Spooling color tube detection method and system based on image analysis

Abstract

The invention relates to the technical field of image analysis, in particular to a spooling color tube detection method and system based on image analysis, and the method comprises the following steps: carrying out the statistics of color component intensity distribution based on a captured bobbin yarn conveying track image of a spooling machine, comparing the difference with a standard illumination color image, and carrying out the proportion adjustment of color components; and obtaining an illumination color correction image. In the invention, by separating and adjusting three color components of red, green and blue in the image, eliminating the influence of illumination conditions, and by calculating the pixel brightness gradient value and correcting the shadow area, the shadow interference in the image is reduced, the track background and the yarn area are accurately distinguished, and the color deviation and uniformity of the ring bobbin are accurately evaluated; and the color deviation tolerance threshold is dynamically adjusted according to the statistical result, so that the flexibility and accuracy of color anomaly detection are improved, and the recognition efficiency and the quality control capability of color anomaly are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image analysis, and particularly to a method and system for detecting bobbin color tubes based on image analysis. Background Art

[0002] The technical field of image analysis involves technologies and algorithms for extracting useful information and features from digital images to automatically analyze and interpret image data. The purpose of image analysis is to extract meaningful information and features from images to achieve functions such as automated detection, classification, tracking, and recognition. This technology is widely used in fields such as medical imaging, satellite remote sensing, industrial automation, traffic monitoring, security systems, and digital media. By preprocessing, segmenting, feature extracting, and classifying images through algorithms, image analysis can improve work efficiency, accuracy, and reduce human errors.

[0003] Among them, the method for detecting bobbin color tubes is a technology specifically used to detect and analyze the color difference and quality of color tubes in textiles, mainly used in the textile industry to ensure color consistency and quality control during the production process of textiles. This method automatically detects color deviations on the color tubes through image analysis technology, thereby preventing color difference problems from affecting the appearance and quality of the final product. By using a camera to capture images of the yarn, analyzing the images, detecting and classifying color changes, it helps the production line adjust process parameters to achieve color standardization, reduce the scrap rate, and improve production efficiency.

[0004] When traditional systems process the color detection of textiles, they are restricted by the influence of light changes and image noise, and cannot guarantee the stability and consistency of detection results. Lack of effective light correction, color deviations in different regions of the image will be misjudged as quality problems, resulting in false alarms or missed detections. The existing technology's handling of shadow interference is relatively rough, and it cannot effectively reduce the shadow impact while retaining image details, resulting in the loss of yarn detail information. The segmentation of the track background and the yarn area is usually based on a global threshold, which is easily affected by the complexity of the background and the diversity of color distributions, with low segmentation accuracy, increasing the error of subsequent color analysis. For the detection of color uniformity and deviations, most methods rely on fixed thresholds set manually, lacking a dynamic adjustment mechanism and unable to adapt to the natural fluctuations of yarn colors, prone to misjudgments or missed detections, leading to low quality control efficiency in the textile industry, increased scrap rate, and affecting production benefits and product quality. Summary of the Invention

[0005] The purpose of the present invention is to solve the disadvantages existing in the prior art, and to propose a method and system for detecting bobbin color tubes based on image analysis.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] A method for detecting the color tubes of a winding machine based on image analysis, comprising the following steps:

[0008] S1: Based on the captured image of the tube yarn conveying track of the winding machine, separate the RGB color components, statistically analyze the intensity distribution of the color components, compare the differences with the standard illumination color image, and adjust the proportions of the color components to obtain an illumination color correction image;

[0009] S2: Based on the illumination color correction image, mark the areas where the gradient change amplitude exceeds the set range as candidate shadow areas, extract and analyze the brightness distribution and shape features, mark the shadow influence areas, and adjust the brightness values of the pixels within each shadow area to balance the brightness and obtain a shadow correction image;

[0010] S3: Based on the shadow correction image, extract the color components of the tube yarn conveying track area of the winding machine, compare with the standard color difference, mark the background area, group the unmarked pixels, and eliminate the areas whose area and shape features do not conform to the characteristics of the roving tube to obtain a roving tube separation image;

[0011] S4: Based on the roving tube edge separation image, compare the differences in the color histograms and take the mean of the differences to obtain the color deviation analysis result, and evaluate the color uniformity of the roving tube by statistically analyzing the color changes in each roving tube area to obtain the color uniformity analysis result;

[0012] S5: Based on the color deviation analysis result and the color uniformity analysis result, adjust the color deviation tolerance threshold according to the quality control objective, evaluate the color abnormality level of the current roving tube, eliminate the mislabeled tubes from the track, and immediately notify the management personnel to obtain the abnormality identification management information.

[0013] As a further solution of the present invention, the illumination color correction image is specifically an image corrected for the red component, the green component, and the blue component. The shadow correction image includes a shadow distribution area, a non-shadow distribution area, and a brightness balance adjustment area. The roving tube separation image includes a track background area, a roving tube edge area, and a connectivity feature area. The color deviation analysis result includes the difference value of the current roving tube color distribution histogram and the difference value of the roving tube color component mean. The color uniformity analysis result includes the color fluctuation amplitude and the color change uniformity distribution index. The abnormality identification management information includes the color abnormality level, the mislabeled tube list, the track elimination record, and the management notification information.

[0014] As a further solution of the present invention, the steps of separating the RGB color components, statistically analyzing the intensity distribution of the color components, comparing the differences with the standard illumination color image, and adjusting the proportions of the color components to obtain the illumination color correction image based on the captured image of the tube yarn conveying track of the winding machine are specifically as follows:

[0015] S101: Based on the captured image of the cheese yarn conveying track of the winder, read the numerical values of the light intensity components of red, green, and blue pixel by pixel, perform intensity value statistics on each color component, and obtain the color component intensity distribution data by summing and normalizing the intensity values of each pixel component;

[0016] S102: Based on the color component intensity distribution data, compare the current intensity value of each color component with the intensity value of the corresponding component in the standard illumination color image of the cheese yarn conveying track of the winder, calculate the difference one by one, count the difference range, and evaluate the color deviation degree using the average difference and the maximum difference to obtain the color component intensity difference data;

[0017] S103: Based on the color component intensity difference data, adjust the pixel intensity values of the red, green, and blue color components in the current image according to the difference ratio. By sequentially calculating and updating the color components of each pixel, synthesize the adjusted red, green, and blue components pixel by pixel to obtain the illumination color corrected image.

[0018] As a further solution of the present invention, based on the illumination color corrected image, the steps of marking the area where the gradient change amplitude exceeds the set range as the candidate shadow area, extracting and analyzing the brightness distribution and shape features, marking the shadow influence area, and adjusting the brightness values of the pixels in each shadow area to balance the brightness to obtain the shadow corrected image are specifically as follows:

[0019] S201: Based on the illumination color corrected image, calculate the pixel brightness gradient value in the image pixel by pixel, record the difference between the brightness value of each pixel and the brightness value of the adjacent pixel as the local gradient, screen the pixel points where the gradient change amplitude exceeds the set range, and mark the pixel point positions as the candidate shadow areas to obtain the candidate shadow area marking information;

[0020] S202: Based on the candidate shadow area marking information, extract the brightness distribution curve in each candidate area, analyze the curve fluctuation amplitude, screen the pixel groups with a fluctuation amplitude smaller than the background brightness change range, and verify in combination with the boundary shape features of the pixel groups. Mark the areas that meet the conditions as the shadow influence range to obtain the shadow influence range marking information;

[0021] S203: Based on the shadow influence range marking information, adjust the pixel brightness values in the shadow area step by step. By calculating the difference between the average brightness of the shadow area and the average brightness of the adjacent non-shadow area, gradually reduce the brightness difference and update the pixel brightness values in the shadow area to compensate for the shadow influence and obtain the shadow corrected image.

[0022] As a further solution of the present invention, based on the shadow correction image, the color component of the bobbin conveying track area of ​​the winding machine is extracted, compared with the standard color difference, the background area is marked, the unmarked pixels are grouped, and the areas whose area and shape characteristics do not conform to the characteristics of the spun yarn tube are eliminated, and the steps of obtaining the spun yarn tube separation image are specifically as follows:

[0023] S301: Based on the shadow correction image, read the red, green and blue color component values ​​in the image pixel by pixel, calculate the difference between each pixel and the standard track color value, mark the pixels with difference values ​​lower than the set threshold as track background areas, and record the corresponding position information to generate track background marking information;

[0024] S302: Based on the track background marking information, a pixel group not marked as a track background area is screened, grouped according to a spatial connectivity rule, a boundary contour of each connected area is extracted, the area value, boundary shape feature and position distribution information of each connected area in the image are counted one by one, and areas whose area and shape features do not conform to the spun yarn tube features are eliminated to obtain the spun yarn tube feature area information;

[0025] S303: Based on the spun yarn tube characteristic region information, the region meeting the spun yarn tube characteristics is segmented, and the continuity and integrity of the region boundary are optimized by interpolation, and the segmented spun yarn tube region is reconstructed into an independent pixel set, and integrated into an image to obtain a spun yarn tube separation image.

[0026] As a further solution of the present invention, based on the spun yarn tube edge separation image, the color histogram difference is compared, and the difference is averaged to obtain the color deviation analysis result, and the color uniformity of the spun yarn tube is evaluated by counting the color changes of each spun yarn tube area. The steps of obtaining the color uniformity analysis result are specifically as follows:

[0027] S401: based on the spun yarn bobbin separation image, extract the red, green and blue color component values ​​of each spun yarn bobbin area pixel by pixel, convert the color component values ​​into a standard color space, record each spun yarn bobbin area to obtain color distribution information, and obtain spun yarn bobbin color distribution data;

[0028] S402: based on the spun yarn tube color distribution data, a color distribution histogram of each spun yarn tube area is calculated, and a pixel-by-pixel difference value is calculated with a standard spun yarn tube color histogram, and the difference values ​​are averaged and recorded as spun yarn tube color deviation values ​​to obtain a color deviation analysis result;

[0029] S403: Based on the spun yarn tube color distribution data, the variation range and fluctuation amplitude of the color component value of each spun yarn tube area in the area are calculated, and compared with the track background color component value, the stability and fluctuation degree of the spun yarn tube color distribution are analyzed, and the color uniformity of each spun yarn tube is evaluated to obtain the color uniformity analysis result.

[0030] As a further solution of the present invention, the formula for evaluating the color uniformity of each roving bobbin is as follows:

[0031]

[0032] where U is the color uniformity score of the roving bobbin, C i represents the color value of the i-th pixel, C mean represents the average value of the color values of all pixels within the roving bobbin area, C max and C min represent the maximum and minimum color values within the area respectively, N is the total number of pixels within the roving bobbin area, and w i is the weight of the i-th pixel.

[0033] As a further solution of the present invention, based on the color deviation analysis result and the color uniformity analysis result, according to the quality control objective, adjusting the color deviation tolerance threshold, evaluating the color abnormality level of the current roving bobbin, removing the marked misaligned bobbins from the track, and immediately notifying the management staff to obtain the abnormal recognition management information, the specific steps are as follows:

[0034] S501: Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control objective, adjust the color deviation tolerance threshold, read the color deviation values of the roving bobbins one by one, compare each deviation value with the adjusted tolerance threshold, record the amplitude beyond the range, and summarize each deviation amplitude to obtain the color deviation amplitude data;

[0035] S502: Based on the color deviation amplitude data, combined with the color uniformity data of the roving bobbins, calculate the color abnormality level of each roving bobbin, compare the abnormality level with the set warning range, mark the roving bobbins beyond the warning range as misaligned bobbins, and obtain the misaligned bobbin marking data;

[0036] S503: Based on the misaligned bobbin marking data, locate the corresponding misaligned bobbin positions in the track image, remove the misaligned bobbins from the track, record the position information of the misaligned bobbin markings, and immediately notify the management staff to obtain the abnormal recognition management information.

[0037] As a further solution of the present invention, the formula for calculating the color abnormality level of each roving bobbin is as follows:

[0038] E = k×(1 - U)+w×|D - T|;

[0039] where E is the color abnormality level of the roving bobbin, U represents the color uniformity score of the roving bobbin, D represents the color deviation value, T is the color deviation tolerance threshold, w is the color deviation value weight, and k is the color uniformity weight coefficient.

[0040] An image analysis-based cheese tube detection system, the image analysis-based cheese tube detection system is used to execute the above-mentioned image analysis-based cheese tube detection method, and the system includes:

[0041] The illumination color correction module separates the RGB color components based on the captured image of the cheese tube conveying track of the winding machine, statistically analyzes the intensity distribution of the color components, compares the differences with the standard illumination color image, and adjusts the ratio of the color components to obtain an illumination color correction image;

[0042] The shadow balance processing module marks the areas with gradient change amplitudes exceeding the set range as candidate shadow areas based on the illumination color correction image, extracts and analyzes the brightness distribution and shape features, marks the shadow influence areas, and adjusts the brightness values of the pixels within each shadow area to balance the brightness and obtain a shadow correction image;

[0043] The background separation processing module extracts the color components of the cheese tube conveying track area based on the shadow correction image, compares the differences with the standard color difference, marks the background areas, groups the unmarked pixels, and eliminates the areas whose area and shape features do not conform to the characteristics of the cheese tube to obtain a cheese tube separation image;

[0044] The color deviation evaluation module compares the differences in color histograms based on the cheese tube edge separation image, takes the average value of the differences to obtain a color deviation analysis result, and evaluates the uniformity of the cheese tube color by statistically analyzing the color changes in each cheese tube area to obtain a color uniformity analysis result;

[0045] The abnormal tube handling module adjusts the color deviation tolerance threshold according to the quality control target based on the color deviation analysis result and the color uniformity analysis result, evaluates the color abnormality level of the current cheese tube, eliminates the marked wrong tubes from the track, and immediately notifies the management personnel to obtain abnormal identification management information.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] In the present invention, by separating and adjusting the red, green, and blue color components in the image, the influence of the lighting conditions is eliminated, the color consistency of the input image is ensured, and the reliability of the analysis is improved. By calculating the pixel brightness gradient value and correcting the shadow area, the shadow interference in the image is reduced, and the clarity and detail retention effect of the image are improved. When detecting the yarn track area, by calculating the difference in pixel-level color values and grouping according to the connectivity rule, the track background and the yarn area are accurately distinguished, the irrelevant areas are effectively removed, and the accuracy of the roving bobbin detection is enhanced. By converting the color information of the yarn area to the standard color space and calculating the color distribution histogram, the color deviation and uniformity of the roving bobbin are accurately evaluated, the detailed classification and grading of color anomalies are realized, and the color deviation tolerance threshold is dynamically adjusted according to the statistical results, which improves the flexibility and accuracy of color anomaly detection, effectively avoids noise interference, and improves the recognition efficiency of color anomalies and the quality control ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0049] Figure 1 It is a schematic diagram of the working process of the present invention;

[0050] Figure 2 It is a detailed flowchart of S1 of the present invention;

[0051] Figure 3 It is a detailed flowchart of S2 of the present invention;

[0052] Figure 4 It is a detailed flowchart of S3 of the present invention;

[0053] Figure 5 It is a detailed flowchart of S4 of the present invention;

[0054] Figure 6 It is a detailed flowchart of S5 of the present invention;

[0055] Figure 7 It is a system flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] The following will describe the technical solutions in the present invention with reference to the drawings.

[0057] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two can be selected.

[0058] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same. "(of)", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same.

[0059] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0060] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0061] Please refer to Figure 1 , the method for detecting the color tube of a winding machine based on image analysis provided by the present invention includes the following steps:

[0062] S1: Based on the captured image of the tube yarn conveying track of the winding machine, separate the red, green, and blue color light components, respectively count the intensity distribution of each color component, evaluate the difference value between the corresponding components of the current image and the standard illumination color image, and according to the illumination color deviation, respectively perform proportional adjustment on the red, green, and blue color components of the current image, and synthesize the adjusted components into a corrected image to obtain an illumination color corrected image;

[0063] S2: Based on the illumination color corrected image, calculate the pixel brightness gradient value in the image, mark the area where the gradient change amplitude exceeds the set range as a candidate shadow area, extract the brightness distribution curve and boundary shape features of the pixels in each candidate shadow area, mark the area where the curve fluctuation amplitude is less than the background brightness change range as the shadow influence range, adjust the brightness value of the pixels in each shadow area, calculate the difference between the average brightness of the shadow area and the average brightness of the adjacent non-shadow area, and perform a decreasing compensation on the difference to obtain a shadow corrected image;

[0064] S3: Based on the shadow-corrected image, extract the color components of the bobbin yarn conveying track area of the winding machine, calculate the difference value between the color value of each pixel in the image and the known standard track color value, mark the pixels with a difference value lower than the set threshold as the track background area, group the pixel clusters not marked as the track background area according to the connectivity rule, and count the area, shape features, and position distribution of each connected area. Eliminate the areas whose area and shape features do not conform to the characteristics of the roving bobbins to obtain the roving bobbin separation image;

[0065] S4: Based on the roving bobbin edge separation image, convert the color components of the roving bobbin area to the standard color space, calculate the color distribution histogram of the current roving bobbin, calculate the difference value between the histogram and the histogram of the standard roving bobbin color, and take the average value of the difference to obtain the color deviation analysis result. By statistically analyzing the color changes in each roving bobbin area and according to the degree of color fluctuation, evaluate the color uniformity of the roving bobbins to obtain the color uniformity analysis result;

[0066] S5: Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control target, adjust the color deviation tolerance threshold, compare the color deviation value of the current roving bobbin with the adjusted tolerance threshold, and combine the color uniformity of the roving bobbins to evaluate the color abnormality level of the current roving bobbin. Mark the roving bobbins with a color abnormality level greater than the set warning range as misaligned bobbins, remove the marked misaligned bobbins from the track, and immediately notify the management personnel to obtain the abnormal identification management information.

[0067] The light color correction image is specifically the image after red component correction, green component correction, and blue component correction. The shadow correction image includes a shadow distribution area, a non-shadow distribution area, and a brightness equalization adjustment area. The roving bobbin separation image includes a track background area, a roving bobbin edge area, and a connectivity feature area. The color deviation analysis result includes the difference value of the current roving bobbin color distribution histogram and the difference value of the average value of the roving bobbin color components. The color uniformity analysis result includes the color fluctuation amplitude and the color change uniformity distribution index. The abnormal identification management information includes the color abnormality level, the misaligned bobbin list, the track removal record, and the management notification information.

[0068] Please refer to Figure 2 , based on the captured image of the bobbin yarn conveying track of the winding machine, separate the red, green, and blue color light components, statistically analyze the intensity distribution of each color component respectively, evaluate the difference value between the current image and the corresponding component of the standard light color image, and according to the light color deviation, perform proportional adjustment on the red, green, and blue color components of the current image respectively. The steps to synthesize the adjusted components into the corrected image to obtain the light color correction image are specifically as follows:

[0069] S101: Based on the captured image of the bobbin conveying track of the winding machine, the values ​​of the three color illumination components of red, green and blue are read pixel by pixel, and the intensity value of each color component is counted, and the intensity value of each pixel component is summed and normalized to obtain the color component intensity distribution data;

[0070] Based on the captured image of the bobbin conveyor track of the winding machine, the intensity values ​​of the red, green and blue color illumination components are extracted pixel by pixel. The color component of each pixel is separated after parsing by reading the image data, and the color component intensities of all pixels are counted item by item. The statistical method is to accumulate the component values ​​of the pixel points and then perform normalization processing. The total value of the statistical component intensity is used as the denominator during the normalization processing, and the component intensity value of each pixel point is standardized. The normalized intensity values ​​are rearranged to generate a distribution array of color components. The distribution array is used as the data source of the component intensity distribution, and the normalization processing steps are completed for the red, green and blue color components respectively. At the same time, the total intensity value parameters and the pixel-by-pixel intensity values ​​of the components used in the normalization process are recorded for difference calculation or verification in subsequent operations to obtain color component intensity distribution data.

[0071] S102: Based on the color component intensity distribution data, the current intensity value of each color component is compared with the intensity value of the corresponding component in the standard illumination color image of the bobbin conveying track of the winding machine, the difference is calculated one by one, the difference range is counted, and the color deviation degree is evaluated by using the average difference and the maximum difference to obtain the color component intensity difference data;

[0072] Based on the color component intensity distribution data, the current intensity value of each color component is compared pixel by pixel with the intensity value of the corresponding component in the standard illumination color image of the bobbin conveying track of the winding machine, and the difference value is calculated one by one. The difference value is obtained by subtracting the standard illumination color component value from the current component intensity value. After obtaining the difference value, each difference value is arranged by pixel to generate a difference array, and the value range in the difference array is statistically analyzed, and the maximum value and average value of the difference value are recorded as key data points. The maximum value is used to judge the degree of abnormality, and the average value is used to reflect the overall deviation level. To ensure the accuracy of the data, the noise value less than the preset threshold in the difference array needs to be eliminated during the statistical process. At the same time, a difference range distribution map is established for each color component, and the total amount and distribution information of the difference value of the corresponding component are recorded one by one. The color component intensity difference data is formed by comparing the difference range and statistical data.

[0073] S103: Based on the color component intensity difference data, the pixel intensity values ​​of the red, green and blue color components in the current image are adjusted according to the difference ratio, and the color component of each pixel is calculated and updated in sequence, and the adjusted red, green and blue components are synthesized pixel by pixel to obtain a light color correction image;

[0074] Based on the color component intensity difference data, the pixel intensity values of the red, green, and blue color components in the current image are adjusted one by one. The adjustment ratio is completed by calculating the ratio coefficient of the current component intensity value to the corresponding component difference data. When adjusting, the intensity value of each color component is updated pixel by pixel. The adjusted data of each color component is respectively stored as an updated component value array. Subsequently, the adjusted red component, green component, and blue component are synthesized pixel by pixel. When synthesizing, the three component values of each pixel are successively added to generate a new pixel color value, and the synthesized pixel values are arranged to generate corrected image data. After the color component adjustment of all image data is completed pixel by pixel, a light color corrected image is obtained.

[0075] Please refer to Figure 3 , based on the light color corrected image, calculate the pixel brightness gradient value in the image. Mark the area where the gradient change amplitude exceeds the set range as the candidate shadow area. For each pixel in each candidate shadow area, extract the brightness distribution curve and boundary shape features. Mark the area where the curve fluctuation amplitude is less than the background brightness change range as the shadow influence range. The steps to adjust the brightness value of each pixel in the shadow area, calculate the difference between the average brightness of the shadow area and the average brightness of the adjacent non-shadow area, and perform a decreasing compensation on the difference to obtain the shadow corrected image are as follows:

[0076] S201: Based on the light color corrected image, calculate the pixel brightness gradient value in the image pixel by pixel. Record the difference between the brightness value of each pixel and the brightness value of the adjacent pixel as the local gradient. Screen the pixel points where the gradient change amplitude exceeds the set range, and mark the position of the pixel points as the candidate shadow area to obtain the candidate shadow area marking information;

[0077] Based on the light color corrected image, read the brightness value of each pixel in the image and calculate the difference between the brightness values of adjacent pixels pixel by pixel. Record this difference as the local gradient. Calculate the horizontal gradient and vertical gradient respectively by traversing rows and columns. After taking the absolute value of the gradient value, count its distribution range. Combine the distribution range to screen the pixel points where the gradient change amplitude exceeds the set range. Mark the screened pixel points as the candidate shadow area. Record the positions of the marked pixel points in matrix form. Verify the connectivity of the marks in the matrix, exclude the discrete pixel points, and retain the connected areas. Count the boundary features of each connected area, such as side length and shape integrity. Exclude the boundary feature areas that do not meet the requirements, and use the connected areas that meet the requirements as the candidate shadow areas to generate the candidate shadow area marking information.

[0078] S202: Based on the candidate shadow area marking information, extract the brightness distribution curve within each candidate area, analyze the fluctuation amplitude of the curve, filter out the pixel groups with a fluctuation amplitude smaller than the background brightness change range, and verify by combining the boundary shape features of the pixel groups. Mark the areas that meet the conditions as the shadow influence range to obtain the shadow influence range marking information;

[0079] Based on the candidate shadow area marking information, extract the brightness distribution curve within each candidate area, plot the distribution graph with the brightness value as the abscissa and the pixel position as the ordinate, analyze the fluctuation amplitude of the curve and compare it with the background brightness change range, filter out the pixel groups with a smaller fluctuation amplitude as the possible shadow areas, extract the boundary shape features of the filtered areas, compare and verify the shape features with the track background features, remove the areas that do not conform to the track background boundary shape, mark the shadow areas that meet the conditions, and integrate and record the pixel value positions and feature information of the shadow areas to obtain the shadow influence range marking information.

[0080] S203: Based on the shadow influence range marking information, gradually adjust the pixel brightness values within the shadow area. By calculating the difference between the average brightness of the shadow area and the average brightness of the adjacent non-shadow area, gradually reduce the brightness difference and update the pixel brightness values in the shadow area to compensate for the shadow influence and obtain the shadow-corrected image.

[0081] Based on the shadow influence range marking information, compare the brightness value of each pixel within the marked area with the average brightness value of the adjacent non-shadow area, calculate the difference between the two, and compensate the difference into the brightness values of the pixels in the shadow area in a decreasing segmented manner. Gradually adjust the brightness values of all pixels within the shadow area to smooth the brightness distribution within the area. After the adjustment of the pixel values within each shadow area is completed, replace the corresponding pixels in the original image with the updated pixel brightness values, reconstruct the overall image brightness distribution, and generate the shadow-corrected image after synthesizing all pixels.

[0082] Please refer to Figure 4 , based on the shadow-corrected image, extract the color components of the bobbin tube yarn conveying track area of the winding machine, calculate the difference value between the color value of each pixel in the image and the known standard track color value, mark the pixels with a difference value lower than the set threshold as the track background area, group the pixel groups that are not marked as the track background area according to the connectivity rule, and count the area, shape features, and position distribution of each connected area group. Remove the areas whose area and shape features do not conform to the characteristics of the roving bobbin. The steps to obtain the roving bobbin separation image are specifically as follows:

[0083] S301: Based on the shadow-corrected image, read the red, green, and blue color component values of each pixel in the image one by one, calculate the difference value between each pixel and the standard track color value, mark the pixels with a difference value lower than the set threshold as the track background area, and record the corresponding position information to generate the track background marking information;

[0084] Based on the shadow-corrected image, read the red, green, and blue color component values of the image pixel by pixel. Compare the color value of each pixel with the standard track color value one by one, calculate the difference in color values, which is obtained by pixel-by-pixel subtraction and stored as a difference array. Judge the difference values of all pixels in the difference array one by one, filter out the pixel points below the set threshold, mark these pixel points as the track background area, and record the position information of these pixel points at the same time. During the marking process, combine the matching degree of the pixel points with the surrounding environment color for review, and eliminate the mismarking that may be caused by noise or sampling error. Integrate all marked track background areas and record them as track background marking information.

[0085] S302: Based on the track background marking information, filter out the pixel groups not marked as the track background area, group them according to the spatial connectivity rule, extract the boundary contours of each connected area, count the area values, boundary shape features, and position distribution information in the image of each connected area one by one, and eliminate the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain the roving bobbin feature area information;

[0086] Based on the track background marking information, extract the pixel groups not marked as the track background. Group the pixel groups according to the spatial connectivity rule. The connectivity rule is determined by judging whether the spatial relationship between the pixel point and its adjacent pixel points is continuous. Extract the boundary contours of each group of connected areas, use the boundary tracking technology to record the shape information of the connected areas, and calculate the area values of the connected areas one by one. Store the area values and boundary shape features separately, and record them in combination with the position distribution information of the connected areas in the image. Filter out the connected areas whose area and shape features conform to the characteristics of the roving bobbin, and record the information of these areas after eliminating other non-conforming areas to obtain the roving bobbin feature area information.

[0087] S303: Based on the roving bobbin feature area information, segment the areas that conform to the roving bobbin characteristics, and optimize the continuity and integrity of the area boundaries through interpolation. Reconstruct the segmented roving bobbin areas into independent pixel sets and integrate them into an image to obtain the roving bobbin separation image;

[0088] Based on the roving bobbin feature area information, segment each area that conforms to the roving bobbin characteristics. During the segmentation process, crop the pixel data according to the boundary information of the area, use interpolation to optimize the continuity of the boundary contour, ensure the smooth transition of the boundary line by interpolating the pixel values between the boundary points, and at the same time reconstruct the independent pixel sets for each segmented area. Rearrange the pixel values of the segmented areas into an independent image matrix, integrate the image data of all segmented areas, and combine the segmentation results of all roving bobbin areas into a complete image to obtain the roving bobbin separation image.

[0089] Please refer to Figure 5 , based on the separated image of the roving bobbin edge, convert the color components of the roving bobbin area to the standard color space, calculate the color distribution histogram of the current roving bobbin, calculate the difference value between the histogram and the histogram of the standard roving bobbin color, and take the average of the differences to obtain the color deviation analysis result. And by statistically analyzing the color changes in each roving bobbin area, and according to the degree of color fluctuation, evaluate the color uniformity of the roving bobbin to obtain the color uniformity analysis result. The specific steps are as follows:

[0090] S401: Based on the separated image of the roving bobbin, extract the red, green, and blue color component values of each roving bobbin area pixel by pixel, convert the color component values to the standard color space, record the color distribution information obtained for each roving bobbin area, and obtain the roving bobbin color distribution data;

[0091] Based on the separated image of the roving bobbin, extract the red, green, and blue color component values of each roving bobbin area pixel by pixel, read and store the three component values of each pixel one by one, and then, according to the conversion requirements of the standard color space, complete the conversion of the extracted component values by looking up the color mapping matrix, record the converted values and associate them with the spatial distribution information of the roving bobbin area, generate a corresponding color distribution array for each roving bobbin area, integrate the distribution arrays into color distribution information, and generate the roving bobbin color distribution data by combining the position of each pixel and the corresponding color space value.

[0092] S402: Based on the roving bobbin color distribution data, calculate the color distribution histogram of each roving bobbin area, calculate the difference value pixel by pixel with the standard roving bobbin color histogram, and record the average of the difference values as the roving bobbin color deviation value to obtain the color deviation analysis result; Based on the roving bobbin color distribution data, calculate the color distribution histogram for each roving bobbin area, statistically count the occurrence frequency of each color component value pixel by pixel, divide the frequency values into multiple intervals according to the range of the standard color space, record the statistical results as a histogram array, and then calculate the difference value pixel by pixel between the color histogram of each roving bobbin area and the standard roving bobbin color histogram. The difference value is calculated by accumulating the differences of the corresponding component values of the pixel points, and the overall color deviation value is obtained by taking the average of the difference values of each area. Record the deviation value as the color deviation analysis result of the roving bobbin.

[0093] S403: Based on the roving bobbin color distribution data, calculate the change range and fluctuation amplitude of the color component values in each roving bobbin area within the area, compare with the track background color component values, analyze the stability and fluctuation degree of the roving bobbin color distribution, evaluate the color uniformity of each roving bobbin, and obtain the color uniformity analysis result;

[0094] Based on the color distribution data of the spindle tubes, calculate the range of variation and the amplitude of fluctuation of the color component values for each spindle tube area one by one. The range of variation is calculated by statistically finding the difference between the maximum and minimum values of each color component value, and the amplitude of fluctuation is completed by calculating the average difference between the component value and the component values of its adjacent pixels pixel by pixel. Subsequently, compare the calculated range of variation and amplitude of fluctuation with the corresponding parameters of the track background color component values, and analyze the stability and fluctuation characteristics of the spindle tube color distribution in combination with the comparison results to estimate the color uniformity of each spindle tube and obtain the color uniformity analysis result.

[0095] The formula for evaluating the color uniformity of each spindle tube is:

[0096]

[0097] Where U is the color uniformity score of the spindle tube, C i represents the color value of the i-th pixel, C mean represents the average of the color values of all pixels within the spindle tube area, C max and C min represent the maximum and minimum color values within the area respectively, N is the total number of pixels within the spindle tube area, and w i is the weight of the i-th pixel.

[0098] Formula:

[0099]

[0100] Detailed Explanation of Parameters and Acquisition Methods

[0101] C i represents the color value of the i-th pixel, which is obtained by directly reading the color value of each pixel from the image data file.

[0102] C mean is the average of the color values of all pixels within the spindle tube area, which is obtained by dividing the sum of the color values of all pixels in this area by the number of pixels N.

[0103] C max and C min represent the maximum and minimum color values within the area respectively, which are obtained by traversing the color values of all pixels in the area and recording the maximum and minimum values.

[0104] N is the total number of pixels within the spindle tube area, which is obtained by counting the number of pixel points included in the area.

[0105] w i is the weight of the i-th pixel, which is a weight coefficient estimated according to the influence weight of the pixel's position in the image. Usually, the weight of each pixel is set to be equal, that is, w i = 1.

[0106] Calculation example:

[0107] Suppose the cop region contains N = 100 pixels, and the color value C i is distributed from 50 to 150, where C mean = 100, C max = 150, and C min = 50, and suppose the weight w of all pixels i = 1. Substitute the values into the above formula to calculate the color uniformity score:

[0108] Calculate the absolute value of the difference between the color value of each pixel and the average value. For the pixel with C i = 75, the absolute value of the difference is |75 - 100| = 25.

[0109] Add up the absolute values of all differences. Suppose the average of these absolute values of differences is 25, then the sum is 25 × 100 = 2500.

[0110] Calculate the normalization factor (C max - C min ) = 150 - 50 = 100.

[0111] Substitute into the formula:

[0112] The calculation result U = 75 indicates that the color uniformity score of the cop region is 75, meaning that the color distribution in this region is relatively uniform compared to the possible maximum color difference. The closer the score is to 100, the higher the color uniformity.

[0113] Please refer to Figure 6 , based on the color deviation analysis result and the color uniformity analysis result, according to the quality control objective, adjust the color deviation tolerance threshold, compare the current color deviation value of the cop with the adjusted tolerance threshold, combine with the color uniformity of the cop, evaluate the color abnormality level of the current cop, mark the cop with a color abnormality level greater than the set warning range as a defective cop, remove the marked defective cop from the track, and immediately notify the management personnel to obtain the abnormal identification management information. The specific steps are as follows:

[0114] S501: Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control objective, adjust the color deviation tolerance threshold, read the color deviation values of the cops one by one, compare each deviation value with the adjusted tolerance threshold, record the amplitude beyond the range, summarize each deviation amplitude, and obtain the color deviation amplitude data;

[0115] Based on the color deviation analysis results and color uniformity analysis results, read the color deviation values of the roving bobbins one by one, compare each deviation value with the adjusted tolerance threshold one by one to obtain the setting range of the tolerance threshold, which is determined by the quality control target and the historical color analysis data. By statistically analyzing the distribution of color deviation values in the historical data, select a specific percentile as the adjusted tolerance threshold, calculate the difference between each deviation value and the tolerance threshold, record the deviation amplitude exceeding the threshold as valid difference data, and at the same time count all the deviation amplitude values exceeding the range, and summarize them item by item to generate complete color deviation amplitude data.

[0116] S502: Based on the color deviation amplitude data and combined with the color uniformity data of the roving bobbins, calculate the color anomaly level of each roving bobbin, compare the anomaly level with the set warning range, and mark the roving bobbins exceeding the warning range as defective bobbins to obtain defective bobbin marking data;

[0117] Based on the color deviation amplitude data and combined with the color uniformity data of the roving bobbins, calculate the color anomaly level of each roving bobbin item by item. The anomaly level is obtained based on the weighted comprehensive result of the deviation amplitude data and the color uniformity data. Normalize the color uniformity data, map the uniformity deviation to a unified grade distribution range, and at the same time calculate the proportion of the color deviation amplitude. Adjust the anomaly level according to the linear relationship between the proportion and the uniformity grade value. Compare the anomaly level results with the preset warning range one by one, screen out the roving bobbins exceeding the warning range and mark them as defective bobbins, and generate defective bobbin marking data including the marking information and anomaly level of each defective bobbin.

[0118] The formula for calculating the color anomaly level of each roving bobbin is:

[0119] E = k×(1 - U) + w×|D - T|;

[0120] Where, E is the color anomaly level of the roving bobbin, U represents the color uniformity score of the roving bobbin, D represents the color deviation value, T is the color deviation tolerance threshold, w is the weight of the color deviation value, and k is the color uniformity weight coefficient.

[0121] Formula:

[0122] E = k×(1 - U) + w×|D - T|;

[0123] Detailed parameter explanation and acquisition method:

[0124] U is the color uniformity score: indicating the degree of color uniformity of the roving bobbin. The calculation method is to analyze the image data, statistically calculate the standard deviation or variance of the pixel color values in the area, and then normalize it to the range of 0 to 100. A high value indicates that the color is relatively consistent, and a low value indicates uneven color distribution.

[0125] D is the color deviation value: The color value of each pixel is extracted from the roving bobbin image by a color difference meter or image processing software, and the difference from the target color (or standard color) is calculated.

[0126] T is the color deviation tolerance threshold: Set by the quality management team according to product quality standards or historical data, representing the maximum acceptable color deviation.

[0127] w is the color deviation weight: This weight defines the influence of different types of color deviations in the total score, usually set according to the impact of color deviation on the appearance of the final product.

[0128] k is the color uniformity weight coefficient: This is a coefficient used to adjust the influence of color uniformity in the total anomaly score, reflecting the degree of emphasis on color uniformity in the manufacturing process.

[0129] Calculation example:

[0130] Suppose: U = 80 (color uniformity score, indicating relatively uniform color); D = 0.25 (actually measured color deviation); T = 0.2 (color deviation tolerance threshold); w = 2 (color deviation weight); k = 0.7 (color uniformity weight);

[0131] Calculate the anomaly level:

[0132] E = 0.7×(1 - 0.8) + 2×|0.25 - 0.2| = 0.7×0.2 + 2×0.05 = 0.24;

[0133] The calculated result E = 0.24 indicates that the color anomaly level of this roving bobbin is 0.24, showing that although there is a slight color deviation, the color uniformity of this roving bobbin is good and does not exceed the warning range (assuming the warning range is above 0.3). Therefore, this roving bobbin is not marked as a defective tube. This indicates that the color uniformity and color deviation are within the allowable range, and the product quality meets the standards.

[0134] S503: Based on the defective tube marking data, locate the corresponding defective tube position in the track image, remove the defective tube from the track, record the position information of the defective tube marking, and immediately notify the management personnel to obtain the anomaly recognition management information;

[0135] Based on the defective tube marking data, locate the corresponding defective tube position in the track image, extract the track position coordinates of the defective tube through the marking information. The corresponding coordinates are mapped from the relationship between the pixel area recorded in the image and the track position. Subsequently, perform pixel screening on the image area where the defective tube is located, set the screened pixel data to zero to achieve the operation of removing the defective tube. At the same time, store the defective tube position and its marking information as an independent record, and send a notice to the management personnel to generate the anomaly recognition management information for processing and monitoring.

[0136] Please refer to Figure 7 , a bobbin color tube detection system based on image analysis. The bobbin color tube detection system based on image analysis is used to execute the above-mentioned bobbin color tube detection method based on image analysis. The system includes:

[0137] The illumination color correction module separates the RGB color components based on the captured image of the bobbin tube yarn conveying track of the winding machine, statistically analyzes the intensity distribution of the color components, compares the differences with the standard illumination color image, and adjusts the proportion of the color components to obtain an illumination color corrected image;

[0138] The shadow balance processing module marks the areas where the gradient change amplitude exceeds the set range as candidate shadow areas based on the illumination color corrected image, extracts and analyzes the brightness distribution and shape features, marks the shadow influence areas, and adjusts the brightness values of the pixels within each shadow area to balance the brightness and obtain a shadow corrected image;

[0139] The background separation processing module extracts the color components of the bobbin tube yarn conveying track area of the winding machine based on the shadow corrected image, compares the differences with the standard color difference, marks the background areas, groups the unmarked pixels, and eliminates the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain a roving bobbin separation image;

[0140] The color deviation evaluation module compares the differences in color histograms based on the roving bobbin edge separation image, takes the average value of the differences to obtain the color deviation analysis result, and evaluates the color uniformity of the roving bobbin by statistically analyzing the color changes in each roving bobbin area to obtain the color uniformity analysis result;

[0141] The abnormal wrong bobbin processing module adjusts the color deviation tolerance threshold according to the quality control target based on the color deviation analysis result and the color uniformity analysis result, evaluates the color abnormality level of the current roving bobbin, eliminates the marked wrong bobbins from the track, and immediately notifies the management personnel to obtain the abnormal identification management information.

[0142] It should be understood that the term "and / or" in this article is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0143] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or a similar expression means any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0144] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above - mentioned processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0145] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms.

[0146] In addition, in each embodiment of the present invention, the functional units can be integrated in one processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0147] As described above, the above are only specific implementation manners of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for detecting the color tube of a winding bobbin based on image analysis, characterized in that, It includes the following steps: S1: Based on the captured image of the creel package yarn conveying track, separate the RGB color components, statistically analyze the intensity distribution of the color components, compare the differences with the standard illumination color image, and adjust the proportions of the color components to obtain an illumination color corrected image; S2: Based on the illumination color corrected image, mark the areas where the gradient change amplitude exceeds the set range as candidate shadow areas, extract and analyze the brightness distribution and shape features, mark the shadow affected areas, and adjust the brightness values of the pixels within each shadow area to balance the brightness and obtain a shadow corrected image; S3: Based on the shadow corrected image, extract the color components of the creel package yarn conveying track area, compare with the standard color difference, mark the background area, group the unmarked pixels, and eliminate the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain a roving bobbin separation image; S4: Based on the roving bobbin edge separation image, compare the differences in the color histograms and take the average of the differences to obtain the color deviation analysis result, and evaluate the color uniformity of the roving bobbin by statistically analyzing the color changes in each roving bobbin area to obtain the color uniformity analysis result; S5: Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control target, adjust the color deviation tolerance threshold, evaluate the color abnormality level of the current roving bobbin, remove the mislabeled bobbins from the track, and immediately notify the management personnel to obtain the abnormality identification management information.

2. The method for detecting the winding color tube based on image analysis according to claim 1, wherein The illumination color corrected image is specifically an image after red component correction, green component correction, and blue component correction. The shadow corrected image includes a shadow distribution area, a non-shadow distribution area, and a brightness balance adjustment area. The roving bobbin separation image includes a track background area, a roving bobbin edge area, and a connectivity feature area. The color deviation analysis result includes the difference value of the current roving bobbin color distribution histogram and the difference value of the roving bobbin color component mean. The color uniformity analysis result includes the color fluctuation amplitude and the color change uniformity distribution index. The abnormality identification management information includes the color abnormality level, the mislabeled bobbin list, the track removal record, and the management notification information.

3. The method for detecting a winding color tube based on image analysis according to claim 1, characterized in that, The steps of based on the captured image of the creel package yarn conveying track, separating the RGB color components, statistically analyzing the intensity distribution of the color components, comparing the differences with the standard illumination color image, and adjusting the proportions of the color components to obtain an illumination color corrected image are specifically as follows: S101: Based on the captured image of the creel package yarn conveying track, read the numerical values of the red, green, and blue color illumination components pixel by pixel, statistically analyze the intensity values of each color component, and obtain the color component intensity distribution data by summing and normalizing the intensity values of each pixel component; S102: Based on the color component intensity distribution data, compare the intensity value of each current color component with the intensity value of the corresponding component in the standard illumination color image of the creel package yarn conveying track, calculate the differences one by one, statistically analyze the difference range, and evaluate the color deviation degree using the average difference and the maximum difference to obtain the color component intensity difference data; S103: Based on the color component intensity difference data, adjust the pixel intensity values of the red, green, and blue color components in the current image according to the difference ratio. By sequentially calculating and updating the color components of each pixel, synthesize the adjusted red, green, and blue components pixel by pixel to obtain a light color corrected image.

4. The method for detecting bobbin color tubes based on image analysis according to claim 1, wherein, Based on the light color corrected image, the steps of marking the area where the gradient change amplitude exceeds the set range as a candidate shadow area, extracting and analyzing the brightness distribution and shape features, marking the shadow affected area, and adjusting the brightness values of the pixels in each shadow area to balance the brightness to obtain a shadow corrected image are specifically as follows: S201: Based on the light color corrected image, calculate the pixel brightness gradient value in the image pixel by pixel. Record the difference between the brightness value of each pixel and the brightness value of the adjacent pixel as the local gradient. Screen the pixel points where the gradient change amplitude exceeds the set range, and mark the pixel point positions as candidate shadow areas to obtain candidate shadow area marking information; S202: Based on the candidate shadow area marking information, extract the brightness distribution curve in each candidate area, analyze the curve fluctuation amplitude, screen the pixel groups with fluctuation amplitudes smaller than the background brightness change range, and verify them in combination with the boundary shape features of the pixel groups. Mark the areas that meet the conditions as the shadow affected range to obtain shadow affected range marking information; S203: Based on the shadow affected range marking information, adjust the pixel brightness values in the shadow area step by step. By calculating the difference between the average brightness of the shadow area and the average brightness of the adjacent non-shadow area, gradually reduce the brightness difference and update the pixel brightness values in the shadow area to compensate for the shadow effect and obtain a shadow corrected image.

5. The method for detecting the winding color tube based on image analysis according to claim 1, characterized in that, Based on the shadow corrected image, the steps of extracting the color components of the bobbin yarn conveying track area of the winding machine, comparing with the standard color difference, marking the background area, grouping the unmarked pixels, and removing the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain a roving bobbin separation image are specifically as follows: S301: Based on the shadow corrected image, read the red, green, and blue color component values in the image pixel by pixel, calculate the difference value between each pixel and the standard track color value, mark the pixels with difference values lower than the set threshold as the track background area, and record the corresponding position information to generate track background marking information; S302: Based on the track background marking information, screen the pixel groups that are not marked as the track background area, group them according to the spatial connectivity rule, extract the boundary contours of each connected area, count the area values, boundary shape features, and position distribution information in the image of each connected area one by one, and remove the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain roving bobbin feature area information; S303: Based on the roving bobbin feature area information, segment the areas that conform to the characteristics of the roving bobbin, and optimize the continuity and integrity of the area boundary through interpolation. Reconstruct the segmented roving bobbin areas into independent pixel sets and integrate them into an image to obtain a roving bobbin separation image.

6. The method for detecting the winding color tube based on image analysis according to claim 1, wherein Based on the edge separation image of the cop, compare the differences in color histograms, take the mean of the differences to obtain the color deviation analysis result, and evaluate the color uniformity of the cop by statistically analyzing the color changes in each cop area to obtain the color uniformity analysis result. The specific steps are as follows: S401: Based on the cop separation image, extract the red, green, and blue color component values of each cop area pixel by pixel, convert the color component values to the standard color space, record the color distribution information obtained for each cop area, and obtain the cop color distribution data. S402: Based on the cop color distribution data, calculate the color distribution histogram of each cop area, calculate the difference value pixel by pixel with the standard cop color histogram, and record the average of the difference values as the cop color deviation value to obtain the color deviation analysis result. S403: Based on the cop color distribution data, calculate the change range and fluctuation amplitude of the color component values in each cop area within the area, compare with the track background color component values, analyze the stability and fluctuation degree of the cop color distribution, evaluate the color uniformity of each cop, and obtain the color uniformity analysis result.

7. The method for detecting the winding color tube based on image analysis according to claim 6, characterized in that, The formula for evaluating the color uniformity of each cop is: Among them, U is the color uniformity score of the flyer tube, C i represents the color value of the i-th pixel, C mean represents the average value of the color values of all pixels within the flyer tube area, C max and C min represent the maximum and minimum color values within the area respectively, N is the total number of pixels within the flyer tube area, w i is the weight of the i-th pixel.

8. The method for detecting the winding color tube based on image analysis according to claim 1, wherein Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control target, adjust the color deviation tolerance threshold, evaluate the color anomaly level of the current cop, remove the mislabeled cop from the track, and immediately notify the management personnel. The specific steps to obtain the anomaly recognition management information are as follows: S501: Based on the color deviation analysis result and the color uniformity analysis result, according to the quality control target, adjust the color deviation tolerance threshold, read the cop color deviation values one by one, compare each deviation value with the adjusted tolerance threshold, record the exceeded range amplitude, and summarize each deviation amplitude to obtain the color deviation amplitude data. S502: Based on the color deviation amplitude data, combined with the cop color uniformity data, calculate the color anomaly level of each cop, compare the anomaly level with the set warning range, and mark the cop that exceeds the warning range as a mislabeled cop to obtain the mislabeled cop marking data. S503: Based on the mislabeled cop marking data, locate the corresponding mislabeled cop position in the track image, remove the mislabeled cop from the track, record the position information of the mislabeled cop marking, and immediately notify the management personnel to obtain the anomaly recognition management information.

9. The method for detecting the winding color tube based on image analysis according to claim 8, wherein, The formula for calculating the color anomaly level of each cop is: E = k×(1―U)+w×|D―T|; Where, E is the cop color anomaly level, U represents the cop color uniformity score, D represents the color deviation value, T is the color deviation tolerance threshold, w is the color deviation value weight, and k is the color uniformity weight coefficient.

10. A cheese cone detection system based on image analysis, characterized in that, According to the cop detection method based on image analysis according to any one of claims 1-9, the system includes: Based on the captured image of the cheese bobbin conveying track of the winding machine, the light color correction module separates the RGB color components, statistically analyzes the intensity distribution of the color components, compares the differences with the standard light color image, and adjusts the ratio of the color components to obtain the light color corrected image; Based on the light color corrected image, the shadow balance processing module marks the areas with gradient change amplitude exceeding the set range as candidate shadow areas, extracts and analyzes the brightness distribution and shape features, marks the shadow affected areas, adjusts the brightness values of the pixels within each shadow area to balance the brightness, and obtains the shadow corrected image; Based on the shadow corrected image, the background separation processing module extracts the color components of the cheese bobbin conveying track area of the winding machine, compares the color difference with the standard, marks the background area, groups the unmarked pixels, and eliminates the areas whose area and shape features do not conform to the characteristics of the roving bobbin to obtain the roving bobbin separation image; Based on the roving bobbin edge separation image, the color deviation evaluation module compares the differences in the color histograms, takes the mean value of the differences to obtain the color deviation analysis result, and evaluates the color uniformity of the roving bobbin by statistically analyzing the color changes in each roving bobbin area to obtain the color uniformity analysis result; Based on the color deviation analysis result and the color uniformity analysis result, the abnormal wrong bobbin processing module adjusts the color deviation tolerance threshold according to the quality control target, evaluates the color abnormality level of the current roving bobbin, eliminates the marked wrong bobbins from the track, and immediately notifies the management personnel to obtain the abnormal identification management information.

Images