Method for synchronously detecting orientation and surface density of regenerated short carbon fiber material

Abstract

The invention relates to the field of carbon fiber felt detection, and particularly discloses a regenerated short carbon fiber material oriented orientation and surface density synchronous detection method, which comprises the following steps: S1, image acquisition and preprocessing: improving image quality through brightness channel enhancement and a CLAHE algorithm, and enhancing a carbon fiber boundary contrast; s2, carbon fiber contour detection: screening effective carbon fiber contours by using Canny edge detection and contour analysis; s3, calculating an orientation angle: automatically extracting the orientation of the carbon fiber by using a contour fitting algorithm, calculating an angle deviation and generating a color coding image; s4, surface density analysis: dividing an image region in combination with local gray change recursion, and subdividing a density change salient region as required; and S5, result output and visualization: constructing a carbon fiber orientation angle distribution histogram and a surface density gray thermodynamic diagram, and extracting a variable coefficient and an extreme value ratio key index. According to the invention, rapid and accurate detection of the regenerated chopped carbon fiber felt structure is realized.

Description

Technical Field

[0001] This invention relates to the field of carbon fiber felt testing technology, specifically a method for simultaneously detecting the orientation and areal density of recycled short-cut carbon fiber materials. Background Technology

[0002] With the widespread application of composite materials in aerospace, automotive manufacturing, and other fields, the demand for carbon fiber felt as a reinforcing material is increasing. The orientation degree and areal density of carbon fiber felt are key parameters affecting the performance of composite materials, and accurate measurement of these parameters is crucial for product quality control. Traditional testing methods are highly destructive, cannot achieve online monitoring during the production process, and have long testing cycles and high costs. Existing non-destructive testing technologies have significant limitations. While beta-ray transmission can achieve non-destructive testing, it poses radiation safety hazards and has high equipment investment costs. Although the gravimetric method for measuring areal density is simple, it only obtains the overall areal density and cannot reflect the local density distribution.

[0003] Based on this, this study proposes a method for simultaneous detection of orientation and areal density of regenerated chopped carbon fiber materials, which can achieve non-destructive, rapid and accurate detection of regenerated chopped carbon fiber felt structures. Summary of the Invention

[0004] To address the aforementioned problems in existing technologies, this invention provides a method for the simultaneous detection of orientation and areal density of recycled chopped carbon fiber materials. This method effectively solves the problems of high destructiveness, limited indicators, poor real-time performance, and high cost of existing carbon fiber felt detection technologies, enabling non-destructive, rapid, and accurate detection of recycled chopped carbon fiber felt structures.

[0005] To achieve the above objectives, this invention proposes a method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials, comprising: S1. Image Acquisition and Preprocessing: Color images of recycled chopped carbon fiber felt are acquired using an industrial camera. The luminance component of the acquired color images of recycled chopped carbon fiber felt is extracted through ROI cropping and YCrCb color space conversion. The CLAHE algorithm is then used to enhance the contrast and eliminate the effects of uneven lighting. S2. Carbon fiber contour detection: The preprocessed image is divided into blocks, and Canny edge detection and contour analysis are used to screen valid carbon fiber contours. The aspect ratio and length threshold are combined to eliminate interference. S3. Orientation angle calculation: Perform least squares straight line fitting on the screened carbon fiber profile, calculate the deviation angle between the result of least squares straight line fitting on the screened carbon fiber profile and the reference direction, and visualize the orientation distribution through color coding. S4. Areal density analysis: The image processed by S1 is subjected to a dynamic block algorithm. The grid size is automatically adjusted according to the gray-level gradient characteristics. The local gray-level density is calculated for each grid cell. The mean, standard deviation, coefficient of variation and extreme value ratio of areal density are statistically analyzed. The local gray-level density calculated for each grid cell is visualized as a heat map. S5. Results Output and Visualization: Comprehensive output of carbon fiber average orientation deviation, angle distribution histogram and areal density thermogram.

[0006] Preferably, in S1, the specific steps of image preprocessing include: S11. Acquire images of the recycled chopped carbon fiber felt to obtain the original color image, and then crop the set cropping area through ROI. S12. Using the Y channel, convert the image after cropping the ROI-defined cropped area to the YCrCb color space. The calculation formula is as follows: Y = 0.299R + 0.587G + 0.114B; S13. The CLAHE algorithm is used to enhance the image. After the image is converted to the YCrCb color space, it is divided into M×N sub-regions. The histogram of each sub-region is cropped. Pixels that exceed the histogram cropping threshold are evenly distributed to other gray levels. Bilinear interpolation is used to eliminate block boundary artifacts. S14. Calculate the variance of the Y channel and determine whether there is uneven illumination in the image; S15. Scale the image according to the target resolution, and use bilinear interpolation to unify the input scale. The calculation formula for bilinear interpolation is as follows: ; In the formula, The weights are determined by the bilinear relationship between pixel positions.

[0007] Preferably, in S14, if the variance calculation result of the Y channel is greater than 300, then the image has uneven illumination. The processing steps to eliminate the influence of non-uniform illumination include: S141. Perform a logarithmic transformation on the image after CLAHE processing: ; S142. Use a Gaussian low-pass filter to estimate the illumination components: ; S143. Calculate the reflection component and apply the gain factor: ; S144. Map the result to the [0,255] interval by normalizing the extreme values.

[0008] Preferably, in S2, the specific steps for carbon fiber contour detection include: S21. Divide the preprocessed grayscale image into multiple square detection regions, with the block size set to 100×100 pixels, and use a sliding window method with a 50% overlap rate to traverse the entire image. S22. Perform Canny edge detection on each block, call the contour retrieval mode RETR_EXTERNAL in OPENCV to obtain the outermost contour, and use the contour approximation method CHAIN_APPROX_SIMPLE to compress horizontal, vertical and diagonal line segments. S23. Perform length filtering, retaining outlines with a length between 30 and 200 pixels, and discarding outlines of other lengths; S24. Perform shape screening, calculate the aspect ratio of the minimum bounding rectangle, and remove non-fibrous contours with an aspect ratio less than 2.

[0009] Preferably, in S3, the specific processing steps for calculating the orientation angle are as follows: S31. Fit the linear equation using the least squares method for each valid contour and calculate the direction vector. ; S32. Perform angle conversion and limit the statistical angles to [-90°, 90°]. The conversion formula is: ; S33. Map 0° to green (0,255,0), ±90° to red (255,0,0), use linear interpolation for intermediate angles, and draw direction indicator lines on the image; S34. For each angle θ, calculate the minimum deviation from the reference direction (90° / -90°), and calculate the average deviation and the percentage of orientation deviation. The calculation formula is as follows: ; S35. Generate a detection image with orientation markers, output an angle distribution histogram, and record the average deviation and deviation percentage.

[0010] Preferably, in S4, the areal density analysis is performed on the areal density of the carbon fiber felt sample image, and an adaptive region partitioning algorithm based on quadtree decomposition is used for accurate quantitative characterization.

[0011] Preferably, in S4, the specific processing steps for areal density analysis include: S41. Establish a two-dimensional coordinate system with the upper left corner of the carbon fiber felt sample image as the origin, and set the initial analysis area to the complete image range. S42. Use a recursive algorithm to perform intelligent region division, calculate the gray-level density statistics of each region, and when the coefficient of variation CV=σ / μ exceeds a set threshold, divide the current region into four equal parts until the uniformity condition is met or the minimum block size is reached; wherein, the gray-level density statistics of each region include the mean. Standard deviation ; S43. Use the area-weighted algorithm to calculate the key statistical parameters of gray density in each region; S44. Construct a discrete density field based on the partitioning results, and generate a continuous density distribution surface through bilinear interpolation; where the white to black gradient corresponds to the normalized surface density value of 0-1. S45. Use real experimental data to compare with the normalized areal density to obtain the true areal density.

[0012] Preferably, in S43, the key statistical parameters of gray density in each region include weighted average density, weighted standard deviation, coefficient of variation, and extreme value ratio.

[0013] Preferably, the weighted average density The weighted standard deviation reflects the overall quality distribution. The coefficient of variation characterizes the intensity of density fluctuations. Quantifying material uniformity, the extreme value ratio Identify local defects.

[0014] Therefore, this invention proposes a method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials, the advantages of which are as follows: (1) The orientation degree and areal density synchronous detection technology of the present invention effectively solves the problems of strong destructiveness, single index, poor real-time performance and high cost of existing carbon fiber felt detection technology, and realizes non-destructive, rapid and accurate detection of recycled short carbon fiber felt structure.

[0015] (2) This invention breaks through the limitations of traditional methods, which require sample destruction, have long detection cycles, and rely on expensive equipment. It significantly improves the speed and robustness of structural information extraction, and provides an efficient and reliable feature extraction scheme for composite material molding quality assessment, data modeling and structural prediction, meeting the needs of the composite material manufacturing field for economical, efficient and multi-dimensional quality inspection.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 A flowchart of a method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials provided by the present invention; Figure 2The image shows the original image of the recycled carbon fiber felt collected in Example 1. Figure 3 This is an image of the pre-processed recycled carbon fiber felt from Example 1; Figure 4 This is a visualization image of the carbon fiber contour extraction in Example 1; Figure 5 This is a histogram of carbon fiber orientation distribution in Example 1; Figure 6 This is a thermogram of the carbon fiber surface density distribution in Example 1. Detailed Implementation

[0018] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of this application.

[0019] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0020] like Figure 1 As shown, the present invention provides a method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials, comprising three main steps: image preprocessing, carbon fiber orientation detection, and areal density detection, specifically including: S1. Image Acquisition and Preprocessing: Color images of recycled chopped carbon fiber felt are acquired using an industrial camera. The luminance component of the acquired color images of recycled chopped carbon fiber felt is extracted through ROI cropping and YCrCb color space conversion. The CLAHE algorithm is then used to enhance the contrast and eliminate the effects of uneven lighting. S11. Acquire images of the recycled chopped carbon fiber felt to obtain the original color image, and then crop the set cropping area through ROI. S12. Using the Y channel, convert the image after cropping the ROI-defined cropped area to the YCrCb color space. The calculation formula is as follows: Y = 0.299R + 0.587G + 0.114B; S13. The CLAHE algorithm is used to enhance the image. After the image is converted to the YCrCb color space, it is divided into M×N sub-regions. The histogram of each sub-region is cropped. Pixels that exceed the histogram cropping threshold are evenly distributed to other gray levels. Bilinear interpolation is used to eliminate block boundary artifacts. S14. Calculate the variance of the Y channel to determine if uneven illumination exists in the image; wherein, if the mean global variance of the Y channel is greater than 300, uneven illumination exists in the image. The processing steps to eliminate the influence of non-uniform illumination include: S141. Perform a logarithmic transformation on the image after CLAHE processing: ; S142. Use a Gaussian low-pass filter to estimate the illumination components: ; S143. Calculate the reflection component and apply the gain factor: ; S144. Map the result to the [0,255] interval by normalizing the extreme values.

[0021] S15. Scale the image according to the target resolution, and use bilinear interpolation to unify the input scale. The calculation formula for bilinear interpolation is as follows: ; In the formula, The weights are determined by the bilinear relationship between pixel positions.

[0022] S2. Carbon fiber contour detection: The preprocessed image is divided into blocks, and Canny edge detection and contour analysis are used to screen valid carbon fiber contours. The aspect ratio and length threshold are combined to eliminate interference. S21. Divide the preprocessed grayscale image into multiple square detection regions, with the block size set to 100×100 pixels, and use a sliding window method with a 50% overlap rate to traverse the entire image. S22. Perform Canny edge detection on each block, call the contour retrieval mode RETR_EXTERNAL in OPENCV to obtain the outermost contour, and use the contour approximation method CHAIN_APPROX_SIMPLE to compress horizontal, vertical and diagonal line segments. S23. Perform length filtering, retaining outlines with a length between 30 and 200 pixels, and discarding outlines of other lengths; S24. Perform shape screening, calculate the aspect ratio of the minimum bounding rectangle, and remove non-fibrous contours with an aspect ratio less than 2.

[0023] S3. Orientation Angle Calculation: The least-squares straight-line fit is performed on the screened carbon fiber profile. The deviation angle between the least-squares straight-line fit result and the reference direction is calculated. The orientation distribution is visualized through color coding. The specific steps for calculating the orientation angle are as follows: S31. Fit the linear equation using the least squares method for each valid contour and calculate the direction vector. ; S32. Perform angle conversion and limit the statistical angles to [-90°, 90°]. The conversion formula is: ; S33. Map 0° to green (0,255,0), ±90° to red (255,0,0), use linear interpolation for intermediate angles, and draw direction indicator lines on the image; S34. For each angle θ, calculate the minimum deviation from the reference direction (90° / -90°), and calculate the average deviation and the percentage of orientation deviation. The calculation formula is as follows: ; S35. Generate a detection image with orientation markers, output an angle distribution histogram, and record the average deviation and deviation percentage.

[0024] S4. Areal density analysis: The image processed by S1 is subjected to a dynamic block algorithm. The grid size is automatically adjusted according to the gray-level gradient characteristics. The local gray-level density is calculated for each grid cell. The mean, standard deviation, coefficient of variation and extreme value ratio of areal density are statistically analyzed. The local gray-level density calculated for each grid cell is visualized as a heat map. In S4, the areal density analysis involves analyzing the areal density of the carbon fiber felt sample image and employing an adaptive region partitioning algorithm based on quadtree decomposition for accurate quantification. The specific processing steps for the areal density analysis include: S41. Establish a two-dimensional coordinate system with the upper left corner of the carbon fiber felt sample image as the origin, and set the initial analysis area to the complete image range. S42. Use a recursive algorithm to perform intelligent region division, calculate the gray-level density statistics of each region, and when the coefficient of variation CV=σ / μ exceeds a set threshold, divide the current region into four equal parts until the uniformity condition is met or the minimum block size is reached; wherein, the gray-level density statistics of each region include the mean. Standard deviation ; S43. Calculate the key statistical parameters of gray density in each region using an area-weighted algorithm; the key statistical parameters of gray density in each region include weighted average density, weighted standard deviation, coefficient of variation, and extreme value ratio. S44. Construct a discrete density field based on the partitioning results, and generate a continuous density distribution surface through bilinear interpolation; where the white-to-black gradient corresponds to the normalized surface density value of 0-1; weighted average density... The weighted standard deviation reflects the overall quality distribution. The coefficient of variation characterizes the intensity of density fluctuations. Quantifying material uniformity, the extreme value ratio Identify local defects.

[0025] S45. Use real experimental data to compare with the normalized areal density to obtain the true areal density.

[0026] S5. Results Output and Visualization: Comprehensive output of carbon fiber average orientation deviation, angle distribution histogram and areal density thermogram.

[0027] like Figures 2-6 As shown, according to the present invention, an embodiment 1 of a method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials is provided below.

[0028] The test samples were made of T300 carbon fiber felt, such as Figure 2 The original image was captured using a high-resolution industrial camera (50 megapixels).

[0029] First, the acquired images are preprocessed. A region of 4000×3000 pixels (height × width) is cropped using the Region of Interest (ROI). Then, the luminance channel is separated using the YCrCb color space to avoid chromaticity interference. Adaptive histogram equalization (CLAHE) is applied to enhance the edges of low-contrast carbon fibers, with parameters set to an 8×8 grid size and a contrast limit of 10. Finally, image downsampling is performed to achieve a uniform input scale, balancing accuracy and computational efficiency. Figure 3 This is the image after preprocessing.

[0030] Then, multi-scale carbon fiber detection is performed. First, the image is divided into 100×100 pixel detection units. The Canny operator is used to extract the carbon fiber contours, with dual thresholds set to (50, 150). Next, valid contours are selected based on length, with an effective length range of 30-200 pixels. Least-squares line fitting is performed on each valid contour to calculate its angle with the vertical baseline (-90° + 90°). Finally, visualization mapping is performed, using a red (±90°) to green (0°) gradient to mark the carbon fiber direction based on the absolute value of the angle. Figure 4 A visualization of carbon fiber. Figure 5 This is a histogram of carbon fiber orientation distribution.

[0031] Next, adaptive density analysis is performed. The post-processed image is initially divided into 16 equally sized sub-regions. A grayscale standard value is calculated for each sub-region and normalized to the [0,1] interval. If the value exceeds a set threshold, further subdivision into four sub-regions is performed. Segmentation stops when the block size is less than or equal to 100 pixels or the standard deviation is less than the threshold. The average grayscale value of each block and the overall average grayscale value are then calculated. The areal density of the carbon fiber is derived from the average grayscale value of the blocks, and the areal density standard deviation, coefficient of variation, and extreme value ratio are statistically analyzed.

[0032] Finally, the calculated carbon fiber areal density is obtained, and the output is as follows: Figure 6 The thermogram of carbon fiber areal density is shown. The orientation degree of the sample, measured by this method, is 79.86%, and the average areal density is 10.15 g / m³. 2 The standard deviation was 1.21, the coefficient of variation was 0.169, and the extreme value ratio was 2.55.

[0033] Therefore, this invention provides a method for simultaneous detection of orientation and areal density of recycled chopped carbon fiber materials, effectively solving the problems of strong destructiveness, single index, poor real-time performance, and high cost of existing carbon fiber felt detection technology. It realizes non-destructive, rapid, and accurate detection of recycled chopped carbon fiber felt structures, significantly improving the speed and robustness of structural information extraction. It provides an efficient and reliable feature extraction scheme for composite material molding quality assessment, data modeling, and structural prediction, meeting the needs of the composite material manufacturing field for economical, efficient, and multi-dimensional quality inspection.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for simultaneously detecting orientation and areal density of recycled short-cut carbon fiber materials, characterized in that, include: S1. Image Acquisition and Preprocessing: Color images of recycled chopped carbon fiber felt are acquired using an industrial camera. The luminance component of the acquired color images of recycled chopped carbon fiber felt is extracted through ROI cropping and YCrCb color space conversion. The CLAHE algorithm is then used to enhance the contrast and eliminate the effects of uneven lighting. S2. Carbon fiber contour detection: The preprocessed image is divided into blocks, and Canny edge detection and contour analysis are used to screen valid carbon fiber contours. The aspect ratio and length threshold are combined to eliminate interference. S3. Orientation angle calculation: Perform least squares straight line fitting on the screened carbon fiber profile, calculate the deviation angle between the result of least squares straight line fitting on the screened carbon fiber profile and the reference direction, and visualize the orientation distribution through color coding. S4. Areal density analysis: The image processed by S1 is subjected to a dynamic block algorithm. The grid size is automatically adjusted according to the gray-level gradient characteristics. The local gray-level density is calculated for each grid cell. The mean, standard deviation, coefficient of variation and extreme value ratio of areal density are statistically analyzed. The local gray-level density calculated for each grid cell is visualized as a heat map. S5. Results Output and Visualization: Comprehensive output of carbon fiber average orientation deviation, angle distribution histogram and areal density thermogram.

2. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 1, characterized in that, In S1, the specific steps of image preprocessing include: S11. Acquire images of the recycled chopped carbon fiber felt to obtain the original color image, and then crop the set cropping area through ROI. S12. Using the Y channel, convert the image after cropping the ROI-defined cropped area to the YCrCb color space. The calculation formula is as follows: Y = 0.299R + 0.587G + 0.114B; S13. The CLAHE algorithm is used to enhance the image. After the image is converted to the YCrCb color space, it is divided into M×N sub-regions. The histogram of each sub-region is cropped. Pixels that exceed the histogram cropping threshold are evenly distributed to other gray levels. Bilinear interpolation is used to eliminate block boundary artifacts. S14. Calculate the variance of the Y channel and determine whether there is uneven illumination in the image; S15. Scale the image according to the target resolution, and use bilinear interpolation to unify the input scale. The calculation formula for bilinear interpolation is as follows: ； In the formula, The weights are determined by the bilinear relationship between pixel positions.

3. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 2, characterized in that, In S14, if the variance calculation result of the Y channel is greater than 300, then the image has uneven illumination. The processing steps to eliminate the influence of non-uniform illumination include: S141. Perform a logarithmic transformation on the image after CLAHE processing: ； S142. Use a Gaussian low-pass filter to estimate the illumination components: ； S143. Calculate the reflection component and apply the gain factor: ； S144. Map the result to the [0,255] interval by normalizing the extreme values.

4. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 1, characterized in that, In S2, the specific steps for carbon fiber contour detection include: S21. Divide the preprocessed grayscale image into multiple square detection regions, with the block size set to 100×100 pixels, and use a sliding window method with a 50% overlap rate to traverse the entire image. S22. Perform Canny edge detection on each block, call the contour retrieval mode RETR_EXTERNAL in OPENCV to obtain the outermost contour, and use the contour approximation method CHAIN_APPROX_SIMPLE to compress horizontal, vertical and diagonal line segments. S23. Perform length filtering, retaining outlines with a length between 30 and 200 pixels, and discarding outlines of other lengths; S24. Perform shape screening, calculate the aspect ratio of the minimum bounding rectangle, and remove non-fibrous contours with an aspect ratio less than 2.

5. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 1, characterized in that, In S3, the specific processing steps for calculating the orientation angle are as follows: S31. Fit the linear equation using the least squares method for each valid contour and calculate the direction vector. ; S32. Perform angle conversion and limit the statistical angles to [-90°, 90°]. The conversion formula is: ； S33. Map 0° to green (0,255,0), ±90° to red (255,0,0), use linear interpolation for intermediate angles, and draw direction indicator lines on the image; S34. For each angle θ, calculate the minimum deviation from the reference direction (90° / -90°), and calculate the average deviation and the percentage of orientation deviation. The calculation formula is as follows: ； S35. Generate a detection image with orientation markers, output an angle distribution histogram, and record the average deviation and deviation percentage.

6. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 1, characterized in that, In S4, the areal density analysis is performed on the areal density of the carbon fiber felt sample image, and an adaptive region partitioning algorithm based on quadtree decomposition is used for accurate quantitative characterization.

7. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 6, characterized in that, In S4, the specific processing steps for areal density analysis include: S41. Establish a two-dimensional coordinate system with the upper left corner of the carbon fiber felt sample image as the origin, and set the initial analysis area to the complete image range. S42. Use a recursive algorithm to perform intelligent region division, calculate the gray-level density statistics of each region, and when the coefficient of variation CV=σ / μ exceeds a set threshold, divide the current region into four equal parts until the uniformity condition is met or the minimum block size is reached; wherein, the gray-level density statistics of each region include the mean. Standard deviation ; S43. Use the area-weighted algorithm to calculate the key statistical parameters of gray density in each region; S44. Construct a discrete density field based on the partitioning results, and generate a continuous density distribution surface through bilinear interpolation; where the white to black gradient corresponds to the normalized surface density value of 0-1. S45. Use real experimental data to compare with the normalized areal density to obtain the true areal density.

8. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 7, characterized in that, In S43, the key statistical parameters of gray density in each region include weighted average density, weighted standard deviation, coefficient of variation, and extreme value ratio.

9. The method for simultaneous detection of orientation and areal density of recycled short-cut carbon fiber materials according to claim 8, characterized in that, The weighted average density The weighted standard deviation reflects the overall quality distribution. The coefficient of variation characterizes the intensity of density fluctuations. Quantifying material uniformity, the extreme value ratio Identify local defects.

Images