampling points within the sliding window.

[0161] S404. Generate a dynamic adaptive threshold based on the global grayscale projection integral mean and the local grayscale projection integral mean.

[0162] The formula for calculating the dynamic adaptive threshold is as follows:

[0163] (19)

[0164] In the formula, For dynamic adaptive threshold, The mean of the global grayscale projection integral. The mean of the local grayscale projection integral. These are the global-local weighting coefficients.

[0165] Furthermore, ,in, This refers to the grayscale integral projection value in the horizontal or vertical direction.

[0166] (20)

[0167] (twenty one)

[0168] In the formula, and These are the dynamic adjustment thresholds in the horizontal and vertical directions, respectively. It is the horizontal grayscale integral projection value. Yes, the grayscale integral projection value in the vertical direction; It is the lag threshold, taken as... .

[0169] S405. Compare the projection values ​​of each point in the gray-scale integral projection value sequence in the target direction with the dynamic adaptive threshold to obtain a binarized gray-scale integral projection sequence in the target direction.

[0170] Each position projection value Its corresponding dynamic adaptive threshold Compare and perform binarization:

[0171] (20)

[0172] Thus, the originally continuously fluctuating grayscale integral projection sequence It is converted into a binary sequence consisting of 0s and 1s with significant square wave characteristics. Each consecutive "1" value region in the sequence corresponds to the distribution of a yarn along the projection direction, while the "0" value region corresponds to the background between the yarns. This transformation greatly simplifies the subsequent extraction of the number and position of yarns.

[0173] S5. Based on the binarized target direction grayscale integral projection sequence, determine the quantity information and position information of the woven prefabricated yarn.

[0174] In this example, after the improved grayscale integral projection algorithm is used to detect the yarn in the preform, it is necessary to accurately extract the yarn quantity and position information, and finally use the density calculation formula for measurement.

[0175] Step S5 includes the following sub-steps:

[0176] S501. Detect waveform transition points on the binarized target direction grayscale integral projection sequence to obtain waveform transition point quantity information.

[0177] The calculation formula for waveform transition point detection is as follows:

[0178] (twenty one)

[0179] In the formula, To detect the transition points of the output waveform; It is the difference between waveform transition points. When the condition is met, the waveform transition point can be output; otherwise, the waveform transition point is empty.

[0180] S502. Determine the quantity of yarn in the woven preform based on the waveform transition point quantity information.

[0181] The formula for extracting the number of yarns in the woven preform is as follows:

[0182] (twenty two)

[0183] In the formula, This represents the total number of valid transition points. Indicates a valid transition point; otherwise, it is 0. This is the edge correction value (if the image edge transition points are incomplete). =1, otherwise =0).

[0184] S503. The region between each pair of adjacent rising edge transition points and falling edge transition points is determined as the candidate interval for a single yarn.

[0185] In step S501, the sequence of transition points detected is sorted according to their position. The interval between each pair of adjacent transition points (the preceding one being a rising edge, the following one a falling edge) is then defined as the [number missing]. Candidate intervals for the root yarn in the projection direction.

[0186] S504. In the second three-dimensional woven preform image, determine the image region corresponding to each of the candidate intervals.

[0187] In this process, each candidate interval obtained in step S503 is mapped back to the second three-dimensional woven preform image:

[0188] If the target direction is horizontal (detecting the weft yarn), then the candidate interval corresponds to the row range in the image.

[0189] If the target direction is vertical (detecting warp), then the candidate interval corresponds to the column range in the image.

[0190] S505. Calculate the average gray level distribution along the target direction in each of the image regions, and determine the index coordinates of the pixel with the largest average gray level value corresponding to the average gray level distribution as the position information of the corresponding yarn.

[0191] The formula for extracting the yarn position is as follows:

[0192] (twenty three)

[0193] In the formula, This refers to the transition point region (from the rising edge to the falling edge) corresponding to a single yarn. The column pixel index coordinates are the columns with the largest average grayscale value in this interval.

[0194] S6. Calculate the yarn density of the prefabricated body based on the yarn quantity information and yarn position information of the woven prefabricated body.

[0195] The density calculation formula is shown below:

[0196] (twenty four)

[0197] In the formula, D is the yarn density of the prefabricated body; N is the number of warp and weft yarns; To extract the pixel distance between the number of warp and weft yarns; This is the actual spatial scale factor.

[0198] Figure 11 The diagrams showing the effects of four types of prefabricated yarn extraction are provided. Please refer to [link / reference]. Figure 11 .

[0199] This example provides a method for detecting yarn density in a three-dimensional woven preform. First, the three-dimensional woven preform image is sequentially subjected to Radon transform rotation correction based on edge detection, adaptive gamma correction enhancement, and three-dimensional block matching collaborative filtering for noise reduction. Then, dynamic threshold segmentation combined with morphological optimization using multi-directional structural elements and pixel gradient weights is employed to achieve accurate separation of the yarn from the background, resulting in a second three-dimensional woven preform image. An improved gray-level integral projection algorithm is used to dynamically generate an adaptive threshold based on the global and local gray-level projection integral mean, binarizing the gray-level integral projection sequence into a projection sequence with significant square wave characteristics. Finally, waveform transitions are applied to the binarized sequence. Point detection and screening accurately extract yarn quantity and location information, and calculate yarn density based on spatial scaling coefficients. Through the above multi-level preprocessing and feature enhancement, the difficulties in yarn extraction caused by image tilt, uneven lighting, and noise interference can be effectively overcome. By introducing a gray-scale integral projection method with dynamic adaptive threshold, the drastically fluctuating projection curve in traditional methods is converted into a stable binary square wave signal, significantly improving the robustness and accuracy of yarn identification and positioning. Through the overall automated detection process, rapid and accurate measurement of yarn density in high-density, non-uniformly arranged three-dimensional woven prefabricated bodies is achieved, providing a reliable technical means for quality assessment and process optimization of aerospace composite material components.

[0200] For those consistent with the above, please refer to Figure 2 , Figure 2 This application provides a schematic diagram of the structure of a three-dimensional woven prefabricated yarn density detection system. (See attached diagram.) Figure 2 As shown, the device includes:

[0201] Acquisition unit 1 is used to acquire three-dimensional woven prefabricated body images.

[0202] Image enhancement unit 2 is used to enhance the image of the three-dimensional woven preform to obtain a first three-dimensional woven preform image.

[0203] Image segmentation unit 3 is used to perform yarn segmentation processing on the first three-dimensional woven preform image to obtain a second three-dimensional woven preform image.

[0204] Image information filtering unit 4 is used to extract information from the second three-dimensional woven preform image using an improved gray-scale integral projection algorithm to obtain a binarized target direction gray-scale integral projection sequence.

[0205] Image information extraction unit 5 is used to determine the quantity information and position information of woven prefabricated yarns based on the binarized target direction grayscale integral projection sequence.

[0206] The density calculation unit 6 is used to calculate the yarn density of the prefabricated body based on the yarn quantity information and yarn position information of the woven prefabricated body.

[0207] For examples consistent with the above embodiments, please refer to... Figure 3 , Figure 3 A schematic diagram of a terminal structure provided in an embodiment of this application is shown in the figure. It includes a processor, an input device, an output device, and a memory. The processor, input device, output device, and memory are interconnected. The memory is used to store a computer program, which includes program instructions. The processor is configured to call the program instructions. The program includes instructions for performing the following steps.

[0208] The above mainly describes the solutions of the embodiments of this application from the perspective of the method execution process. It is understood that, in order to achieve the above functions, the terminal includes the corresponding hardware structure and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments provided herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware 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 implementation should not be considered beyond the scope of this application.

[0209] This application embodiment can divide the terminal into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0210] This application also provides a computer storage medium storing a computer program for electronic data exchange, which causes a computer to perform some or all of the steps of any of the three-dimensional woven preform yarn density detection methods described in the above method embodiments.

[0211] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program that causes a computer to perform some or all of the steps of any of the three-dimensional woven preform yarn density detection methods described in the above method embodiments.

[0212] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0213] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0214] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus 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 or other forms.

[0215] 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.

[0216] Furthermore, the functional units in the various embodiments of the application 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. The integrated unit can be implemented in hardware or as a software program module.

[0217] If the integrated unit is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory 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 application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0218] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0219] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for detecting the yarn density of a three-dimensional woven prefabricated body, characterized in that, include: Acquire three-dimensional images of woven prefabricated structures; Image enhancement is performed on the three-dimensional woven preform image to obtain a first three-dimensional woven preform image; The first three-dimensional woven preform image is subjected to yarn segmentation processing to obtain the second three-dimensional woven preform image; An improved gray-scale integral projection algorithm is used to extract information from the second three-dimensional woven preform image to obtain a binarized target direction gray-scale integral projection sequence; Based on the binarized target direction grayscale integral projection sequence, determine the quantity information and position information of the woven preform yarns; The yarn density of the prefabricated body is calculated based on the yarn quantity information and yarn position information of the woven prefabricated body. The step of image enhancement of the three-dimensional woven preform image to obtain a first three-dimensional woven preform image includes: The Radon transform based on edge detection is used to perform rotation correction on the three-dimensional woven preform image to obtain the rotation correction result; Adaptive gamma correction is used to enhance the image of the rotation correction result, resulting in an enhanced image. The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain a first three-dimensional woven preform image. The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain a first three-dimensional woven prefabricated image, including: The image enhancement result is divided into blocks. For each reference target image block, highly similar image blocks are searched for matching and combined to obtain a three-dimensional array. The three-dimensional array is subjected to three-dimensional transformation processing. After reducing image noise by a collaborative hard thresholding filtering method, a three-dimensional inverse transformation is performed to obtain the estimated value of the two-dimensional similar block. The weighted average of multiple estimates for each similar block is taken, and the filtering results of each image block are aggregated to obtain the basic estimate of the image; In the basic estimation, the positions of similar blocks are found using block matching. Using the positions of the similar blocks, a first three-dimensional array and a second three-dimensional array are obtained from the noisy image and the basic estimation image, respectively. The first three-dimensional array and the second three-dimensional array are subjected to three-dimensional transformation processing. The three-dimensional array in the basic estimate is used as the energy spectrum of the real signal. The noisy image is processed by collaborative Wiener filtering using the energy spectrum. Then, a three-dimensional inverse transformation is performed and the image is returned to the original position of the similar block to obtain the final estimated value, which is used as the first three-dimensional woven prefabricated image.

2. The method for detecting yarn density of a three-dimensional woven preform according to claim 1, characterized in that, The formula for the adaptive gamma correction is as follows: ， In the formula, The corrected output brightness value. The brightness value of the input image; These are the Gamma transformation coefficients; For normalization Brightness value; for the median; Gamma correction factor This represents the average value of the target brightness.

3. The method for detecting yarn density of a three-dimensional woven preform according to claim 1, characterized in that, The step of performing yarn segmentation processing on the first three-dimensional woven preform image to obtain the second three-dimensional woven preform image includes: The first three-dimensional woven prefabricated image is processed using a dynamic threshold segmentation method to obtain an initial binary image; Calculate the pixel gradient weight image based on the first three-dimensional woven preform image; Based on the pixel gradient weight image, the initial binary image is subjected to morphological optimization processing to obtain the second three-dimensional woven preform image.

4. The method for detecting yarn density of a three-dimensional woven preform according to claim 3, characterized in that, The step of performing morphological optimization processing on the initial binary image based on the pixel gradient weight image to obtain the second three-dimensional woven prefabricated image includes: Construct n structuring elements in different directions; Based on the n structural elements in different directions and the pixel gradient weight image, the initial binary image is processed using an opening operation method that first performs erosion and then dilation to obtain the second three-dimensional woven prefabricated image.

5. The method for detecting yarn density of a three-dimensional woven preform according to claim 1, characterized in that, The improved gray-scale integral projection algorithm is used to extract information from the second three-dimensional woven prefabricated image to obtain a binarized target direction gray-scale integral projection sequence, including: Obtain the grayscale integral projection value sequence of the second three-dimensional woven preform image in the target direction; The mean of gray-scale integral projection is calculated based on the sequence of gray-scale integral projection values ​​in the target direction to obtain the global mean of gray-scale integral projection. For the gray-level integral projection value sequence in the target direction, the local gray-level projection integral mean of each sampling point in the gray-level integral projection value sequence in the target direction is calculated using the sliding window method; A dynamic adaptive threshold is generated based on the global grayscale projection integral mean and the local grayscale projection integral mean; The projection values ​​of each point in the gray-scale integral projection value sequence in the target direction are compared with the dynamic adaptive threshold to obtain the binarized gray-scale integral projection sequence in the target direction.

6. The method for detecting yarn density of a three-dimensional woven preform according to claim 5, characterized in that, The dynamic adaptive threshold is generated based on the global grayscale projection integral mean and the local grayscale projection integral mean. The calculation formula for the dynamic adaptive threshold is as follows: ， In the formula, For dynamic adaptive threshold, The mean of the global grayscale projection integral. The mean of the local grayscale projection integral. These are the global-local weighting coefficients.

7. The method for detecting yarn density of a three-dimensional woven preform according to claim 1, characterized in that, The step of determining the quantity information and position information of the woven prefabricated yarns based on the binarized target direction grayscale integral projection sequence includes: Waveform transition point detection is performed on the binarized target direction grayscale integral projection sequence to obtain the number of waveform transition points; Based on the waveform transition point count information, determine the yarn count information for the woven preform; The region between each pair of adjacent rising edge transition points and falling edge transition points is determined as the candidate interval for a single yarn. In the second three-dimensional woven preform image, an image region corresponding to each of the candidate intervals is determined; Calculate the average grayscale distribution along the target direction in each image region, and determine the index coordinates of the pixel with the largest average grayscale value corresponding to the average grayscale distribution as the position information of the corresponding yarn.

8. A three-dimensional woven prefabricated yarn density detection system, characterized in that, include: The acquisition unit is used to acquire three-dimensional woven prefabricated images; The image enhancement unit is used to enhance the image of the three-dimensional woven preform to obtain a first three-dimensional woven preform image; The image segmentation unit is used to perform yarn segmentation processing on the first three-dimensional woven preform image to obtain the second three-dimensional woven preform image; The image information filtering unit is used to extract information from the second three-dimensional woven preform image using an improved gray-scale integral projection algorithm to obtain a binarized target direction gray-scale integral projection sequence. The image information extraction unit is used to determine the quantity information and position information of the woven prefabricated yarns based on the binarized target direction grayscale integral projection sequence. The density calculation unit is used to calculate the yarn density of the prefabricated body based on the yarn quantity information and yarn position information of the woven prefabricated body; The step of image enhancement of the three-dimensional woven preform image to obtain a first three-dimensional woven preform image includes: The Radon transform based on edge detection is used to perform rotation correction on the three-dimensional woven preform image to obtain the rotation correction result; Adaptive gamma correction is used to enhance the image of the rotation correction result, resulting in an enhanced image. The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain a first three-dimensional woven preform image. The image enhancement result is denoised using a three-dimensional block filtering algorithm to obtain a first three-dimensional woven prefabricated image, including: The image enhancement result is divided into blocks. For each reference target image block, highly similar image blocks are searched for matching and combined to obtain a three-dimensional array. The three-dimensional array is subjected to three-dimensional transformation processing. After reducing image noise by a collaborative hard thresholding filtering method, a three-dimensional inverse transformation is performed to obtain the estimated value of the two-dimensional similar block. The weighted average of multiple estimates for each similar block is taken, and the filtering results of each image block are aggregated to obtain the basic estimate of the image; In the basic estimation, the positions of similar blocks are found using block matching. Using the positions of the similar blocks, a first three-dimensional array and a second three-dimensional array are obtained from the noisy image and the basic estimation image, respectively. The first three-dimensional array and the second three-dimensional array are subjected to three-dimensional transformation processing. The three-dimensional array in the basic estimate is used as the energy spectrum of the real signal. The noisy image is processed by collaborative Wiener filtering using the energy spectrum. Then, a three-dimensional inverse transformation is performed and the image is returned to the original position of the similar block to obtain the final estimated value, which is used as the first three-dimensional woven prefabricated image.

Images