Cotton yarn color difference detection system based on image data

Abstract

This invention discloses a cotton yarn color difference detection system based on image data, belonging to the field of image processing technology. It addresses the limitations of traditional yarn color difference detection, such as the inability to accurately monitor yarn dyeing process images in real-time and over a wide area, and the difficulty in effectively controlling the production process based on image analysis results. The system includes an acquisition module, an image detection module, a matching module, a color difference analysis module, a storage library, an alarm module, and a color difference diagnosis module. The image detection module utilizes image processing technology for sharpness evaluation and analysis, enabling real-time monitoring and control of yarn dyeing quality. The matching module facilitates comprehensive detection of yarn color differences by stitching together the detected images. The color difference analysis module analyzes the color difference deviation values ​​of the analyzed images, enabling color difference alarms and yarn quality grading. The color difference diagnosis module facilitates the analysis of parameter sets indicating color differences, allowing for targeted processing by staff and improving production efficiency.

Description

Technical Field

[0001] This invention belongs to the field of image processing technology, specifically relating to a cotton yarn color difference detection system based on image data. Background Technology

[0002] In the textile industry, quality inspection of cotton yarn is crucial for ensuring the quality of the final product. Color difference, as one of the key factors affecting the appearance quality of yarn, directly relates to whether the product meets market and consumer demands. With the rapid development of the textile industry, traditional manual inspection methods, due to their low efficiency, strong subjectivity, and high labor intensity, can no longer meet the needs of modern production. During the winding process after yarn dyeing, color difference inspection is required to ensure the consistency of yarn color and the quality of the final product. Currently, the color difference inspection technologies available on the market mainly include manual visual inspection and automatic inspection based on colorimeters. Manual visual inspection relies on the experience of the inspectors and is easily affected by fatigue and subjective judgment, leading to unstable inspection results. While automatic inspection methods based on colorimeters can provide relatively objective measurement results, their detection range is limited and it is difficult to adapt to large-scale, continuous inspection needs.

[0003] Existing yarn color difference detection methods often fail to provide real-time and large-scale accurate monitoring of the yarn dyeing process and the ability to adjust yarn dyeing production based on image analysis results. Furthermore, existing yarns are not easily graded for quality after production, resulting in a lack of assurance regarding product consistency. Therefore, we propose a cotton yarn color difference detection system based on image data. Summary of the Invention

[0004] The purpose of this invention is to provide a cotton yarn color difference detection system based on image data to solve the problems mentioned in the background art. To achieve the above objective, this invention provides the following technical solution: a cotton yarn color difference detection system based on image data, including an acquisition module, an image detection module, a matching module, a color difference analysis module, a storage library, an alarm module, and a color difference diagnosis module;

[0005] Acquisition module: Real-time acquisition of image information of the wound yarn through multiple sets of high-speed cameras, obtaining yarn detection images captured by the high-speed cameras in real time and sending them to the matching module. At the same time, according to the preset acquisition interval, a set of detection images is periodically extracted from each set of high-speed cameras and sent to the image analysis module.

[0006] Image detection module: acquires the image sharpness evaluation value of the detected image, and adjusts the environmental parameters according to the magnitude of the image sharpness evaluation value of the detected image;

[0007] Matching module: Matches and stitches together multiple sets of detections sent by the acquisition module at the same time into a complete analysis image;

[0008] Color difference analysis module: Performs color difference analysis on the analyzed image, calculates the color difference deviation value of the analyzed image, and judges the degree of yarn color difference based on the magnitude of the color difference deviation value, and performs alarm and quality rating;

[0009] Storage library: Stores data such as test images, standard images, environmental parameter sets, and color difference deviation value diagnostic parameter sets;

[0010] Alarm module: When insufficient image clarity or color difference deviation exceeds a preset threshold is detected, it will issue a voice announcement and control the relevant equipment to stop working;

[0011] Color difference diagnosis module: Establishes a color difference deviation value diagnosis model and a color deviation increase total product diagnosis model to diagnose the cause parameters of color difference.

[0012] Preferably, the data acquisition module operates as follows:

[0013] Multiple high-speed cameras are installed and synchronously controlled at the outlet of the yarn dyeing machine and between the winding machine to monitor all areas that need to be monitored during yarn winding. The real-time images of the yarn captured by the high-speed cameras are obtained and marked as detection images i, where i represents the label of the detection image, i=1, 2, 3, ..., n. The detection images collected by the multiple high-speed cameras are sent to the matching module. At the same time, the image clarity detection interval is preset. According to the preset acquisition interval, a set of detection images is periodically extracted from each set of high-speed cameras and sent to the image analysis module.

[0014] Preferably, the image detection module operates as follows:

[0015] Some objects with fixed colors in the detection image are marked as detection objects. The detection object image is extracted from the detection image and converted from the RGB color space to the HSV color space. Several recognition points are evenly set in the detection object image. For each recognition point, its H, S and V channel values ​​in the HSV color space are obtained. For all recognition points in the detection object image, the sum of the H, S and V channel values ​​of each channel is calculated. Then, the average value of each channel value is calculated as the H, S and V channel values ​​of the detection object image.

[0016] The image of the detected object is matched with all standard images of the detected objects in the storage library to obtain the standard image of the corresponding detected object. The H, S, and V channel values ​​of the standard image of the corresponding detected object are extracted. The deviation values ​​between the H, S, and V channel values ​​of the detected object image and the H, S, and V channel values ​​of the corresponding standard image are calculated and labeled as the H channel deviation value HPi, the S channel deviation value SPi, and the V channel deviation value VPi of the detected image. The image sharpness evaluation value TXi is obtained using the formula: TXi = HPi*a1 + SPi*a2 + VPi*a3, where a1, a2, and a3 are preset weight coefficients.

[0017] The image sharpness evaluation value TXI is compared with the preset image sharpness deviation threshold. If the image sharpness evaluation value TXI is large, it is substituted into the environmental parameter model to obtain the set of environmental parameters affecting image sharpness. The set of environmental parameters affecting image sharpness and the current time are sent to the storage library. At the same time, an environmental adjustment command is generated and sent to the alarm module. After receiving the environmental adjustment command, the alarm module will broadcast the environmental adjustment voice. After hearing the environmental adjustment voice broadcast, the staff will retrieve the set of environmental parameters affecting image sharpness from the storage library according to the start time of the environmental adjustment voice broadcast and perform corresponding processing based on the set of environmental parameters affecting image sharpness.

[0018] Preferably, the process of establishing the environmental parameter model is as follows:

[0019] A series of environmental parameter data that may affect image sharpness were collected from historical records, including light intensity, air humidity, ambient temperature, lens dust content, and camera technical parameters such as aperture size, focal length, and ISO setting. The Pearson correlation coefficient statistical method was used to quantify the correlation between each environmental parameter and the image sharpness evaluation value TXI. Based on the results of the correlation analysis, the three environmental parameters with the highest correlation were selected as the input feature set of the model to construct a neural network model. The collected data was used to train the neural network model, and the model performance was optimized by adjusting the network weights and biases. After training and adjustment, the environmental parameter model was obtained.

[0020] Preferably, the matching module operates as follows:

[0021] Edge feature points are extracted from the detection images sent by the acquisition module at the same time. Feature points in detection images captured by different high-speed cameras are matched to establish spatial correspondence between the detection images. Based on the spatial correspondence between the detection images, the geometric transformation relationship between the detection images is calculated. Using the geometric transformation relationship between the detection images, the detection images sent by the acquisition module at the same time are aligned to a common reference coordinate system. The overlapping areas of the detection images are fused at the pixel level. Finally, the aligned and fused multiple sets of detection images are stitched together into a complete large-scale image and marked as analysis image t. Analysis image t is then sent to the color difference analysis module; where t represents the time number of the analysis image, t=1, 2, 3, ..., n.

[0022] Preferably, the color difference analysis module operates as follows:

[0023] Several color difference detection points are uniformly set in the image analysis. RGB color recognition is performed on each color difference detection point to obtain the RGB color value of each color difference detection point. By comparing the RGB color value of each color difference detection point with the RGB color value of the target yarn, the difference is calculated. Color difference detection points with positive difference values ​​are marked as brighter detection points, and those with negative difference values ​​are marked as darker detection points. A preset RGB value deviation range is set for each brighter detection point and each darker detection point. Each brighter detection point is matched with the preset deviation RGB value range of the brighter detection point, and the same applies to each darker detection point.

[0024] Brighter detection points that are outside the preset RGB value deviation range of the brighter detection points are marked as secondary brighter points. The number of secondary brighter points, PLc_t, is counted, where c represents the preference of the color difference detection points, c = 1, 2, 3, ..., n. Simultaneously, the sum of the distances of all secondary brighter points from the preset RGB value deviation range of the brighter detection points is obtained and marked as the total brighter value, PLZ_c_t. Similarly, the number of darker secondary points, PA_c_t, and the total darker value, PAZ_c_t, are obtained using the formula: The color difference deviation value FPZ of the analyzed image is obtained. t Where w1 and w2 are preset weighting coefficients, the color difference deviation value FPZ of the analyzed image is used. t Compare the color difference deviation alarm value of the analyzed image with the preset color difference deviation alarm value. If the color difference deviation value of the analyzed image is FPZ... t If the deviation is large, a yarn color difference warning command is generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning command, it will issue a color difference voice broadcast and simultaneously control the yarn winding machine and yarn dyeing machine to stop working. At the same time, it will analyze the color difference deviation value FPZ of the image. t When the value is large, the color difference deviation (FPZ) of the analyzed image will be used. t Send to the color difference diagnosis module.

[0025] Preferably, if the color difference deviation of the analyzed image is small, the processing method is as follows:

[0026] If we analyze the color difference deviation value FPZ of the image t If the value is small, the color difference deviation value of the analyzed image is substituted into the curve graph between the color difference deviation value and time. The values ​​of the color difference deviation value of the analyzed image at each time point are connected to obtain the image color deviation threshold line. A preset analysis image color deviation threshold is drawn on the image color deviation threshold line. The area enclosed by the image color deviation threshold line and the preset analysis image color deviation threshold is marked as the color deviation increase area. All color deviation increase areas are counted and marked as the color deviation increase total product. A preset color deviation increase total product alarm threshold is set, and the color deviation increase total product is compared with the preset color deviation increase total product alarm threshold.

[0027] When the total color deviation exceeds the preset alarm threshold, a yarn color difference deviation warning command will be generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning, it will broadcast a voice message about the increased yarn color difference and control the yarn winding machine and yarn dyeing machine to stop working. At the same time, the total color deviation will be sent to the color difference diagnosis module.

[0028] When the yarn winding machine finishes, if the total color deviation is still less than the preset total color deviation alarm threshold, three yarn quality levels are set, and the total color deviation is matched with the total color deviation range corresponding to the three set yarn quality levels to obtain the quality level of the finished yarn.

[0029] Preferably, the color difference diagnosis module operates as follows:

[0030] Using the environmental parameter model as a reference model, a diagnostic model for color difference deviation and a diagnostic model for color deviation product of the analyzed image are similarly established. The color difference deviation or color deviation product of the analyzed image received by the color difference diagnosis module is substituted into the corresponding diagnostic model to obtain the diagnostic parameter set for color difference deviation or color deviation product of the analyzed image. The diagnostic parameter set for color difference deviation or color deviation product of the analyzed image is then sent to the storage library. Staff can retrieve the diagnostic parameter set for color difference deviation or color deviation product of the analyzed image from the storage library according to the start time of the voice broadcast for color difference or the start time of the voice broadcast for increased yarn color difference, and process the yarn color difference.

[0031] Compared with the prior art, the beneficial effects of the present invention are:

[0032] (1) The cotton yarn color difference detection system based on image data acquires the image information of the wound yarn in real time through multiple sets of high-speed cameras to obtain the detection image during the yarn production process. The image clarity evaluation value of the detection image is analyzed to obtain the clarity of the detection image. If it is not clear, the image clarity evaluation value is substituted into the environmental parameter model to obtain the set of environmental parameters that affect the image clarity. Then, the environmental parameters are optimized according to the set of environmental parameters that affect the image clarity to ensure the clarity of the image acquisition and provide a more accurate analysis image for subsequent yarn detection and analysis. At the same time, the images captured by multiple high-definition cameras can be stitched together through the matching module, which helps to comprehensively detect the color difference of the yarn.

[0033] (2) The color difference detection system for cotton yarn based on image data has a color difference analysis module that analyzes the color difference deviation value of the analyzed image and compares it with the corresponding preset alarm value. It implements two alarm mechanisms: instantaneous color difference alarm and cumulative alarm. The instantaneous color difference alarm targets the color difference deviation at a single point in time. When the color difference deviation value of the analyzed image exceeds the threshold, an alarm is immediately triggered and production is suspended to reduce defective products. The cumulative alarm focuses on the cumulative effect of color difference over time. If the total color deviation exceeds the set threshold, an alarm will also be triggered to ensure long-term quality control. Finally, the yarn is graded according to the accumulation of the total color deviation, thereby effectively distinguishing and guaranteeing the quality of yarn products.

[0034] (3) The cotton yarn color difference detection system based on image data can easily diagnose the parameter set of color difference through the color difference diagnosis module. Based on the parameter set of color difference, it is easy for staff to carry out targeted color difference processing, which improves the efficiency of yarn production adjustment and thus speeds up the overall efficiency of yarn production. Attached Figure Description

[0035] Figure 1 This is a flowchart of the present invention;

[0036] Figure 2 This is a graph showing the curves connecting the color deviation thresholds of the present invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] Please see Figure 1 - Figure 2 Example 1:

[0039] This invention provides a cotton yarn color difference detection system based on image data, including an acquisition module, an image detection module, a matching module, a color difference analysis module, a storage library, an alarm module, and a color difference diagnosis module;

[0040] The acquisition module uses multiple sets of high-speed cameras to acquire image information of the wound yarn in real time. The specific process is as follows:

[0041] Multiple high-speed cameras are installed and synchronously controlled at the outlet of the yarn dyeing machine and between the winding machine to monitor all areas that need to be monitored during yarn winding. The real-time images of the yarn captured by the high-speed cameras are obtained and marked as detection images i, where i represents the label of the detection image, i=1, 2, 3, ..., n. The detection images collected by the multiple high-speed cameras are sent to the matching module. At the same time, the image clarity detection interval is preset. According to the preset acquisition interval, a set of detection images is periodically extracted from each set of high-speed cameras and sent to the image analysis module.

[0042] The image detection module performs image sharpness detection on the detection images extracted by each group of high-speed cameras within the acquisition interval, specifically as follows:

[0043] Some objects with fixed colors in the detection image are marked as detection objects. The detection object image is extracted from the detection image and converted from the RGB color space to the HSV color space. Several recognition points are evenly set in the detection object image. For each recognition point, its H (hue), S (saturation), and V (brightness) channel values ​​in the HSV color space are obtained. For all recognition points in the detection object image, the sum of the values ​​of each channel (H, S, V) is calculated. Then, the average value of each channel value is taken as the H, S, and V channel values ​​of the detection object image.

[0044] The database stores standard images of all detected objects. The detected object image is matched with all the standard images in the database to obtain the corresponding standard image. The H, S, and V channel values ​​of the corresponding standard image are extracted. The deviation values ​​between the H, S, and V channel values ​​of the detected image and the corresponding standard image are calculated and labeled as the detected image H channel deviation value HPi, the detected image S channel deviation value SPi, and the detected image V channel deviation value VPi. The image sharpness evaluation value TXi is obtained using the formula: TXi = HPi*a1 + SPi*a2 + VPi*a3, where a1, a2, and a3 are preset weighting coefficients; a larger image sharpness evaluation value TXi indicates lower image sharpness.

[0045] The image sharpness evaluation value TXI is compared with a preset image sharpness deviation threshold. If the image sharpness evaluation value TXI is large, it indicates insufficient image sharpness. In this case, the image sharpness evaluation value TXI is substituted into the environmental parameter model to obtain the set of environmental parameters affecting image sharpness. The set of environmental parameters affecting image sharpness, along with the current time, is sent to the storage library. At the same time, an environmental adjustment command is generated and sent to the alarm module. After receiving the environmental adjustment command, the alarm module will broadcast an environmental adjustment voice announcement. After hearing the environmental adjustment voice announcement, the staff can retrieve the set of environmental parameters affecting image sharpness from the storage library based on the start time of the environmental adjustment voice announcement and perform corresponding processing based on the set of environmental parameters affecting image sharpness to improve the sharpness of the high-speed camera image acquisition. If the image sharpness evaluation value TXI is small, the current image sharpness of the high-speed camera is considered satisfactory, and no additional measures are required.

[0046] The process of establishing the environmental parameter model is as follows: a series of environmental parameter data that may affect image clarity were collected from historical records. These data include light intensity, air humidity, ambient temperature, lens dust content, and camera technical parameters such as aperture size, focal length, and ISO setting. The historical data of the collected environmental parameters were preprocessed.

[0047] Based on the preprocessed data, the Pearson correlation coefficient statistical method was used to quantify the correlation between each environmental parameter and the image sharpness evaluation value TXI. Based on the results of the correlation analysis, the top three environmental parameters with the highest correlation were selected as the input feature set of the model to construct a neural network model. Subsequently, the collected data was used to train the neural network model. By adjusting the weights and biases of the network, the performance of the model was optimized. After training and adjustment, the environmental parameter model was obtained.

[0048] It should be noted that the image detection module can perform clarity detection on the detection images captured by multiple high-speed cameras in the acquisition module to ensure the clarity of the detection images acquired by the acquisition module. This provides a clearer detection image for the subsequent color difference analysis module, thereby improving the accuracy of the color difference analysis. Example 2:

[0049] Matching module: Matches and stitches the detection images sent by the acquisition module at the same time into a complete large-area image. This image covers the monitoring area of ​​all high-speed cameras. The specific process is as follows:

[0050] Preprocessing is performed on the detection images sent by the acquisition module at the same time, including noise reduction, distortion correction, and contrast enhancement, to improve image quality. Edge feature points are extracted from the detection images, including edge texture and edge color. After the edge feature points are extracted, the SIFT feature matching algorithm is used to match the feature points in the detection images captured by different high-speed cameras, establishing a spatial correspondence between the detection images. Based on the spatial correspondence between the detection images, the geometric transformation relationship between the detection images is calculated, including translation, rotation, and scaling. Using the geometric transformation relationship between the detection images, the detection images sent by the acquisition module at the same time are aligned to a common reference coordinate system. Pixel-level fusion processing is performed on the overlapping areas of the detection images to eliminate stitching artifacts. Finally, the aligned and fused multiple sets of detection images are stitched together into a complete large-area image, labeled as analysis image t, and sent to the color difference analysis module. Here, t represents the time number of the analysis image, t=1, 2, 3, ..., n; the analysis image covers the monitoring area of ​​all high-definition cameras.

[0051] It should be noted that the matching module stitches together the detection images captured by multiple high-speed cameras to form a more comprehensive yarn production analysis image, which facilitates the subsequent color difference analysis module to conduct comprehensive detection of the yarn on the entire production line.

[0052] The color difference analysis module performs color difference analysis on the image to obtain the color difference deviation value. Then, based on the magnitude of the color difference deviation value, it determines whether the degree of yarn color difference requires an alarm. Simultaneously, based on the correlation between the color difference deviation value and time, it rates the yarn quality. The specific process is as follows:

[0053] Several color difference detection points are uniformly set in the image analysis. RGB color recognition is performed on each color difference detection point to obtain the RGB color value of each color difference detection point. By comparing the RGB color value of each color difference detection point with the RGB color value of the target yarn, the difference is calculated. Color difference detection points with positive difference values ​​are marked as brighter detection points, and those with negative difference values ​​are marked as darker detection points. A preset RGB value deviation range is set for each brighter detection point and each darker detection point. Each brighter detection point is matched with the preset deviation RGB value range of the brighter detection point, and the same applies to each darker detection point.

[0054] Brighter detection points that are outside the preset RGB value deviation range of the brighter detection points are marked as secondary brighter points. The number of secondary brighter points, PLc_t, is counted, where c represents the preference of the color difference detection points, c = 1, 2, 3, ..., n. Simultaneously, the sum of the distances of all secondary brighter points from the preset RGB value deviation range of the brighter detection points is obtained and marked as the total brighter value, PLZ_c_t. Similarly, the number of darker secondary points, PA_c_t, and the total darker value, PAZ_c_t, are obtained using the formula: The color difference deviation value FPZ of the analyzed image is obtained. t Where w1 and w2 are preset weighting coefficients, the color difference deviation value FPZ of the analyzed image is used. t Compare the color difference deviation alarm value of the analyzed image with the preset color difference deviation alarm value. If the color difference deviation value of the analyzed image is FPZ... t If the deviation is large, a yarn color difference warning command is generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning command, it will issue a color difference voice broadcast and simultaneously control the yarn winding machine and yarn dyeing machine to stop working. At the same time, it will analyze the color difference deviation value FPZ of the image. t When the value is large, the color difference deviation (FPZ) of the analyzed image will be used. t Send to the color difference diagnosis module;

[0055] If we analyze the color difference deviation value FPZ of the image t If the value is small, the color difference deviation of the analyzed image is substituted into the curve of color difference deviation versus time, such as... Figure 2 The color difference deviation values ​​of the analyzed image at various time points are connected to obtain the image color deviation threshold line. A preset image color deviation threshold is drawn on the image color deviation threshold line. The area enclosed by the image color deviation threshold line is marked as the color deviation increase area. All color deviation increase areas are counted and marked as the color deviation increase total product. The color deviation increase total product may increase with time. A preset color deviation increase total product alarm threshold is set, and the color deviation increase total product is compared with the preset color deviation increase total product alarm threshold.

[0056] When the total color deviation exceeds the preset alarm threshold, a yarn color difference deviation warning command will be generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning, it will broadcast a voice message about the increased yarn color difference and control the yarn winding machine and yarn dyeing machine to stop working. At the same time, the total color deviation will be sent to the color difference diagnosis module.

[0057] When the yarn winding machine finishes, if the total color deviation is still less than the preset total color deviation alarm threshold, three yarn quality levels are set: Level 1, Level 2, and Level 3. Each yarn quality level corresponds to a range of total color deviation. The higher the level, the larger the range of total color deviation, the greater the color difference, and the worse the quality. The total color deviation is matched with the range of total color deviation corresponding to the three yarn quality levels to obtain the quality level of the finished yarn.

[0058] It should be noted that the color difference analysis module analyzes the color difference deviation value of the analyzed image and compares it with the corresponding preset alarm value, implementing two alarm mechanisms: instantaneous color difference alarm and cumulative alarm. The instantaneous color difference alarm targets the color difference deviation at a single point in time. When the color difference deviation value of the analyzed image exceeds the threshold, an alarm is immediately triggered and production is suspended to reduce defective products. The cumulative alarm focuses on the cumulative effect of color difference over time. If the total color deviation exceeds the set threshold, an alarm will also be triggered to ensure long-term quality control. Finally, based on the accumulation of the total color deviation, the yarn is graded for quality, thereby effectively distinguishing and ensuring the quality of yarn products.

[0059] The color difference diagnosis module works as follows:

[0060] Using the establishment of the environmental parameter model as a reference model establishment method, similarly, a diagnostic model for color difference deviation value and a diagnostic model for color deviation product of the analyzed image are established.

[0061] The color difference deviation value or color deviation product of the analysis image received by the color difference diagnosis module is substituted into the corresponding diagnosis model to obtain the color difference deviation value diagnosis parameter set or color deviation product diagnosis parameter set of the analysis image. The color difference deviation value diagnosis parameter set or color deviation product diagnosis parameter set is then sent to the storage library. Staff can retrieve the color difference deviation value diagnosis parameter set or color deviation product diagnosis parameter set from the storage library according to the start time of the color difference voice broadcast or the start time of the yarn color difference increase voice broadcast to process the yarn color difference.

[0062] It should be noted that the color difference diagnosis module can easily diagnose the parameter set that causes color difference. Based on the parameter set, staff can easily carry out targeted color difference processing, which improves the efficiency of yarn production adjustments and thus speeds up the overall efficiency of yarn production.

[0063] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A cotton yarn color difference detection system based on image data, comprising an acquisition module, an image detection module, a matching module, a color difference analysis module, a storage library, an alarm module, and a color difference diagnosis module, characterized in that: Acquisition module: Real-time acquisition of image information of the wound yarn through multiple sets of high-speed cameras, obtaining yarn detection images captured by the high-speed cameras in real time and sending them to the matching module. At the same time, according to the preset acquisition interval, a set of detection images is periodically extracted from each set of high-speed cameras and sent to the image analysis module. Image detection module: acquires the image sharpness evaluation value of the detected image, and adjusts the environmental parameters according to the magnitude of the image sharpness evaluation value of the detected image; Matching module: Matches and stitches together multiple sets of detections sent by the acquisition module at the same time into a complete analysis image; Color difference analysis module: Performs color difference analysis on the analyzed image, calculates the color difference deviation value of the analyzed image, and judges the degree of yarn color difference based on the magnitude of the color difference deviation value, and performs alarm and quality rating; The color difference analysis module works as follows: Several color difference detection points are uniformly set in the image analysis. RGB color recognition is performed on each color difference detection point to obtain the RGB color value of each color difference detection point. By comparing the RGB color value of each color difference detection point with the RGB color value of the target yarn, the difference is calculated. Color difference detection points with positive difference values ​​are marked as brighter detection points, and those with negative difference values ​​are marked as darker detection points. A preset RGB value deviation range is set for each brighter detection point and each darker detection point. Each brighter detection point is matched with the preset deviation RGB value range of the brighter detection point, and the same applies to each darker detection point. Mark the brighter detection points that are outside the preset RGB value deviation range of the brighter detection points as secondary brighter points, and count the number of secondary brighter points. c represents the color difference detection point number, c=1, 2, 3, ..., n3; simultaneously, the distance values ​​of all the brighter secondary points from the preset RGB value deviation range of the brighter detection point are obtained and marked as the total brighter value. Similarly, the number of darker sub-points can be obtained. Compared to the total dark value Using the formula: , Obtain the color difference deviation value of the analyzed image. Where w1 and w2 are preset weighting coefficients, the color difference deviation of the analyzed image will be used. The color difference deviation alarm value of the analyzed image is compared with the preset color difference deviation value. If the color difference deviation value of the analyzed image is... If the deviation is large, a yarn color difference warning command is generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning command, it will issue a color difference voice broadcast and simultaneously control the yarn winding machine and yarn dyeing machine to stop working. At the same time, it will analyze the color difference deviation value of the image. When the value is large, the color difference deviation of the analyzed image will be analyzed. Send to the color difference diagnosis module; where t represents the time number of the image being analyzed, t=1, 2, 3, ..., n2; Storage library: Stores test images, standard images, environmental parameter sets, and color difference deviation value diagnostic parameter sets; Alarm module: When insufficient image clarity or color difference deviation exceeding a preset threshold is detected, a voice announcement is made and the relevant equipment is controlled to stop working; Color difference diagnosis module: A color difference deviation diagnostic model and a color deviation product diagnostic model are established to diagnose the cause parameters of color difference.

2. The cotton yarn color difference detection system based on image data according to claim 1, characterized in that: The data acquisition module works as follows: Multiple high-speed cameras are installed and synchronously controlled between the yarn dyeing machine outlet and the winding machine to monitor all areas requiring monitoring during yarn winding. Real-time images of the yarn captured by the high-speed cameras are obtained and labeled as detection images i, where i represents the label of the detection image, i=1, 2, 3, ..., n1. The detection images collected by the multiple high-speed cameras are sent to the matching module. At the same time, the image clarity detection interval is preset. According to the preset acquisition interval, a set of detection images is periodically extracted from each set of high-speed cameras and sent to the image analysis module.

3. The cotton yarn color difference detection system based on image data according to claim 2, characterized in that: The image detection module works as follows: Some objects with fixed colors in the detection image are marked as detection objects. The detection object image is extracted from the detection image and converted from the RGB color space to the HSV color space. Several recognition points are evenly set in the detection object image. For each recognition point, its H, S and V channel values ​​in the HSV color space are obtained. For all recognition points in the detection object image, the sum of the H, S and V channel values ​​of each channel is calculated. Then, the average value of each channel value is calculated as the H, S and V channel values ​​of the detection object image. The image of the detected object is matched with all standard images of the detected objects in the storage library to obtain the standard image of the corresponding detected object. The H, S, and V channel values ​​of the standard image of the corresponding detected object are extracted. The deviation values ​​between the H, S, and V channel values ​​of the detected object image and the H, S, and V channel values ​​of the corresponding standard image are calculated and labeled as the H channel deviation value HPi, the S channel deviation value SPi, and the V channel deviation value VPi of the detected image. The image sharpness evaluation value TXi is obtained using the formula: TXi = HPi × a1 + SPi × a2 + VPi × a3, where a1, a2, and a3 are preset weight coefficients. The image sharpness evaluation value TXI is compared with the preset image sharpness deviation threshold. If the image sharpness evaluation value TXI is large, it is substituted into the environmental parameter model to obtain the set of environmental parameters affecting image sharpness. The set of environmental parameters affecting image sharpness and the current time are sent to the storage library. At the same time, an environmental adjustment command is generated and sent to the alarm module. After receiving the environmental adjustment command, the alarm module will broadcast the environmental adjustment voice. After hearing the environmental adjustment voice broadcast, the staff will retrieve the set of environmental parameters affecting image sharpness from the storage library according to the start time of the environmental adjustment voice broadcast and perform corresponding processing based on the set of environmental parameters affecting image sharpness.

4. The cotton yarn color difference detection system based on image data according to claim 3, characterized in that: The process of establishing the environmental parameter model is as follows: A series of environmental parameter data affecting image sharpness were collected from historical records, including: light intensity, air humidity, ambient temperature, lens dust content, and camera technical parameters, including aperture size, focal length, and ISO setting. The Pearson correlation coefficient method was used to quantify the correlation between each environmental parameter and the image sharpness evaluation value TXI. Based on the correlation analysis results, the three most correlated environmental parameters were selected as the input feature set for the model, thereby constructing a neural network model. The collected data was used to train the neural network model, and the model's performance was optimized by adjusting the network weights and biases. After training and adjustment, the environmental parameter model was obtained.

5. The cotton yarn color difference detection system based on image data according to claim 4, characterized in that: The matching module works as follows: Edge feature points are extracted from the detection images sent by the acquisition module at the same time. Feature points in detection images captured by different high-speed cameras are matched to establish spatial correspondence between the detection images. Based on the spatial correspondence between the detection images, the geometric transformation relationship between the detection images is calculated. Using the geometric transformation relationship between the detection images, the detection images sent by the acquisition module at the same time are aligned to a common reference coordinate system. The overlapping areas of the detection images are fused at the pixel level. Finally, the aligned and fused multiple sets of detection images are stitched together into a complete large-scale image and marked as analysis image t. Analysis image t is then sent to the color difference analysis module; where t represents the time number of the analysis image, t=1, 2, 3, ..., n2.

6. The cotton yarn color difference detection system based on image data according to claim 5, characterized in that: If the color difference deviation of the analyzed image is small, the processing method is as follows: If we analyze the color difference deviation of the image If the value is small, the color difference deviation value of the analyzed image is substituted into the curve graph between the color difference deviation value and time. The values ​​of the color difference deviation value of the analyzed image at each time point are connected to obtain the image color deviation threshold line. A preset analysis image color deviation threshold is drawn on the image color deviation threshold line. The area enclosed by the image color deviation threshold line and the preset analysis image color deviation threshold is marked as the color deviation increase area. All color deviation increase areas are counted and marked as the color deviation increase total product. A preset color deviation increase total product alarm threshold is set, and the color deviation increase total product is compared with the preset color deviation increase total product alarm threshold. When the total color deviation exceeds the preset alarm threshold, a yarn color difference deviation warning command will be generated and sent to the alarm module. When the alarm module receives the yarn color difference deviation warning, it will broadcast a voice message about the increased yarn color difference and control the yarn winding machine and yarn dyeing machine to stop working. At the same time, the total color deviation will be sent to the color difference diagnosis module. When the yarn winding machine finishes, if the total color deviation is still less than the preset total color deviation alarm threshold, three yarn quality levels are set, and the total color deviation is matched with the total color deviation range corresponding to the three set yarn quality levels to obtain the quality level of the finished yarn.

7. The cotton yarn color difference detection system based on image data according to claim 6, characterized in that: The color difference diagnosis module works as follows: Using the environmental parameter model as a reference model, a diagnostic model for color difference deviation and a diagnostic model for color deviation product of the analyzed image are similarly established. The color difference deviation or color deviation product of the analyzed image received by the color difference diagnosis module is substituted into the corresponding diagnostic model to obtain the diagnostic parameter set for color difference deviation or color deviation product of the analyzed image. The diagnostic parameter set for color difference deviation or color deviation product of the analyzed image is then sent to the storage. Staff retrieve the diagnostic parameter set for color difference deviation or color deviation product of the analyzed image from the storage according to the start time of the voice broadcast of color difference or the start time of the voice broadcast of increased yarn color difference, and process the yarn color difference.

Images