A real-time error correction method for sole stitch codes based on CDS color quantization

Abstract

The invention belongs to the technical field of sole stitch code monitoring, and specifically discloses a sole stitch code real-time error correction method based on CDS color quantization, comprising: obtaining first color information, second color information, and third color information; obtaining a color image of a target object, obtaining a first continuous value of the third color information in the color image with pixel points as units, determining the size of a detection unit by the first continuous value, dividing a coverage area by the detection unit as units, and setting a first detection threshold; selecting a judgment color according to the posture of the target object in the color image, obtaining a second continuous value of the judgment color in the coverage area by the detection unit as units; judging whether the second continuous value is greater than the first detection threshold, and if so, the target object has incomplete stitch codes; the method has the following advantages: through real-time monitoring, stitch code defects are automatically monitored and displayed, production efficiency and material utilization rate are improved, costs are reduced, product quality is ensured, and the method is suitable for the production of various cloth shoe soles.

Description

Technical Field

[0001] The invention relates to the technical field of sole stitch code monitoring, and in particular to a sole stitch code real-time error correction method based on CDS color quantization. Background Art

[0002] In automated production systems, especially when it comes to stitching soles or insoles, although these systems can significantly improve production speed and efficiency, they also bring some challenges. Although automated stitching systems optimize traditional manual and semi-automatic production processes, they still face many challenges in large-scale production to ensure that each product meets high quality standards. Especially in a fast-paced production environment, the risk of missing stitches, stitch deflection or other stitch defects increases and is difficult to display and discover in a large amount of data, affecting the overall quality of the product.

[0003] Therefore, a real-time error correction method for sole stitch code based on CDS color quantization is proposed to solve the above-mentioned problems. Summary of the invention

[0004] The present invention aims to provide a real-time error correction method for sole stitch codes based on CDS color quantization to solve or improve at least one of the above-mentioned technical problems.

[0005] In view of this, a first aspect of the present invention is to provide a real-time error correction method for sole stitch codes based on CDS color quantization.

[0006] The first aspect of the present invention provides a real-time error correction method for sole stitch codes based on CDS color quantization, comprising the following steps: pre-acquiring first color information on the surface of a target object, second color information on the surface of a platform carrying the target object, and third color information on the surface of a wire material of the target object with stitch codes; continuously photographing a color image of the target object at each moment, and transmitting it to a display end for real-time display; judging whether there is a stitch code error in the color image at the current moment based on the distribution of the first color information, the second color information, and the third color information in the color image, and generating an identification pattern if there is a stitch code error; marking the identification pattern in the color image based on the display position of the stitch code error in the color image; transmitting the marked color image to the display end to be displayed together with the color image before marking at each moment.

[0007] In any of the above technical solutions, the determination of whether there is a pin code error in the color image at the current moment includes: determining whether the third color information in the color image completely covers the first color information, and treating incomplete coverage as a pin code error.

[0008] In any of the above technical solutions, the real-time error correction method for sole stitch codes based on CDS color quantization further includes: constructing a model for determining whether to generate the identification pattern; and the model includes: ; wherein said is the calculation score output by the model, used to determine whether to generate the identification pattern; is a fuzzy inference function, used to affect the calculation score of the model output; Characterizing the consistency between the color information of the wire surface in the current color image and the pre-acquired third color information; Characterizing the correlation between the pin code pattern in the current color image and the expected regularity and structural continuation; Characterizing the consistency between the posture of the target object in the current color image and the standard posture; Characterizes the completeness of the stitch code wire covering the entire target object.

[0009] In any of the above technical solutions, by setting The threshold T matched with the color image is used to determine whether the color image at each moment generates a logo pattern; when the When the value of is greater than the threshold value T, a logo pattern is generated in the color image at the current moment; when When the value of is less than or equal to the threshold T, no identification pattern is generated.

[0010] In any of the above technical solutions, the threshold T is dynamically adjusted by using Bayesian optimization or grid search by comparing historical color image data with the error status of the needle code in production.

[0011] In any of the above technical solutions, the surface of the target object is divided into n detection units, and the detection units are numbered as i; and the formula of the model is updated and determined on the multiple detection units by the following steps: , , and stated As input variables, the membership function of each input variable is obtained respectively; the fuzzy output of each membership function is obtained by fuzzy rules and membership reasoning, and all the fuzzy outputs are aggregated to generate fuzzy results; the formula of the model is adjusted according to the fuzzy results and the actual needle code error detection results, and the formula of the model is updated to ; Logarithmic function, function.

[0012] In any of the above technical solutions, The membership function of is: ; wherein said is the third color information acquired in advance; is the color information of the wire surface in the current color image.

[0013] In any of the above technical solutions, The membership function of is: ; wherein said is the local area of ​​the needle code extracted from the color image; the CNN output is the probability value of needle code pattern recognition.

[0014] In any of the above technical solutions, The membership function of is: ; wherein said is the posture matching function; is the posture point predicted by the detection unit; is the actual posture point of the detection unit; N is the number of detection units containing posture points.

[0015] In any of the above technical solutions, The membership function of is: ; wherein said The pin code area of ​​the current frame color image; The stitch code area of ​​the previous frame of color image; DTW is to calculate the continuity difference between two frames.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] By acquiring color information of the target object surface, the carrier platform, and the needle code wire, the accuracy of subsequent analysis is ensured. The color image of the target object is continuously captured and transmitted to the display end in real time, allowing the monitoring personnel to instantly observe every detail of the production process. By analyzing the color distribution in the color image, needle code errors, such as wire breakage or missing needles, can be accurately judged and identified, and the corresponding identification pattern is generated in the image. These marked images are then transmitted to the display end and displayed together with the images before marking, providing an intuitive comparative view, allowing operators to quickly understand and respond to problems in the production process.

[0018] It significantly improves the efficiency and accuracy of needle code monitoring, makes remote monitoring possible, and greatly improves the pertinence and real-time nature of production data processing. In addition, it reduces production inefficiencies and quality problems caused by manual detection and delayed response, ensuring the continuity of the production process and high quality standards of the products. Through remote automatic monitoring and real-time marking, it provides an efficient and reliable quality control solution for cloth shoe sole production, which is particularly suitable for application requirements when facing large-scale production data.

[0019] Additional aspects and advantages of embodiments according to the present invention will become apparent in the following description or may be learned through practice of embodiments according to the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0021] Figure 1 is a flow chart of the method steps of the present invention;

[0022] Figure 2 is a flow chart of sub-steps of step S103 in the method of the present invention;

[0023] Figure 3 It is a schematic diagram of a target object of the present invention in an upright posture;

[0024] Figure 4 A schematic diagram of a target object of the present invention in an inverted posture;

[0025] Figure 5 It is a schematic diagram of color calibration of the present invention;

[0026] Figure 6 is a system structure block diagram of the present invention;

[0027] Figure 7 The figure is a schematic diagram of the structure of an electronic device of the present invention. DETAILED DESCRIPTION

[0028] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0029] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0030] See also Figure 1-Figure 7 , the following describes a real-time error correction method for sole stitch codes based on CDS color quantization according to some embodiments of the present invention.

[0031] The embodiment of the first aspect of the present invention proposes a real-time error correction method for sole stitch codes based on CDS color quantization. In some embodiments of the present invention, Figure 1-Figure 5 As shown, the method comprises the following steps:

[0032] S101, pre-acquire first color information of the surface of the target object, second color information of the surface of the platform carrying the target object, and third color information of the surface of the wire material of the stitch target object. The target object may be an insole or a cloth sole of a cloth shoe, and the platform may be a sewing machine or a testing platform of the stitch target object.

[0033] Here, the first color information refers to the color of the insole or the cloth sole of the cloth shoe. In the automatic sewing or detection system, this color information is used to identify and distinguish different materials or different product parts; the second color information refers to the surface color of the sewing machine or the detection table. This information is used for background distinction, helping the system to identify the boundary between the object and the background; the third color information refers to the color of the thread used for sewing. This information helps the automatic sewing machine to select the appropriate thread for operation.

[0034] As can be seen from the above, the second color information (platform color) and the third color information (wire color) are combined to ensure the accuracy of sewing or detection operations. For example, in the automatic sewing process, by comparing the color of the target object with the color of the wire, the machine can correctly adjust the sewing path, avoid color conflicts, and improve the aesthetics and quality of the product; the second color information is used when the system positions and aligns objects. By ensuring that the color of the platform surface has sufficient contrast with the target object, the system can more accurately capture and analyze images, thereby improving operational accuracy; the color information of the wire is combined with the color information of the target object to achieve visual harmony and technical accuracy. For example, the system can automatically select the most matching wire color based on the color of the target object, or adjust the wire color when performing special sewing effects.

[0035] S102, continuously photographing a color image of the target object at each moment, and transmitting the image to a display terminal for real-time display.

[0036] Here, high-resolution cameras are used to continuously capture color images of the target object (such as the sole) during the coding process, ensuring that every step from start to finish is recorded. This helps capture any anomalies or deviations that occur during the process. The captured color images are instantly transmitted to a display such as a monitoring screen or control system interface. This allows operators or automated systems to view the production process in real time, allowing them to quickly identify and respond to possible problems.

[0037] As mentioned above, continuous image capture provides a continuous data stream for subsequent image processing and analysis, which is the basis for real-time monitoring and error correction. The instant transmission of images allows decisions on the production line to be made based on the latest visual information, thereby increasing response speed and processing accuracy.

[0038] S103, judging whether there is a needle code error in the color image at the current moment according to the distribution of the first color information, the second color information and the third color information in the color image, and generating a marking pattern if there is a needle code error.

[0039] Here, the main color on the surface of the target object is identified, usually the basic color of the sole or insole. This helps determine the area where the stitching should be performed and provides a background reference for the identification of stitching errors. It helps the system distinguish the target object from its environment or background, such as the color of the conveyor belt or workbench on the production line. This distinction is a prerequisite for accurate image processing. Interruptions in the continuity of the wire color may indicate stitching errors.

[0040] As can be seen from the above, it provides the system with basic identification of the target area, which is compared with the third color information (wire color) to determine the quality and integrity of the needle code. It clarifies the boundary of the target object, so as to more accurately analyze the needle code status inside the target object. It is used in combination with the first color information to detect and confirm the accuracy and integrity of the needle code operation.

[0041] When a pin code error is found in the color image based on the color information distribution, such as wire breakage or pin code holes, it will generate an identification pattern. This pattern serves as an intuitive visual prompt to instruct the operator or the automatic correction system to make corresponding adjustments or repairs; by accurately monitoring the color distribution in the color image, the system can reduce defective products caused by pin code errors and improve the quality control of the production process; real-time monitoring and automated error detection greatly reduce the need for manual inspection, speed up the production process, and reduce the time wasted due to delayed corrections.

[0042] Specifically, the identification pattern can be designed in different shapes or colors to distinguish different types of stitch errors. For example, a stitch break may be indicated by a red X mark, a stitch hole may be indicated by a yellow circle, and a stitch deflection may be indicated by a blue arrow indicating the deflection direction. This visual error identification makes the identification of the problem intuitive and quick. The identification pattern will appear directly at the specific location of the stitch error in the color image. The system accurately locates the error by analyzing the distribution of the first, second and third color information in the image, and then overlays the identification pattern at that location. This ensures that the repair measures can be implemented precisely to the problem area.

[0043] S301, obtaining a color image of the target object at the current moment, where the color image is an image obtained by a camera, determining a coverage area of ​​the target object in the color image based on first color information and second color information, and using image recognition to identify the size and shape of the coverage area of ​​the target object in the color image.

[0044] Here, a camera captures high-resolution color images of the current work area. These images provide visual information of the target object and its surroundings; define the color properties of the target object (such as an insole or the fabric sole of a cloth shoe). This includes hue, saturation, and brightness; define the color characteristics of the platform that carries the target object (such as a sewing machine or a testing table); and use image recognition algorithms to analyze the color image to identify and measure the size and shape of the area covered by the target object in the image.

[0045] As can be seen above, color images are the basis for all subsequent image processing and analysis steps. High-quality image acquisition helps ensure the accuracy of image recognition; by comparing with the second color information (platform color), the first color information helps the image recognition system distinguish between the target object and the background. This is a key step in determining the coverage area of ​​the target object; the second color information provides a color reference for the background, allowing the image recognition system to effectively separate the target object from the background. It helps to perform precise spatial positioning and size measurement; image recognition technology relies on clear first and second color information to perform tasks. The algorithm determines the boundaries of the object by analyzing this color information, and then calculates the specific parameters of the coverage area, such as area, perimeter, shape, etc.

[0046] In any of the above embodiments, the color image of the target object includes a first image obtained when knitting the wire and a second image obtained after the knitting is completed, and the first image and the second image are distinguished by the stitch code and the stitch code completion state; and the posture of the target object includes an upright posture and an inverted posture, and the upright posture is as follows: Figure 3 In the state shown, the sole area of ​​the shoe is on top and the heel area is on the bottom; the inverted posture is as follows Figure 4 In the state shown, the sole area of ​​the shoe is below the heel area and above, so that the front and back sides of the same pair of insoles can be distinguished.

[0047] In this embodiment, the first image acquired before the knitting process begins is mainly used to record the initial state of the sole or insole. This image captures the exact position and layout of the wire before it is applied, providing necessary reference information for the setting of the knitting machine. For example, on an automated sole production line, the first image can help the operating system detect whether the wire has been correctly loaded and positioned, thereby preventing knitting errors; the second image acquired after the knitting operation is completed shows the final state of the finished product. This image is used to evaluate the quality and overall completion of the knitting, such as the tightness of the wire and the accuracy of the pattern. When producing high-performance sports shoes, the second image can be used to ensure that each wire is accurately knitted according to the design specifications, thereby ensuring the performance and durability of the sole; in the upright posture, the sole area of ​​the sole faces up and the heel area faces down. This posture is suitable for appearance and surface quality inspection, and is often used to evaluate the color consistency and appearance defects of the sole. For example, when producing highly decorative women's soles, the upright posture can be used to ensure that the appearance design of the sole meets fashion requirements; the inverted posture makes the sole area face down and the heel area face up, which is convenient for checking the internal structure and knitting quality of the sole. For example, when producing professional hiking shoes, the inverted posture can be used to ensure the structural strength and durability of the sole.

[0048] By comparing and analyzing the first and second images, the system is able to evaluate any deviations and errors in the knitting process. This comparison can reveal whether the wire deviates from the predetermined path or whether any stitches are missed. For example, when manufacturing basketball soles with complex patterns, any small deviation may affect the performance of the sole. By comparing the images, it can be ensured that every detail meets the design specifications; using images in both upright and inverted postures, the system can fully inspect the sole from different angles to ensure that the product meets quality standards both visually and structurally. For example, when producing children's shoes, a comprehensive quality inspection can ensure that the soles meet the high standards of safety and comfort for children.

[0049] S302, obtaining the first continuous value of the third color information in units of pixels in the color image, determining the size of the detection unit by the first continuous value, the detection unit being a square unit with a side length being the length of continuous pixels of the third color information in a direction perpendicular to the needle code, dividing the coverage area in units of detection units and setting a first detection threshold, the first detection threshold being the number of detection units with the first color information between two non-adjacent detection units with the third color information in a direction perpendicular to the needle code.

[0050] Here, this step involves identifying and recording the continuous pixel points where the third color (usually the color of the wire) appears in the color image. These continuous pixel points provide preliminary data on the position and distribution of the wire. In the automatic sewing machine, by identifying the continuous distribution of the specific color wire used for sewing in the image, the initial position and possible deviation of the wire can be determined; the size of the detection unit is based on the first continuous value, that is, the length of the continuous pixel points of the wire color. The detection unit size defined in this way ensures that the color information inside each unit is uniform, which helps the subsequent quality analysis. If 50 consecutive pixels of the wire color are detected, a square detection unit with a side length of 50 pixels can be set to perform a more detailed analysis in this area; by defining the detection unit, the system can divide the entire coverage area into multiple square units, and the color inside each unit will be analyzed separately. The first detection threshold is used to evaluate whether the interval between the wires meets the predetermined standard. When making soles with complex patterns, the detection threshold is set to ensure that the distance between the individual wires is not too large or too small, thereby maintaining the integrity and beauty of the pattern.

[0051] As can be seen above, the continuous third color information (thread color) helps define the size of the detection unit, which directly affects the accuracy of subsequent detection. The boundaries of the square unit are precisely set to ensure that the color analysis within each unit is accurate; through the segmentation of the detection unit, the system can more carefully monitor and analyze the placement of each section of the thread during the knitting process. The first detection threshold further helps to determine whether the interval between the threads meets the quality standards. For example, the detection threshold may specify that no more than a certain number of units of the first color information (the color of the target object) should appear between adjacent detection units.

[0052] Specifically, the step of obtaining the first continuous value of the third color information in units of pixels in the color image includes:

[0053] Get the stitch direction of the wire on the target object, usually in the direction of the line between the sole area and the heel area.

[0054] Acquire the first continuous number of pixel points whose colors are the third color information along the direction perpendicular to the needle code; and use the first continuous number as the first continuous value.

[0055] In view of the above specific description, the purpose of determining the stitch direction is to clarify the sewing path of the wire, which is usually the direction of the line between the sole area and the heel area of ​​the target object (such as the sole). This direction guides the direction of extracting the wire color information in subsequent operations. During the sewing process of the sole, the determination of the stitch direction helps the machine to correctly arrange the wire and ensure that the layout of the wire is consistent with the design pattern, especially in complex patterns such as wavy lines or pattern designs; by collecting continuous pixels of the third color information (usually the color of the wire) perpendicular to the stitch direction, the width and consistency of the wire are determined. This continuous number is regarded as the actual thickness or width of the wire in the image. When manufacturing the soles of running shoes, by accurately measuring the continuous pixels of the wire color along the vertical direction, it can be ensured that the wire is evenly laid and the width meets the design specifications.

[0056] In any of the above embodiments, the stitch direction is the direction in which the wire material continuously appears from the sole area to the heel area on the surface of the target object.

[0057] In this embodiment, the stitch direction refers to the direction in which the wires are arranged on the sole, which usually extends from the sole area to the heel area. This direction is not only related to the aesthetics of the design, but also to the structural stability and functionality of the sole. In the production of sports shoes, the stitch direction is usually designed to be along the natural curvature of the foot, which can increase the elasticity of the sole and adapt to the pressure distribution during different sports; the continuous appearance of the wires in the stitch direction means that the wires are arranged on the sole continuously and without breakpoints.

[0058] Determining the stitch direction provides guidance for the placement of the wires, ensuring that they follow a predetermined path. This placement helps achieve the aesthetic and functional requirements of the design, such as enhancing the flexibility and support of the sole; the definition of the stitch direction allows for more accurate monitoring of the continuous appearance of the wires on the sole. By monitoring the continuity of the wires, problems in production, such as wire breaks or incorrect placement, can be discovered in a timely manner, allowing for quick adjustments or repairs; the correct definition of the stitch direction and the continuous laying of the wires directly affect the performance and durability of the sole. The correct stitch direction can improve the overall structural stability of the sole, especially in sports shoes that require high-performance support, such as basketball or football shoes.

[0059] S303, selecting a judgment color according to the posture of the target object in the color image, the judgment color can be used for the current pin code incompleteness judgment, and obtaining the second continuous value of the judgment color in the coverage area in units of detection units.

[0060] Here, the posture of the target object (such as the sole) in the color image is identified to determine how to obtain the judgment color. The posture may include upright or inverted, and each posture may expose different stitch code problems. In an automated production line, the sole may be placed under the scanning device in different postures. The system needs to identify these postures to adjust the image processing algorithm to ensure the accurate collection of color data, so as to accurately evaluate the stitch code quality; the color used to judge the integrity of the stitch code is selected according to the posture of the sole. The judgment color is usually used to distinguish different materials or stitch code states during the production process. For example, areas with broken or misaligned threads may show different colors. In complex sole designs, the judgment color can be used to identify the stitching integrity of specific areas. If a part should be a continuous blue line but is interrupted, the system identifies the color change in this area as a potential sewing error; based on the determined judgment color, the second continuous value of this color in the covered area is measured. This continuous value indicates the length of time a specific color appears continuously, and is used to evaluate whether the stitch code is continuous and unbroken. In the production inspection of soles, if the evaluation color (such as red used to identify problem areas) appears continuously in a predetermined area for more than the normal value, the system may mark this area as a quality problem area that requires further inspection or repair.

[0061] As can be seen from the above, by accurately identifying the posture of the target object, the system can select the judgment color that best suits the current posture, which can more accurately identify and evaluate the problem of the stitch code. The posture affects the visibility and interpretation of the color. The correct color selection is key to correctly determine the integrity of the stitch code; the continuity analysis of the judgment color provides a direct quantitative means for monitoring the integrity of the stitch code. By measuring the length of continuous appearance of a specific color, the system can intuitively evaluate the integrity of the stitch code and promptly discover and correct problems in production; combined with the recognition of the target object's posture and the precise measurement of the judgment color, the entire system can effectively monitor and ensure the quality standards of the production process. This method is particularly suitable for the production of products with high quality requirements, such as the production of high-end sports shoes or special soles.

[0062] Specifically, the step of obtaining the second continuous value of the judgment color in the coverage area in units of detection units includes:

[0063] Determine whether the target object has completed the needle code wire at the current moment, so as to perform detection according to different operation states.

[0064] If the execution is not completed, the color is judged to be the first color information, and the detection unit needs to be color corrected. Specifically, when more than one color information appears in a detection unit, it is completely filled with the color with the largest proportion in the detection unit, and the entire column of detection units that are all the first color information along the needle code direction is deleted to retain the column of detection units where the first color information and the second color information appear alternately; along the needle code direction, the second continuous number of detection units whose color is the first color information is obtained, and the second continuous number is used as the second continuous value.

[0065] If the execution is completed, the color of the detection unit is obtained as the third continuous number for judging the color along the direction perpendicular to the needle code, and the third continuous number is used as the second continuous value.

[0066] In view of the above specific description, it is necessary to determine whether the stitch code wire operation on the target object has been completed, because the time point of the detection directly affects the accuracy of the detection result. If the stitch code operation has not been completed, the continuity measurement may reflect the intermediate process rather than the final result. On the production automation line, the system monitors the sewing status of the wire through sensors or image recognition technology. Once the stitch code operation is detected to have stopped, it is considered that the operation at this stage has been completed. At this time, it is most appropriate to perform continuity measurement; measure the second continuous value of the judgment color in the detection unit. This value refers to the number of pixels where the judgment color appears continuously in a given area. This continuous value is used to judge the integrity and consistency of the stitch code operation. In the sole manufacturing process, if the length of the continuous appearance of the judgment color (the color of the marking wire) in a certain area is lower than the predetermined threshold, it may indicate that the wire is broken or not laid correctly according to the design pattern.

[0067] When the stitching operation is not completed, if there are multiple color information in the detection unit, the system will select the color with the largest proportion to fill the unit completely to correct and unify the color representation. This step ensures color consistency and detection accuracy when producing soles with certain specific patterns, such as stripes or patterns. Color correction ensures that each detection unit displays the dominant color and reduces visual errors caused by color mixing; the system will delete the entire column of detection units that are all first color information along the stitching direction. This is done to retain the columns of detection units with alternating first and second color information, thereby ensuring the continuity and complexity of the pattern. When producing soles with complex edges or details, removing a single color column can avoid overly monotonous or incorrect pattern reproduction, ensuring the diversity of the design and the correct implementation of the pattern; the system measures the number of detection units with continuous first color information along the stitching direction, and uses this number as the second continuous value. This value helps to evaluate the continuity of color information and the correct implementation of the pattern. In the production of high-end fashion soles, the continuity of patterns and colors is critical to maintaining the overall beauty of the design and brand recognition. By accurately measuring the number of consecutive occurrences of the first color information, it can be ensured that each part of the pattern is correctly executed according to the design specifications.

[0068] Confirming that the stitching operation has been completed is a prerequisite to ensure that the continuity data for the next evaluation is based on the final product state rather than the intermediate state. On the automated sole sewing line, the completion signal can be triggered by the sensor, indicating that all stitches have been completed as designed, and the final quality inspection can be started at this time; the number of consecutive appearances of the judged color in each detection unit is measured perpendicular to the stitching direction, that is, the third consecutive number. This data is used to evaluate the continuity and uniformity of the color on the sole. When producing sports shoes with specific color patterns, the continuity evaluation of the color can ensure that each part of the pattern is not broken or missing, thereby meeting high quality standards.

[0069] Specifically, the step of determining whether the target object has completed the stitch code wire at the current moment includes:

[0070] The second detection threshold is set according to the proportion of the above-mentioned coverage area in the first image to avoid deletion affected by different object distances and different target object models when acquiring images.

[0071] Along the direction perpendicular to the needle code, the maximum continuous number of first color information of the color of the detection unit obtained in the coverage area is taken as the third continuous value.

[0072] It is determined whether the third continuous value is less than the second detection threshold value. If so, the target object has completed the needle coding wire. If not, the target object still has a whole area that has not been needle coded, and the needle coding needs to be continued.

[0073] In view of the above specific description, the second detection threshold is used to evaluate the proportion of the coverage area in the first image. This threshold helps determine whether the stitching operation is completed and whether the image is suitable for subsequent analysis and quality inspection. The setting of this threshold takes into account the impact of different object distances and different target object models on image acquisition to ensure data consistency and comparability. On the automated sole manufacturing line, the proportion of the stitching area in the image will vary depending on the sole size and design. By setting an adaptable detection threshold, the system can be applied to soles of different sizes and models to ensure the accuracy of each image analysis; by analyzing the proportion of the coverage area in the first image, the system can evaluate whether the image of the target object is complete and whether the stitching wire operation has reached the expected completion state. Proportion analysis is an intuitive indicator to determine whether the production stage is completed. When stitching complex patterns, the image proportion can directly reflect the stitching progress. For example, if the stitching should cover 80% of the sole area, and the current image only shows 50% coverage, this may indicate that the stitching operation has not been completed.

[0074] In the coverage area, the maximum number of consecutive occurrences of the first color information (usually representing the stitch code wire or base color) is measured in the direction perpendicular to the stitch code. This continuous number reflects the continuity and integrity of the stitch code and is a key indicator for judging whether the stitch code operation is completed. When making soles with complex patterns, continuous color information can indicate whether the pattern has been completed as designed. For example, if the design requires that a certain area should display red wire continuously, the system confirms whether the predetermined design requirements are met by calculating the maximum number of consecutive occurrences of this color.

[0075] The third continuous value refers to the maximum number of consecutive occurrences of the first color information (representing the stitch code wire or background color) in the coverage area along the direction perpendicular to the stitch code. This measurement reflects the continuity of the stitch code wire and is used to evaluate the integrity of the stitch code. When producing patterned soles, it ensures that the pattern is continuous without gaps. Measuring the continuity of pattern colors, such as the red line should appear continuously to a certain length, can directly reflect the accuracy of production; the second detection threshold is a pre-set standard used to compare the third continuous value. This threshold is based on product design requirements and quality standards to help determine whether the stitch code operation has reached the expected completion state. In high-end clothing production, the stitch code in the details needs to be very precise. Setting a detection threshold for delicate patterns can ensure that every detail meets the design requirements, thereby ensuring the high quality of the product.

[0076] Specifically, the second detection threshold is set according to the proportion of the above-mentioned coverage area in the first image.

[0077] In any of the above embodiments, when the target object needle code wire material is executed, the posture of the target object in the first image is an upright posture, and the judged color is the first color information.

[0078] In this embodiment, the upright posture means that the target object (such as the sole) is correctly oriented in the first image, with the sole area at the top and the heel area at the bottom. This posture makes the key features of the target object and the stitching wire easy to observe and evaluate. On the production line, the correct posture ensures that all subsequent automated inspection equipment such as cameras and sensors can correctly identify and analyze the characteristics of the target object, such as wire position, stitching quality and pattern accuracy.

[0079] The judging color is usually selected as the most critical or most prominent color in production, which is used to mark the completed stitch code wire. The first color information is used as the judging standard to identify and evaluate the quality and integrity of the stitch code. In the production of multi-color soles, if black is used as the main color of the structural stitch code, the system sets this color as the judging color. This enables the inspection system to focus on the black area, evaluate its continuity and neatness, and thus determine whether the stitch code work is complete.

[0080] In any of the above embodiments, when the target object needle code wire is executed, the posture of the target object in the second image is an inverted posture, and the judged color is the third color information.

[0081] In this embodiment, in the second image, the target object (such as the sole) is shown in an inverted posture, that is, the heel area is located at the top and the sole area is located at the bottom. This posture is usually used to check the internal structure of the sole and the bottom surface details of the stitching to ensure the integrity of the internal quality and structure. In the production of high-performance sports shoes, the inverted posture enables the detection system to carefully check the inner structure of the sole and confirm that all internal stitching and reinforcement materials have been correctly placed and sewn, which is key to the functionality and durability of the sole; the third color information is used as a judgment color, which is usually a color specially selected according to production requirements to indicate key design or structural features. In the inverted posture, the third color helps to identify manufacturing defects or special materials that may be hidden at the bottom. When producing work shoes with anti-slip soles, the third color (such as bright yellow or green) may be used to identify specific anti-slip materials. By detecting the arrangement and continuity of these colors, it can be confirmed whether the anti-slip layer is intact.

[0082] The inverted posture provides an in-depth view of the bottom of the target object, which is an ideal perspective for evaluating bottom details and internal structural integrity. Combined with the third color information, the bottom features can be judged more accurately, especially when the color is used to mark specific areas or materials; ensuring that the target object is correctly analyzed in the inverted posture, combined with the special evaluation color, the inspection is not limited to the surface, but also includes parts that are usually difficult to observe. This method improves the comprehensiveness and depth of inspection, ensuring that every detail of the product meets the quality standards; through the comprehensive analysis of the inverted posture and the third color information, the production monitoring system can maintain a high standard of quality control throughout the production process. This control mechanism helps to identify and correct potential manufacturing problems in a timely manner, reduce scrap rates, and ensure that each product meets the design specifications.

[0083] In any of the above embodiments, the identification pattern is matched with the posture of the target object to represent the needle code integrity of the target object.

[0084] In this embodiment, the identification pattern is usually a specific color or mark used to highlight key stitch code areas or special materials on the target object. These marks help the automated system identify and evaluate the quality and integrity of the stitch code. When producing complex sports shoes with multi-layer structures, different layers may be represented by different identification colors. For example, the structural support area may be marked in blue, while the cushioning layer may use green to ensure that each layer of material is correctly placed and stitched according to the design specifications; the posture of the target object determines how the detection system captures the image and analyzes the identification pattern. The correct posture can ensure the visibility of the identification pattern and the accuracy of the analysis. When making safety shoes, the sole may need to be placed in an inverted posture to check the integrity of the anti-slip layer. If the anti-slip layer uses a specific identification color, the inverted posture can ensure that these areas are correctly identified during the detection.

[0085] The effectiveness of the identification pattern depends largely on the correct posture display of the target object. Only in the correct posture can the identification color or mark be fully captured and correctly interpreted by the system. This combined use not only improves the accuracy of detection, but also optimizes the efficiency of the automated process; combining the identification pattern and the posture of the target object, the system can comprehensively evaluate the integrity of the needle code. This includes checking whether the identification area is continuous without breaks, whether the color meets the specifications, and whether the pattern is complete. For example, a complete identification line shows a seamless needle code operation, while breaks or color differences may indicate manufacturing defects; by real-time monitoring of the representation of the identification pattern and the posture of the target object, the production system can be adjusted instantly to ensure that each step of the operation meets quality standards. This feedback mechanism reduces scrap rate, improves production efficiency, and ensures that the final product meets the expected quality requirements.

[0086] S304, determining whether the second continuous value is greater than the first detection threshold, if so, the needle code of the target object is incomplete and is set as a needle code error.

[0087] Here, the second continuous value usually refers to the length of continuous color information of the stitch code wire in a certain detection unit or along a certain direction. This value reflects the continuity of the stitch code and the integrity of the pattern. When making sports shoe soles with specific color patterns, the continuous length of the wire color along the stitch code direction is measured to confirm that the pattern is not interrupted; the first detection threshold is a preset standard used to compare the second continuous value to determine the integrity of the stitch code. If the second continuous value is lower than this threshold, it indicates that there may be problems with the stitch code, such as broken wire or misalignment. In the production of high-quality leather products, any incomplete stitch code may affect the appearance and durability. Set a threshold based on product quality standards to ensure that the stitch code of each part is intact.

[0088] As can be seen from the above, comparing the second continuous value and the first detection threshold is a key step in evaluating the integrity of the stitch code. This comparison process directly affects quality control decisions, such as whether the stitch code needs to be re-coded or other corrective measures are required; if the second continuous value is less than the first detection threshold, the system will automatically generate an identification pattern to mark the area where the stitch code is incomplete. These identification patterns can be used for subsequent repair work or quality inspections to ensure that problems are solved in a timely manner. On the automated textile production line, the system detects the color continuity break in a certain area through real-time monitoring, and automatically generates a warning mark in the control system to guide the operator to check and repair; it not only improves product quality assurance, but also optimizes the production process. Reduce rework and waste caused by quality problems, and improve overall production efficiency and economic benefits.

[0089] Furthermore, after the target object is incompletely knitted, it is determined whether the color image at the current moment is marked based on the following model, and the model is constructed using the following steps:

[0090] Step 1: Color feature extraction and quantitative analysis:

[0091] Using principal component analysis (PCA) or independent component analysis (ICA), the color information of the target object surface, the carrier platform surface, and the needle code wire is extracted from the color image to reduce redundancy and extract the most important color feature vector.

[0092] Color histogram matching and mean color difference analysis are introduced to accurately evaluate the color changes of the pin code wire under different lighting conditions and shooting angles, thereby obtaining a highly quantitative color similarity score ( ). This score quantifies the degree of deviation of image color information by evaluating the statistical differences of each color channel (such as RGB, HSV).

[0093] ;

[0094] in, and are the pre-acquired color information and the value of the current color image respectively, n is the number of detection units, and i is the detection unit number.

[0095] Step 2: Image texture and pin code pattern recognition:

[0096] Through wavelet transform or Gabor filter, the texture features of the pin code area in the image are extracted to identify the direction, interval and consistency of the pin code.

[0097] The convolutional neural network (CNN) is used to perform pattern recognition on the shape and continuity of the stitch code, and the outline of the stitch code lines is extracted by combining edge detection algorithms (such as Canny or Sobel) to obtain the stitch code consistency score ( ). The output of this step can accurately identify the details of broken or missed stitches.

[0098] ;

[0099] in, is the local area of ​​the pin code extracted from the image, and the CNN output is the probability value of pin code pattern recognition.

[0100] Step 3: Spatial position and posture evaluation:

[0101] Use deep learning models (such as ResNet) or posture estimation algorithms (such as OpenPose) to determine the posture of the target object (upright or inverted), and perform posture recognition by extracting the key points of the target object in the image.

[0102] At the same time, the position change of the object in the current image is estimated through geometric transformation, such as rotation, translation and scale change, to generate a posture score ( ). Correct posture corresponds to a higher score, while abnormal posture may lead to needle code errors.

[0103] ;

[0104] in, is the posture matching function, and are the predicted and actual pose points respectively, and N is the number of detection units containing pose points.

[0105] Step 4: Continuity detection and time series analysis:

[0106] For the temporal continuity detection of the needle code, the hidden Markov model (HMM) or dynamic time warping (DTW) is used to analyze the time series changes in the image frame to evaluate the continuous integrity of the needle code wire.

[0107] For each image at each time point, the continuity score of the stitch code wire (ScontS_{cont}Scont) is calculated. The model determines whether there is a break or discontinuity by fitting the stitch code shape changes between the current frame and the previous frame.

[0108] ;

[0109] in, and For the pin code area of ​​the current and previous frame images, DTW calculates the continuity difference between the two frames.

[0110] Step 5: Generate logo pattern through deep learning:

[0111] Based on the results of the above scores, a decision tree model or random forest method is used to make multi-dimensional judgments.

[0112] When the model determines that a needle code error occurs, a generative adversarial network (GAN) is used to automatically generate a logo pattern that matches the error area of ​​the needle code and mark it on the color image. GAN generates an adaptive logo based on the size and shape of the error area.

[0113] ;

[0114] in, The final comprehensive score is is the error area in the image.

[0115] Step 6, fuzzy logic and final decision:

[0116] Using fuzzy logic system, each score , , , Input into the fuzzy inference system to generate the final comprehensive score .

[0117] By setting a threshold value T, it is determined whether the comprehensive score exceeds the threshold value. If it exceeds the threshold value, an identification pattern is generated and displayed to indicate a needle code error.

[0118] ;

[0119] in, is a fuzzy inference function, and the output is the final decision score.

[0120] Step 7: Threshold setting and judgment criteria:

[0121] By comparing historical data with the needle code errors in actual production, Bayesian optimization or grid search is used to dynamically adjust the threshold T to adapt to different production environments and needle code specifications.

[0122] Generate logo pattern conditions: ;

[0123] Specifically, the final function is the following formula:

[0124] ;

[0125] Among them, the color continuity score : reflects whether the color information of the needle code wire of the target object is consistent with the expectation. The more continuous the needle code color is, the higher the score is; the continuity of the needle code pattern : reflects whether the stitch pattern continues according to the expected rules and structure. The more regular the stitch pattern, the higher the score; posture matching : reflects whether the posture of the target object is consistent with the standard posture. The higher the posture matching degree, the higher the score; pin code coverage consistency : Reflects the completeness of the stitch code wire covering the entire surface. The higher the consistency, the higher the score.

[0126] In summary, by combining machine learning, deep learning, fuzzy logic and advanced image processing technology, the complex evaluation model proposed in this invention can automatically analyze and generate highly targeted identification patterns. The model introduces precise color and texture analysis, gesture recognition and time series monitoring to ensure that pin code errors are detected and marked in real time, automatically and intelligently during the production process, greatly improving production efficiency and product quality control.

[0127] Specifically, the surface of the target object is divided into n detection units, and the detection units are numbered as i; and the formula of the model is updated and determined on the plurality of the detection units through the following steps:

[0128] The , , and stated as input variables, and respectively obtain the membership function of each of the input variables.

[0129] The fuzzy output of each membership function is obtained through fuzzy rules and membership reasoning, and all the fuzzy outputs are aggregated to generate a fuzzy result.

[0130] The formula of the model is adjusted according to the fuzzy result and the actual needle code error detection result, and the formula of the model is updated to ; Logarithmic function, function.

[0131] As can be seen above, fuzzy rules describe the logical relationship between different input states based on defined membership functions. These rules are constructed based on expert knowledge or empirical data to guide how to infer the final output state from various input states. After the actual input value is converted into a fuzzy state through a membership function, these fuzzy rules are applied for logical reasoning. Each rule provides a fuzzy output based on the fuzzy state of the input, usually through logical operators (such as "and" and "or") to handle the combination of multiple conditions. Since multiple rules may be applicable at the same time, the system needs to aggregate the fuzzy outputs of all applicable rules into a single fuzzy output result. This step is usually achieved by taking the maximum value of each output to ensure that the output result can represent the combined impact of all applicable rules.

[0132] Collect data including color matching, stitch pattern continuity, posture matching, and stitch coverage consistency scores, and record the corresponding stitch errors. These data will provide the necessary empirical basis for the model, and use the collected data to evaluate the accuracy of the existing model. The performance of the model can be evaluated through statistical analysis such as cross-validation and error calculation to ensure that the model can reflect the actual production, and adjust the parameters of the membership function and the fuzzy rules of the model based on the evaluation results. This may include adjusting the shape of the membership function and optimizing the logical structure of the rules to improve the model's ability to capture the relationship between variables in actual production.

[0133] S104, marking the identification pattern in the color image according to the display position of the needle code error in the color image.

[0134] Here, stitch errors are first detected by analyzing the color continuity values ​​in the color image. This includes identifying breaks, discontinuities or abnormal patterns in the color information, which may indicate broken stitch wires, missed stitches or other errors. Once stitch errors are detected, the system will accurately calculate the specific locations of these errors in the color image. This usually relies on advanced image processing techniques such as edge detection, color segmentation and pattern recognition. A corresponding identification pattern is generated based on the error type and location and marked at the error location in the color image. The identification pattern can take different shapes, colors or symbols to distinguish different types of stitch errors.

[0135] As can be seen from the above, error detection is the prerequisite for marking the identification pattern. Only when the error is accurately detected can the system proceed to the next step of marking. Error positioning directly affects the accurate placement of the identification pattern. Accurate positioning of the error ensures that the identification pattern can be directly covered on the error part, thereby ensuring the pertinence and efficiency of the correction measures. The generation and application of the identification pattern is the last step in the error handling process. It visualizes the previous error detection and positioning information and provides it to the operator or the automatic repair system as a direct visual guide.

[0136] S105, transmitting the marked color image to a display terminal to be displayed together with the color image before marking at each moment.

[0137] Here, after the pin code error is detected, the corresponding identification pattern is generated in the color image to mark the specific location of the error. This step is completed in real time, ensuring that the error information is updated in time. The processed image is transmitted to the display end, which may be a large screen in the control room or an operator's monitoring device. This allows the monitoring personnel to see the changes and markings in the image in real time. The display end shows the color image before and after the marking at the same time, which allows the operator to compare and evaluate the changes before and after the error correction, and better understand the nature of the problem and the effect of the treatment.

[0138] As can be seen above, the visualization directly related to the error makes the problem area clearly marked in the image, providing a direct visual reference for subsequent correction operations. The transmission of images ensures a seamless connection from error detection to error display, enhancing the response speed and accuracy of the monitoring system. The joint display helps increase the efficiency of operators and provides an intuitive comparison to more accurately evaluate and verify the effectiveness of corrective measures.

[0139] The real-time error correction method for sole stitching based on CDS color quantization provided by the present invention optimizes and corrects possible errors in the stitching process by acquiring and analyzing color information, and pre-acquires the first color information of the surface of the target object, the second color information of the surface of the platform carrying the target object, and the third color information of the surface of the stitching wire. Then, by capturing a real-time color image of the target object, and using the first and second color information to accurately determine the coverage area of ​​the target object in the color image.

[0140] The first continuous value of the third color information is obtained from the color image in units of pixels, and the size of the detection unit is determined accordingly. Subsequently, the coverage area is segmented based on these detection units, and a first detection threshold is set. On this basis, a judgment color is selected according to the posture of the target object in the color image, and the second continuous value of the judgment color in the coverage area is obtained.

[0141] Whether the stitch code wire has been executed is evaluated by judging whether the second continuous value is greater than the first detection threshold. If the second continuous value exceeds the threshold, it indicates that the target object has an incomplete stitch code problem, and an identification pattern is generated accordingly. Through this precise color quantification and continuous value analysis method, the present invention can effectively identify and correct problems such as missing stitches or deflection in the stitch code process, thereby ensuring production efficiency and product quality, and is particularly suitable for large-scale regular production processing.

[0142] Specifically, Figure 5 As shown in the figure, O100 b60w30 is the color number of the first color information, and O100 b10w80 is the color number of the third color information.

[0143] The embodiment of the second aspect of the present invention provides a system 2 for implementing a real-time error correction method for sole stitch codes based on CDS color quantization. In some embodiments of the present invention, Figure 6 As shown, the system 2 includes:

[0144] The color calibration module 201 is used to calibrate the first color information, the second color information and the third color information.

[0145] The target recognition module 202 is used to recognize the target object in the color image and divide the coverage area of ​​the target object.

[0146] The area segmentation module 203 is used to segment the coverage area into multiple detection units.

[0147] The stitch code recognition module 204 is used to obtain the second continuous value of the first color information or the third color information, and determine whether the stitch code wire of the target object is complete.

[0148] In the system provided by the present invention, the color calibration module is responsible for calibrating the first color information (surface of the target object), the second color information (surface of the carrying platform), and the third color information (surface of the wire). Color calibration is the basis for ensuring the accuracy of subsequent image processing. Through accurate color information, the system can accurately identify and analyze the needle code process. Color calibration provides the target recognition module with necessary information, so that the module can more accurately identify the target object and its background, thereby improving the recognition efficiency and accuracy of the overall system.

[0149] The target recognition module is used to identify the target object in the acquired color image and divide the coverage area of ​​the target object. This step is the prerequisite for accurate pin code detection, ensuring that the subsequent pin code analysis and error correction are performed on the correct area. The target recognition module determines the location and range of the coverage area, providing a clear working basis for the area segmentation module, so that the subsequent segmentation and pin code recognition can be performed more accurately.

[0150] The region segmentation module is responsible for further subdividing the coverage area divided by the target recognition module into multiple detection units. These detection units are the basic units of the pin code integrity analysis. The region segmentation module directly affects the efficiency and accuracy of the pin code recognition module. A reasonable detection unit size can effectively improve the detection accuracy of pin code defects.

[0151] The pin code recognition module is responsible for obtaining the second continuous value of the first color information or the third color information, and judging whether the pin code wire of the target object is complete. The measurement of this continuous value is a key indicator for judging the quality of the pin code. The pin code recognition module is the core of the entire system, and its analysis results directly determine whether it is necessary to generate a logo pattern and take error correction measures. The output of this module depends on the precise input of the first three modules.

[0152] Embodiments of the third aspect of the present invention provide electronic devices. In some embodiments of the present invention, such as Figure 7 As shown, an electronic device is provided, which includes: electronic devices such as desktop computers, notebooks, handheld computers, and cloud servers. The electronic device 3 may include but is not limited to a processor 301 and a memory 302. Those skilled in the art will understand that Figure 7 The electronic device 3 is merely an example and does not limit the electronic device 3 , and may include more or less components than those shown in the figure, or different components.

[0153] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0154] The memory 302 may be an internal storage unit of the electronic device 3, for example, a hard disk or memory of the electronic device 3. The memory 302 may also be an external storage device of the electronic device 3, for example, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 3. The memory 302 may also include both an internal storage unit of the electronic device 3 and an external storage device. The memory 302 is used to store computer programs and other programs and data required by the electronic device.

[0155] The embodiment of the fourth aspect of the present invention proposes a computer-readable storage medium. In some embodiments of the present invention, a computer-readable storage medium is provided, and when the computer-readable storage medium is executed by the processor 301, the steps of the above method are implemented. Therefore, the computer-readable storage medium provided by the fourth aspect of the present invention has all the technical effects of the above steps, which will not be repeated here.

[0156] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0157] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0158] In the embodiments provided in the present disclosure, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0159] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0160] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be included in the protection scope of the present disclosure.

Claims

1. A real-time error correction method for sole stitch codes based on CDS color quantization, characterized in that: The steps include: Pre-acquire first color information of the surface of the target object, second color information of the surface of the platform carrying the target object, and third color information of the surface of the wire material of the target object; Continuously capture the color image of the target object at each moment and transmit it to the display end for real-time display; According to the distribution of the first color information, the second color information and the third color information in the color image, determining whether there is a needle code error in the color image at the current moment, and generating a marking pattern if there is a needle code error; Marking the identification pattern in the color image according to the display position of the pin code error in the color image; The color image after marking is transmitted to the display end to be displayed together with the color image before marking at each moment.

2. The real-time error correction method for sole stitch code based on CDS color quantization according to claim 1, characterized in that: The determining whether there is a pin code error in the color image at the current moment includes: It is determined whether the third color information in the color image completely covers the first color information, and incomplete coverage is regarded as a pin code error.

3. The real-time error correction method for sole stitch code based on CDS color quantization according to claim 2, characterized in that: Also includes: Constructing a model for determining whether to generate the identification pattern; And the model includes: ； Among them, the is the calculation score output by the model, used to determine whether to generate the identification pattern; is a fuzzy inference function, used to affect the calculation score of the model output; Characterizing the consistency between the color information of the wire surface in the current color image and the pre-acquired third color information; Characterizing the correlation between the pin code pattern in the current color image and the expected regularity and structural continuation; Characterizing the consistency between the posture of the target object in the current color image and the standard posture; Characterizes the completeness of the stitch code wire covering the entire target object.

4. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 3 is characterized in that: By setting the A threshold T is matched to determine whether the color image at each moment generates a logo pattern; When the When the value of is greater than the threshold value T, a logo pattern is generated in the color image at the current moment; When the When the value of is less than or equal to the threshold T, no identification pattern is generated.

5. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 4 is characterized in that: By comparing the historical color image data and the needle code error status in production, the threshold T is dynamically adjusted using Bayesian optimization or grid search.

6. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 3, characterized in that: The surface of the target object is divided into n detection units, and the detection units are numbered as i; and the formula of the model is updated and determined on the multiple detection units through the following steps: The , , and stated as input variables, and respectively obtaining the membership function of each of the input variables; Obtaining the fuzzy output of each membership function through fuzzy rules and membership reasoning, and aggregating all the fuzzy outputs to generate a fuzzy result; The formula of the model is adjusted according to the fuzzy result and the actual needle code error detection result, and the formula of the model is updated to ; Logarithmic function, function.

7. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 6, characterized in that: Said The membership function of is: ； Among them, the is the third color information acquired in advance; is the color information of the wire surface in the current color image.

8. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 6, characterized in that: Said The membership function of is: ； Among them, the is the local area of ​​the needle code extracted from the color image; the CNN output is the probability value of needle code pattern recognition.

9. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 6, characterized in that: Said The membership function of is: ； Among them, the is the posture matching function; is the posture point predicted by the detection unit; is the actual posture point of the detection unit; N is the number of detection units containing posture points.

10. The real-time error correction method for sole stitch codes based on CDS color quantization according to claim 6, characterized in that: Said The membership function of is: ； Among them, the The pin code area of ​​the current frame color image; The stitch area of ​​the color image; DTW is to calculate the continuity difference between two frames.

Images