Machine vision-based fabric seam alignment control system

Abstract

This invention relates to the field of alignment control technology, specifically a machine vision-based fabric seam alignment control system, including a grayscale peak recognition module, a seam edge recognition module, an image segment alignment judgment module, a tension direction correction module, and an alignment trajectory adjustment module. In this invention, by constructing grayscale peak trajectories and analyzing the stability and slope changes between peaks, the rhythmic region of the pattern is accurately extracted, enhancing the image processing's ability to recognize edge distribution. Structural recognition is performed using the directional features and coordinate positions of grayscale abrupt change points, avoiding reliance on static feature templates and improving the adaptability of edge recognition. Alignment judgment between image segments is performed based on the directional consistency and coordinate trend relationship of abrupt change points, enhancing the dynamic recognition of inter-segment continuity. Changes in the angle between the tension direction and the target line drive the selection of backup points and the updating of target points, improving the directional adaptability of the target point under tension disturbances.

Description

Technical Field

[0001] This invention relates to the field of alignment control technology, and in particular to a fabric turning seam alignment control system based on machine vision. Background Technology

[0002] Alignment control technology is an important branch of automated control systems, mainly used for precise position coordination and alignment between multiple objects. This technology is widely used in printing, circuit board manufacturing, textile processing, packaging, and mechanical assembly, playing a core role, especially in processes requiring high-precision positioning and matching. With the development of industrial automation and intelligent manufacturing, alignment control systems are gradually integrating advanced sensing technologies (such as lasers and vision sensors) and image processing algorithms to achieve rapid identification, error detection, and dynamic compensation control of target objects, thereby improving product quality and production efficiency.

[0003] Among them, the machine vision-based fabric turning and seam alignment control system is a control system that uses visual recognition to automatically detect and adjust the position of the seam during the fabric turning process. Its core is to capture images of the fabric seam through machine vision equipment, analyze them in combination with image processing algorithms, and achieve precise alignment by controlling the actuator. The main application of this system is in textile production lines, especially in the fabric turning and splicing process, to improve the quality of fabric splicing, reduce manual intervention, improve the level of production automation and overall processing efficiency.

[0004] Existing technologies mostly rely on static template matching or edge threshold extraction. Image features are difficult to adapt to areas with irregular textures or slight grayscale changes, leading to errors in feature boundary recognition. In image segment recognition, static coordinate difference judgment is commonly used, lacking a directional trend judgment mechanism and failing to respond to structural offset issues. In tension disturbance scenarios, existing path control mostly relies on target position offset adjustment, ignoring the trend of tension direction changes, which easily causes target command offset. Path advancement adjustment is based on angle offset, without introducing multi-factor boundary conditions, resulting in delayed execution path correction. Under the combined effect of image structure offset and tension changes, existing technologies have problems such as response lag, structural misalignment, and path instability in fabric seam alignment scenarios. Summary of the Invention

[0005] To address the shortcomings of existing technologies, such as reliance on static template matching or edge threshold extraction, which makes it difficult for image features to adapt to areas with irregular textures or subtle grayscale variations, leading to errors in feature boundary recognition, static coordinate difference judgment is commonly used in image segment recognition, lacking a directional trend judgment mechanism and failing to respond to structural offset issues. In tension disturbance scenarios, existing path control relies heavily on target position offset adjustment, ignoring the trend of tension direction changes, which easily causes target command offset. Path advancement adjustment is based on angle offset, without introducing multi-factor boundary conditions, resulting in delayed execution path correction. Under the combined effect of image structural offset and tension changes, existing technologies suffer from response lag, structural misalignment, and path instability in fabric seam alignment scenarios. This invention provides a machine vision-based fabric seam alignment control system. The technical solution is as follows:

[0006] On the one hand, a machine vision-based fabric turning and sewing head alignment control system is provided, including:

[0007] The grayscale peak recognition module acquires image frames of the fabric turning process, identifies the continuous position of grayscale peaks in the image, compares the stability of the spacing between adjacent peaks with the change of grayscale gradient slope, determines whether the symmetry is stable, and obtains the effective segment identifier of the pattern.

[0008] The seam edge recognition module constructs the trajectory of gray-level difference between adjacent pixels based on the effective segment identifier of the pattern, locates the key position points of two abrupt changes, analyzes the gray-level trend and relative structural features between the two abrupt changes, and obtains the set of structural abrupt change feature points.

[0009] The image segment alignment judgment module, based on the set of structural mutation feature points, calls the vertical coordinates of the mutation points in two images, compares whether their positional differences fall within the inter-segment offset range, detects whether the mutation direction remains consistent in the continuous image cycle, determines the image segment with stable structural direction and controllable positional changes, and obtains matching image segment data.

[0010] The tension direction correction module collects the tension direction vector based on the region coordinates in the matching image segment data, constructs the angle sequence formed by the vector and the line connecting the seam target point to the center of the fabric, analyzes whether the angle reverses in the image sequence, and obtains the target point repositioning result.

[0011] On the other hand, the effective segment identifier of the pattern includes the peak position index, the boundary spacing interval, and the gray-scale symmetrical segment; the structural mutation feature point set includes the mutation point coordinate group, the mutation point direction information, and the gray-scale difference change label; the matchable image segment data includes the inter-segment alignment coordinate group, the structural direction correspondence, and the image periodic distribution label; and the target point relocation result includes the tension principal vector indication, the target point selection number, and the angle change sequence.

[0012] On the other hand, the grayscale peak recognition module includes:

[0013] The grayscale sequence extraction submodule acquires image frames of the fabric turning process, extracts the grayscale distribution of each column in the Y-axis direction of the image, calculates the maximum grayscale amplitude for each column of pixel grayscale set, locates the grayscale amplitude change arrangement trend in consecutive columns, and generates a grayscale amplitude trend set between columns.

[0014] The grayscale trajectory construction submodule collects the changing direction between the differences between each group of columns based on the grayscale amplitude trend set between columns, arranges them in order according to the column order, groups the continuous segments of the changing direction, and integrates all direction segment sequences to construct the column order direction transfer trajectory group.

[0015] The structural symmetry judgment submodule extracts the interval span between continuous wave peaks in the column sequence based on the column sequence direction transfer trajectory group, collects the directional change length of the rising and falling segments on both sides of the wave peak, calculates the column offset in the symmetrical direction, and combines the relationship between the directional arrangement and the column sequence span in the structural symmetry segment to determine whether the symmetry condition is met and obtains the valid segment identifier of the pattern.

[0016] On the other hand, the seam edge recognition module includes:

[0017] The column sequence construction submodule extracts the gray value sequence of each column in the corresponding image area according to the effective segment identifier of the pattern, obtains the gray value arrangement set of the corresponding row number in each column according to the column order, and pairs the gray values ​​under the same row number in consecutive columns by index to form a column sequence array to obtain the column pixel gray value arrangement set.

[0018] The gradient trajectory construction submodule calls the column-oriented pixel grayscale arrangement set, establishes a difference set for grayscale between adjacent pixels, obtains the changing direction of grayscale difference of continuous pixels, records the start and end row numbers of the changing segment according to the direction of grayscale difference, connects them in order to form a directional arrangement chain, and obtains pixel grayscale difference trend data.

[0019] Based on the pixel grayscale difference trend data, the structural feature filtering submodule identifies abrupt boundary where grayscale direction changes, records its vertical coordinates and corresponding grayscale transformation direction, extracts grayscale arrangement segments between two adjacent abrupt boundary, determines whether the arrangement direction is continuous and consistent, and filters stable structures whose arrangement has not reversed, thereby obtaining a set of structural abrupt feature points.

[0020] On the other hand, the image segment alignment determination module includes:

[0021] The coordinate difference detection submodule extracts the vertical coordinates of the corresponding column order in the two images based on the structural abrupt feature point set, calculates the vertical coordinate spacing of the corresponding points under the premise that the column order numbers are consistent, determines whether the coordinate spacing falls within the inter-segment offset allowable range, and obtains the column position offset range.

[0022] The direction change judgment submodule calls the column position offset interval, extracts the gray-scale gradient direction marker of the corresponding change point in each column in the continuous image period, records the gradient direction change results of the same column order in each period, identifies the position number of the point where the direction changes, and counts the column number segments with continuous and consistent directions to obtain a list of direction stable segments.

[0023] The structural segment matching and filtering submodule matches the column sequence combination with the smallest column offset and unchanged direction according to the list of directional stable segments. It determines whether the abrupt change point in the combination satisfies the dual conditions of consistent direction and continuous offset within the period, determines the controllable image segment, and obtains the matching image segment data.

[0024] On the other hand, the determination of whether the mutation point in the combination satisfies the dual conditions of consistent direction and continuous offset within the period is based on the formula:

[0025] ;

[0026] Calculate continuous offset feature values ​​to determine controllable image segments, and obtain matching image segment data. Representing the The offset continuity feature value used to determine the offset continuity in the combination of column segments Representing the In the sequence combination of the nth column segment, the first The column offset of each abrupt change point in a periodic segment is represented by the horizontal pixel value of the image. This indicates the offset of the preceding segment. Representing the In the sequence combination of the nth column segment, the first The image orientation angle of each periodic abrupt change point is obtained from the gradient direction of the neighborhood of the abrupt change point in the image. The direction and angle of its preceding image segment. To prevent positive constants with a denominator of zero, This indicates the number of calculable effective periodic segments in the sequence combination.

[0027] On the other hand, the tension direction correction module includes:

[0028] The direction vector extraction submodule collects the start and end positions of the tension direction of the coordinate points based on the region coordinates in the matching image segment data, calculates the straight line direction between the start and end coordinates, and calibrates the projection direction of the vector in the image plane to obtain the tension direction projection trajectory.

[0029] The angle sequence construction submodule calls the tension direction projection trajectory to obtain the angle between each tension vector and the line connecting the corresponding seam target point to the center of the fabric, calculates the direction of change of the angle, sorts it according to the image frame number, determines whether the angle direction changes in consecutive frames, marks the position of the direction change, and generates the angle turning change distribution.

[0030] The point relocation submodule filters vector segments whose directions have not reversed based on the distribution of angle changes, matches coordinate points in the backup points that are adjacent in the included angle direction and have the same extension direction, calls the extension direction coordinates of the backup points for cross comparison, locates the points with overlapping directions in the current cycle, and obtains the target point relocation result.

[0031] On the other hand, the coordinates of the matching backup points that are adjacent in the included angle direction and have the same extension direction are selected using the following formula:

[0032] ;

[0033] Calculate the difference values ​​of adjacent angles, call the extended direction coordinates of the backup points for cross-comparison, and locate the points where directions coincide in the current cycle. Represents the current candidate point The difference value between the angle direction and the backup point is adjacent to the value. Indicate candidate points The direction of the included angle, This represents the average value of the included angles of all the backup points. Indicate candidate points The extended distance, This represents the set of extended distances for each of the backup points. Indicate candidate points The rate of change in its extension direction, This represents the average rate of change of all backup points in their extension direction.

[0034] On the other hand, the system also includes:

[0035] Based on the target point repositioning result, the alignment trajectory adjustment module calls the execution path direction and movement record of the previous cycle, calculates the centerline of the angle between the main tension direction and the original path direction, determines whether the centerline is located within the movement boundary interval, limits the path advancement direction, sets the adjustment trajectory along the centerline direction, and obtains the corrected travel trajectory configuration.

[0036] The modified travel trajectory configuration includes adjusting the path vector, the direction of the centerline angle, and the path boundary parameter group.

[0037] On the other hand, the alignment trajectory adjustment module includes:

[0038] The path parameter extraction submodule extracts the image coordinates of the corresponding point in the current frame based on the target point relocation result, calls the execution path direction data and movement distance under the same column number in the previous period, calculates the angle difference between the tension direction and the original path direction, and obtains tension and path angle comparison data.

[0039] The centerline direction determination submodule calls the tension and path angle comparison data, calculates the symmetrical angle centerline of each pair of tension directions and original path directions according to the period number, detects whether the starting point and ending point of the centerline in the motion coordinate area are within the boundary interval, and marks the centerline segment number that meets the motion constraints to obtain the restricted interval advancement centerline set.

[0040] The trajectory configuration generation submodule advances the centerline set according to the restricted interval, filters the coordinate segments in continuous frames that have consistent centerline direction and whose boundaries are not touched, extracts the advancement path between the start and end points of the segment, and connects it with the target point coordinates to form a path chain, thereby obtaining the corrected travel trajectory configuration.

[0041] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0042] By constructing grayscale peak trajectories and analyzing the stability of spacing and slope changes between peaks, the rhythmic regions of the pattern are accurately extracted, enhancing the image processing's ability to identify edge distribution. Structural recognition is performed using the directional features and coordinate positions of grayscale abrupt change points, avoiding reliance on static feature templates and improving the adaptability of edge recognition. Alignment judgment between image segments is made based on the directional consistency and coordinate trend relationship of abrupt change points, enhancing the dynamic recognition of inter-segment continuity. Changes in the angle between the tension direction and the target line drive the selection of backup points and the updating of target points, improving the directional adaptability of target points under tension disturbances. The path correction process introduces the judgment of the angle centerline and combines it with the advancement boundary constraint, improving the stability of the path response. The processing logic constructs a linkage mechanism between image features, tension changes, and path correction, enhancing the structural alignment stability during dynamic fabric changes. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the system of the present invention;

[0045] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0046] Figure 3 This is a flowchart of the grayscale peak recognition module of the present invention;

[0047] Figure 4 This is a flowchart of the seam edge recognition module of the present invention;

[0048] Figure 5 This is a flowchart of the image segment alignment determination module of the present invention;

[0049] Figure 6 This is a flowchart of the tension direction correction module of the present invention;

[0050] Figure 7 This is a flowchart of the alignment trajectory adjustment module of the present invention. Detailed Implementation

[0051] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0052] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0053] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0054] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0055] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0056] This invention provides a machine vision-based fabric seam alignment control system, such as... Figure 1 As shown, the system includes:

[0057] The grayscale peak recognition module acquires image frames of the fabric turning process, extracts the grayscale distribution of each column in the Y-axis direction of the image, constructs the grayscale projection change trajectory between columns, identifies the continuous position of grayscale peaks in the image, compares the stability of the spacing between adjacent peaks with the change of grayscale gradient slope, determines whether the symmetry in the continuously distributed peak structure is stable, and obtains the effective segment identifier of the pattern.

[0058] The seam edge recognition module extracts the grayscale sequence of the corresponding area based on the effective segment identifier of the pattern, constructs the change trajectory of grayscale difference between adjacent pixels, locates the key position points of two abrupt changes, records their coordinates and gradient direction in the vertical direction, analyzes the grayscale trend and relative structural features between the two abrupt change points, selects stable gradient combinations, and obtains the set of structural abrupt change feature points.

[0059] The image segment alignment judgment module is based on the structural mutation feature point set. It calls the vertical coordinates of the mutation points in two images, compares whether their positional differences fall within the inter-segment offset range, detects whether the mutation direction remains consistent in the continuous image cycle, determines the image segment with stable structural direction and controllable positional change, and obtains matching image segment data.

[0060] The tension direction correction module collects the tension direction vector based on the region coordinates in the matching image segment data, constructs the angle sequence formed by the vector and the line connecting the seam target point to the center of the fabric, analyzes whether the angle reverses in the image sequence, and calls the points in the spare points whose angle direction is close to and whose extension direction coincides to obtain the target point repositioning result.

[0061] The alignment trajectory adjustment module, based on the target point repositioning result, calls the execution path direction and movement record of the previous cycle, calculates the centerline of the angle between the main tension direction and the original path direction, determines whether the centerline is within the motion boundary interval, limits the path advancement direction, sets the adjustment trajectory along the centerline direction, and obtains the corrected travel trajectory configuration.

[0062] The effective segment identifiers for the pattern include peak position index, boundary spacing interval, and gray-scale symmetrical segments. The set of structural abrupt change feature points includes abrupt change point coordinate group, abrupt change point direction information, and gray-scale difference change labels. The matching image segment data includes inter-segment alignment coordinate group, structural direction correspondence, and image periodic distribution labels. The target point relocation results include tension principal vector indication, target point selection number, and angle change sequence. The corrected travel trajectory configuration includes adjusting the path vector, centerline angle direction, and path boundary parameter group.

[0063] A mutation point refers to a location in the gray-level gradient curve where the gray-level change suddenly increases or decreases, corresponding to the light-dark transition area at the edge of the fabric pattern or seam texture. A critical location point refers to the coordinate position of the mutation point in the vertical direction (Y-axis) of the image. This point represents the key area where the image structure produces texture changes in that direction, and is typically used to determine edge direction, structural continuity, and seam alignment reference. An image segment refers to a local area in the image divided according to the pattern rhythm, usually determined by the start and end column range marked by the "effective segment number of the pattern rhythm." It is the basis for subsequent structural consistency judgment and seam alignment judgment. Each image segment contains a relatively complete seam pattern structure and its gradient changes, and is the basic unit for the machine vision system to judge the continuity and alignment of seams between segments. A backup point refers to a set of candidate target point coordinates preset near the fabric edge or seam, qualified to be replaced. The system will take action if the target point experiences a drastic reversal in the angular direction or tension disturbance, causing the existing target point to become invalid. At that time, the point with the smallest included angle and aligned direction is selected from the set as the new target to ensure that the tension direction and the alignment point adjustment are coordinated; the execution path refers to the actual movement trajectory planned by the actuator to align the seam head during the movement of the fabric. It is represented by the change sequence of the propulsion direction vector in a continuous cycle. The execution path is composed of the travel direction of the previous cycle, the target point position, and the tension vector. It is the core control object for dynamically adjusting the alignment action. The path adjustment is based on the centerline of the included angle direction; the column order refers to the column position order of the image from left to right, numbered sequentially by pixel unit. The column represents the arrangement of pixels in each vertical direction in the image. For example, the image column order is the column array numbering in the entire image frame according to the column number. It usually starts from the first column on the left with the number 1 (or 0), and is numbered 2, 3, 4... until the rightmost column of the image. In an image with a resolution of 640×480, the column order number is 1~640, and each column order position corresponds to a whole column of 480 vertical pixels.

[0064] like Figure 2 and Figure 3 As shown, the grayscale peak recognition module includes:

[0065] The grayscale sequence extraction submodule acquires image frames of the fabric turning process, extracts the grayscale distribution of each column in the Y-axis direction of the image, calculates the maximum grayscale amplitude for each column of pixel grayscale set, locates the grayscale amplitude change arrangement trend in consecutive columns, and generates a grayscale amplitude trend set between columns.

[0066] The grayscale data extraction operation is performed sequentially on each column of pixels along the Y-axis in the image frame. The set of grayscale values ​​of all pixels in each column is represented as a grayscale column vector. For each column of grayscale vector, the maximum and minimum grayscale values ​​of all pixels in that column are extracted first, and the maximum grayscale amplitude value of that column is calculated by the difference between the two. For example, for a column containing pixels with grayscale values ​​of [45, 60, 55, 80, 90, 100], the maximum value is 100, the minimum value is 45, and the maximum grayscale amplitude is 100-45=55. The grayscale amplitudes of all columns are stored in column order to form an initial sequence. Next, the grayscale values ​​of adjacent columns are compared sequentially in all columns. To determine the trend of grayscale variation, the difference between the grayscale values ​​of column i and column (i+1) is calculated, and the positive or negative direction of the difference is recorded. If the difference is positive, it indicates that the grayscale amplitude is increasing; otherwise, it indicates that it is decreasing. For example, the grayscale amplitudes of column 1 and column 2 are 48 and 55 respectively, with a difference of +7, indicating an increasing trend. The grayscale amplitudes of column 2 and column 3 are 55 and 52 respectively, with a difference of -3, indicating a decreasing trend. This process is repeated until all columns are traversed. Through this difference calculation, the grayscale amplitude variation trend sequence between columns is obtained. This sequence is used to represent the spatial evolution process of grayscale fluctuations in local structures in the fabric image. Finally, this trend sequence is stored as a set of grayscale amplitude trends between columns according to the original column order of the image.

[0067] The grayscale trajectory construction submodule collects the changing direction between the differences between each column group based on the grayscale amplitude trend set between columns, arranges them in order according to the column order, groups the continuous segments of the changing direction, and integrates all direction segment sequences to construct the column order direction transfer trajectory group.

[0068] The direction of change indicated by the grayscale amplitude difference between adjacent columns is read sequentially. All directions of change are concatenated according to the image column order to form a one-dimensional original sequence of change directions. Segments with consistent continuous change directions are grouped, and each group is called a direction segment group. For example, if the original sequence is [↑, ↑, ↑, ↓, ↓, ↑, ↑], it can be divided into three segments: [↑, ↑, ↑], [↓, ↓], and [↑, ↑]. Each segment records its starting column, ending column, and direction of change. By scanning the entire trend sequence and accumulating the segment group information in sequence, a direction segment sequence group is formed. Based on this, according to the arrangement order of the direction segment groups in the image column order, they are integrated to construct a column order direction transfer trajectory group. This trajectory group retains multi-dimensional information such as the consistency of direction, segment span, and change order of each grayscale fluctuation segment, which facilitates subsequent analysis of structural symmetry. For example, if a certain area in the image presents a grayscale change direction sequence of [↑10 columns, ↓8 columns, ↑10 columns], then the structure has strong symmetry and can be intuitively identified through this trajectory group.

[0069] The structural symmetry judgment submodule extracts the interval span between continuous wave peaks in the column sequence based on the column sequence direction transfer trajectory group, collects the directional change length of the rising and falling segments on both sides of the wave peak, calculates the column offset in the symmetrical direction, and combines the relationship between the directional arrangement and the column sequence span in the structural symmetry segment to determine whether the symmetry condition is met and obtains the valid segment identifier of the pattern.

[0070] Analyze the positions of wave peaks in adjacent directional segments, extract the interval span between consecutive wave peaks in the column sequence, record the index position of each wave peak in the column sequence, and calculate the span value by the difference in the number of columns between two wave peaks. If the wave peaks appear in the 20th and 36th columns respectively, the interval span is 16 columns. Then, collect the gray-scale change direction on both sides of the wave peak and the length of each directional segment, and count the number of columns of the continuous rising or falling segments on the left and right sides of the wave peak. For example, if the left side is [↑5 columns, ↑3 columns] and the right side is [↓3 columns, ↓5 columns], then the corresponding rising and falling segment lengths are 8 columns respectively. Based on the data, calculate the symmetrical column offset value of the directional segments on both sides, that is, the distance difference from the starting point of the symmetrical boundary to the center of the wave peak. For example, if the distance from the starting point of the left side to the wave peak is... If there are 8 columns on the right, the offset is 0. The criterion is whether the offset is less than the set threshold Δ. When Δ is set to no more than 3 columns, if the offset is ≤3, it is considered to meet the symmetry condition. The setting of Δ is based on the image resolution and the stable range of the fabric width. If the horizontal resolution of the image is 640 columns and the maximum swing range of the fabric width is ±10mm, the corresponding image offset is ±6 columns. Setting Δ=3 can ensure that the offset is within the controllable range. Finally, combined with the symmetrical arrangement relationship between the directional segments and whether the column span is equal, if the difference in length between the symmetrical directional segments on both sides is no more than 1 column, and the direction is consistent and the arrangement order is symmetrical, then the column segment is determined to meet the symmetry condition, and its column range is marked as the valid pattern segment.

[0071] like Figure 2 and Figure 4 As shown, the seam edge recognition module includes:

[0072] The column sequence construction submodule extracts the gray value sequence of each column in the corresponding image area according to the valid segment identifier of the pattern, obtains the gray value arrangement set of the corresponding row number in each column according to the column order, and pairs the gray values ​​under the same row number in consecutive columns by index to form a column sequence array to obtain the column pixel gray value arrangement set.

[0073] First, determine the left and right boundaries and top and bottom boundaries of the identified region in the image frame. Locate each column of pixels within the region and read the grayscale value of each pixel in each column from top to bottom, forming a vertical grayscale sequence for each column. Establish a column-oriented grayscale set based on image columns. In practical applications, if an image is 640 columns wide (px) and 480 rows high (px), and the effective segment is from column 200 to column 220, then extract 480 grayscale values ​​from each column within this 21-column range as a column-oriented array. Each column array is arranged in top-to-bottom order by row number. Subsequently, in adjacent columns... The algorithm searches for grayscale value pairs corresponding to the same row number, pairs them one by one, and forms a pixel grayscale pairing structure. That is, under each row number, it records the grayscale combination of the same height in two columns. For example, in row number 100, the grayscale value of column 200 is 128 and column 201 is 132, so they are paired as a set of grayscale points. The pairing is continued to complete for all row numbers, and finally a column-oriented grayscale array is formed. The width of the array is the number of effective columns, and the length is the number of rows in the image. The overall structure is a column-oriented grayscale sequence grid, which is suitable for identifying the horizontal change trend of grayscale between columns at the same height and completing the construction of the column-oriented pixel grayscale arrangement set.

[0074] The gradient trajectory construction submodule calls the column-oriented pixel grayscale arrangement set, establishes a difference set for the grayscale between adjacent pixels, obtains the changing direction of the grayscale difference of continuous pixels, records the start and end row numbers of the changing segment according to the direction of grayscale difference, connects them in order to form a directional arrangement chain, and obtains pixel grayscale difference trend data.

[0075] For each column, comparing the grayscale values ​​of two adjacent pixels from top to bottom, the results are marked as increasing, decreasing, or unchanged to indicate the directional trend of grayscale differences. In image processing, if in column 150, the pixel in row 50 has a grayscale value of 110 and the pixel in row 51 has a grayscale value of 114, the latter is greater than the former, therefore the grayscale trend is increasing. Conversely, if the next pixel is 112, the trend is decreasing. If the grayscale values ​​are the same, it is marked as stable. After this operation is performed for each column, the direction markers are concatenated according to row number order to identify consecutive grayscale direction segments with the same trend. Record the starting and ending row numbers of each directional segment, and determine whether the grayscale change of each segment has directional continuity. If the grayscale in a continuous directional segment always increases or decreases, it is recorded as a complete directional segment. For example, if the grayscale in the 205th column gradually increases from the 20th to the 30th row and the trend of change is consistent, then this segment is a directional segment. Repeat this process until all columns are traversed, and connect all directional segments in the order of the image columns to form a directional arrangement chain. This chain comprehensively records the continuous directional structure of grayscale change with height, constituting pixel grayscale difference trend data.

[0076] The structural feature filtering submodule identifies abrupt boundary changes in grayscale direction based on pixel grayscale difference trend data, records its vertical coordinates and corresponding grayscale transformation direction, extracts grayscale arrangement segments between two adjacent abrupt boundary changes, determines whether the arrangement direction is continuous and consistent, and filters stable structures whose arrangement has not reversed, thereby obtaining a set of structural abrupt feature points.

[0077] The algorithm iterates through each column of the directional arrangement chain, locating the boundary point where the grayscale change direction first reverses. For example, the turning point where multiple upward segments are followed by a downward direction is considered a mutation boundary. The vertical row number and the direction of grayscale change of the boundary point are recorded. For example, if the direction changes from upward to downward from row 100, then 100 is recorded as the mutation row number and the direction changes from upward to downward. The operation is repeated to record all mutation boundary points. Then, the grayscale arrangement segments corresponding to two adjacent mutation boundary points are extracted, and it is checked whether their overall grayscale change direction is consistent. If there is no reversal, the arrangement direction is considered to be continuous and stable. For example, if the grayscale continuously decreases from row number 110 to row number 122 without any reversal trend, it is determined to be a stable structural segment. The length is checked to see if it exceeds the stability threshold. If the threshold is set to 5 rows, then the length exceeding this value is valid. Finally, all grayscale arrangement segments with no reversal and the length meeting the set requirements are selected. The structural feature points representing the segment are extracted from the middle position, and their column number, row number, and change direction are recorded. These are then summarized to form a set of structural mutation feature points.

[0078] like Figure 2 and Figure 5 As shown, the image segment alignment determination module includes:

[0079] The coordinate difference detection submodule extracts the vertical coordinates of corresponding column order in two images based on the structural abrupt feature point set, calculates the vertical coordinate spacing of corresponding points under the premise that the column order number is consistent, determines whether the coordinate spacing falls within the inter-segment offset allowable range, and obtains the column position offset range.

[0080] Extract the row number information of the mutation point that has the same column number as the previous period image frame, ensuring that all point pairs come from the same column number and have the prerequisite for comparison in consecutive image periods. For each column number, extract the vertical coordinate value of the corresponding mutation point from the two period images respectively, and calculate the vertical row number difference of the mutation point corresponding to the column number in the two periods. For example, in column 240, the mutation point in the current image frame is located in row 105 and in row 102 in the previous period image frame. The vertical coordinate distance between the two points is 3 rows. This difference is called the coordinate distance. After calculating the coordinate distance of all corresponding points column by column, compare whether the difference falls within the allowable range of inter-segment offset. The range setting needs to be set according to the image resolution and the fabric tension fluctuation characteristics. For example, when the image resolution is 480 rows and the vertical fluctuation within the fabric period is controlled within 5 rows, the allowable offset range can be set to no more than 5 rows. That is, only the column number with a coordinate distance of no more than 5 rows is retained, and point pairs with offset exceeding the range are excluded. The column number and the corresponding offset range with the offset value in the vertical direction satisfying the set range are obtained.

[0081] The direction change judgment submodule calls the column position offset interval, extracts the gray-scale gradient direction mark of the corresponding change point in each column in the continuous image period, records the gradient direction change results of the same column order in each period, identifies the position number of the point where the direction changes, and counts the column number segments with continuous and consistent directions to obtain a list of direction stable segments.

[0082] In each periodic image frame, the grayscale gradient direction labels of the abrupt change points in that column are read column by column. The direction labels of the same column number in consecutive periodic images are compared to see if they have changed. For example, if the direction of column 260 in the current period is upward and the direction of the previous period was also upward, then the direction is consistent. If the direction of the previous period was downward, then the direction has changed. The column number of each point with a change in direction is recorded column by column. The column numbers with unchanged direction are accumulated to form a continuous region segment. The starting column number, ending column number, length, and direction consistency mark of each segment are counted. Each region segment with consistent direction is then numbered and stored to form a list of direction-stable segments. When setting the threshold, the span of the structurally stable interval must be considered. If the length of continuous and consistent direction is required to be no less than 5 columns to form a valid segment, then regions with a length of less than 5 columns must be excluded. For example, if the direction is consistent in columns 200 to 206, a stable segment with a length of 7 columns is formed, while columns 212 to 214, which only have 3 columns, are not adopted. All column segments that meet the continuous length requirement and have consistent direction are combined to form a list of direction-stable segments.

[0083] The structural segment matching and filtering submodule matches the column sequence combination with the smallest column offset and unchanged direction based on the list of directional stable segments. It then determines whether the abrupt change point in the combination satisfies the dual conditions of consistent direction and continuous offset within the period, identifies controllable image segments, and obtains matching image segment data.

[0084] To determine whether abrupt changes in a combination satisfy both the conditions of consistent direction and continuous offset within the period, the following formula is used:

[0085] ;

[0086] Calculate continuous offset feature values ​​to determine controllable image segments, and obtain matching image segment data. Representing the The offset continuity feature value used to determine the offset continuity in the combination of column segments Representing the In the sequence combination of the nth column segment, the first The column offset of each abrupt change point in a periodic segment is represented by the horizontal pixel value of the image. This indicates the offset of the preceding segment. Representing the In the sequence combination of the nth column segment, the first The image orientation angle of each periodic abrupt change point is obtained from the gradient direction of the neighborhood of the abrupt change point in the image. The direction and angle of its preceding image segment. To prevent positive constants with a denominator of zero, This indicates the number of calculable effective periodic segments in the sequence segment combination;

[0087] Let the column sequence combination number be The number of periodic segments is This indicates that there are a total of 5 consecutive periodic segments participating in the combination judgment. Next, the column offset and direction angle information of the mutation point of each periodic segment are extracted in sequence. The column offset of the mutation point is then... The data is obtained by using the horizontal pixel difference between the horizontal pixel position of the mutation point in the image and the horizontal pixel difference of the mutation point in the previous period. Assuming the pixel unit is pixel value, five data segments measured from image processing are obtained as follows: 3.2, 4.1, 5.0, 6.0, and 7.2 pixels, corresponding to the image orientation angle data. The angles are 85.0 degrees, 84.5 degrees, 86.2 degrees, 86.0 degrees, and 85.5 degrees, respectively. The direction angles are calculated from the direction vector of the main edge in the image, using the image gradient vector direction as the angle extraction method. The unit is degrees. Continuous offset feature values ​​are calculated, where... We use 0.01 as a positive constant to prevent the denominator from being zero. The summation in the formula is performed four times, starting from the second segment to the fifth segment, and substituted sequentially as follows:

[0088] Sub-item 2: ;

[0089] Section 3 Sub-item: ;

[0090] Section 4 Sub-item: ;

[0091] Section 5 Sub-item: ;

[0092] Add the four sub-items together: ;

[0093] Then take the average:

[0094] ;

[0095] Let the threshold value for determining the continuity of the offset be... This represents the average value of the continuity index of abrupt change point shifts in 100 sets of images acquired under the same scene conditions. with standard deviation The upper limit of the set mean, i.e. Current results The condition for offset continuity is satisfied. On the other hand, the determination of directional consistency uses a threshold. Based on the fact that the differences in the directional angles of the images in each period segment are 0.5 degrees, 1.7 degrees, 0.2 degrees, and 0.5 degrees, except for the third segment which exceeds the threshold, the rest are all within the threshold range. To ensure overall consistency, a maximum allowable number of single exceedances should be set to no more than one. The current situation meets this condition, and this result indicates that the combination... The mutation points of the column sequence segments within 5 cycles meet the requirements in terms of both direction consistency and offset continuity, confirming that the combined segments are controllable and obtaining matching image segment data.

[0096] like Figure 2 and Figure 6 As shown, the tension direction correction module includes:

[0097] The direction vector extraction submodule collects the start and end positions of the tension direction of the coordinate points based on the region coordinates in the matching image segment data, calculates the straight line direction between the start and end coordinates, and calibrates the projection direction of the vector in the image plane to obtain the tension direction projection trajectory.

[0098] First, the key coordinate point pairs located in each image segment are read. These coordinate point pairs are defined as the starting and ending points of the tension application. Then, the column and row information of the starting and ending coordinates are called in sequence. The line segment formed by the two points in the image plane is defined as the tension direction line. By reading the coordinate changes between the two points in the column and row directions, the orientation of the line segment is determined to be classified as upper left to lower right, lower left to upper right, horizontal to the right, vertical downward, etc. The image projection direction of each line segment is marked according to this classification. For example, if the tension starting point is in column 240, row 100, and the ending point is in column 250, row 130, and the column direction difference is positive and the row direction difference is positive, then the tension line segment is determined to be from upper left to lower right. This projection direction is stored as a label for the tension vector. All tension line segments in the image segment are arranged in column order to form the tension direction projection trajectory.

[0099] The angle sequence construction submodule calls the tension direction projection trajectory to obtain the angle between each tension vector and the line connecting the corresponding seam target point to the center of the fabric, calculates the direction of change of the angle, sorts it according to the image frame number, determines whether the angle direction changes in consecutive frames, marks the position of the direction change, and generates the angle turning change distribution.

[0100] The projection direction of each tension vector and the coordinates of the seam target point in its corresponding image segment are read sequentially. Then, the coordinates of the fabric center point in the same frame are read. A line connecting the seam target point to the fabric center point is constructed. By reading the changes in column and row numbers of the start and end points of the two line segments, spatial direction labels are constructed. The directions of the two line segments are then classified as left-up, right-down, and right-direction, respectively. The corresponding angle relationships between the two direction labels are matched and recorded as angle values. The image frames are then sorted from smallest to largest according to their numbers. The angle value in each frame is compared with the angle value in the previous frame to determine whether the angle has changed from rising to falling or vice versa. If the direction changes, the current frame is marked as a point of abrupt change in angle direction. For example, if the angle is 42 degrees in frame 5, 45 degrees in frame 6, and 41 degrees in frame 7, then frame 7 is determined to be a point of abrupt change, and its image number is recorded. According to the image frame number order, all positions where the angle direction changes are formed into an angle turning change distribution.

[0101] The point relocation submodule filters vector segments whose direction has not reversed based on the distribution of angle turning changes, matches coordinate points in the backup points that are adjacent in the included angle direction and have the same extension direction, calls the extension direction coordinates of the backup points for cross comparison, locates the points with overlapping directions in the current cycle, and obtains the target point relocation result.

[0102] To match the coordinates of adjacent points in the included angle direction and with the same extension direction among the backup points, the following formula is used:

[0103] ;

[0104] Calculate the difference values ​​of adjacent angles, call the extended direction coordinates of the backup points for cross-comparison, and locate the points where directions coincide in the current cycle. Represents the current candidate point The difference in angular direction between the current candidate point and the backup point refers to a comprehensive quantitative index that measures the degree of deviation between the current candidate point and the backup point in three dimensions: angular direction, extension distance, and rate of change. It evaluates the consistency of the angular direction, the similarity of geometric extension characteristics, and the dynamic coordination of displacement rates between the candidate point and the backup point. Indicate candidate points The direction of the included angle, This represents the average value of the included angles of all the backup points. Indicate candidate points The extended distance (i.e., the Euclidean distance between the point and the previous coordinate point). This represents the set of extended distances for each of the backup points. Indicate candidate points The rate of change in its extension direction (i.e., the change in extension distance per unit time or unit step). This represents the average rate of change of all backup points in their extension direction;

[0105] Determine candidate points The direction of the included angle If the value is obtained from the actual angle... At the same time, three backup points are set, with their included angles being respectively... , , Then calculate its average value:

[0106] ;

[0107] The square of the difference in the included angle is:

[0108] ;

[0109] Obtain candidate points extension distance Assuming The distance of the backup point extension is , , The squares of their differences are as follows:

[0110] ;

[0111] ;

[0112] ;

[0113] The mean square error of the extended distance is:

[0114] ;

[0115] The square root is:

[0116] ;

[0117] Extracting the elongation rate of candidate points The backup point extension rate is , , The average value is:

[0118] ;

[0119] The square of the rate offset is:

[0120] ;

[0121] The maximum speed is:

[0122] ;

[0123] The normalization term is:

[0124] ;

[0125] Calculation results:

[0126] ;

[0127] This result indicates the current candidate point The overall directional offset, path distance error, and speed fluctuation with the selected backup point are controlled within a reasonable range. If the matching threshold is 0.6, the value is within... If it is below the matching threshold The determination of direction overlap can be used for target point relocation.

[0128] like Figure 2 and Figure 7 As shown, the alignment trajectory adjustment module includes:

[0129] The path parameter extraction submodule extracts the image coordinates of the corresponding point in the current frame based on the target point relocation result, calls the execution path direction data and movement distance under the same column number in the previous cycle, calculates the angle difference between the tension direction and the original path direction, and obtains tension and path angle comparison data.

[0130] The system obtains the new image coordinates of the target point in the current image frame and reads its corresponding column and row numbers. These coordinates are used as the positioning reference point for the tension direction. Then, it calls the path execution data in the previous image frame that is consistent with the current column number, extracts the path direction unit vector information and the translation distance data of the corresponding point in the previous period. Then, it uses the change in column and row numbers between the start and end points of the current tension direction as the direction reference, reads the components of the two sets of direction vectors, namely the tension vector and the path vector of the previous period, and performs the angle calculation operation between the vector directions to obtain the angle difference between the tension direction and the original path direction at each point. For example, if the current tension direction is from the upper left to the lower right and the original path direction is to the right, with an angle of 45 degrees, then it is recorded as the tension and path angle of this column is 45 degrees. After performing the same angle difference extraction in all columns, a tension and path angle comparison dataset is formed. This dataset contains column numbers, tension direction markers, path direction markers and the numerical information of the angles between them.

[0131] The centerline direction determination submodule calls the tension and path angle comparison data, calculates the symmetrical angle centerline of each pair of tension directions and original path directions according to the period number, detects whether the starting point and ending point of the centerline in the motion coordinate area are within the boundary interval, and marks the centerline segment number that meets the motion constraints to obtain the restricted interval propulsion centerline set.

[0132] For each pair of directions, the centerline direction is calculated, which determines the symmetrical angle division direction corresponding to the angle between the tension direction and the original path direction. This direction is the centerline direction between the two vectors. Then, the coordinates of the start and end points in the image coordinate plane are derived based on the centerline direction. It is determined whether the horizontal and vertical components are within the preset boundary constraints within the image frame. If the image width is 640 columns and the height is 480 rows, and the boundary constraint interval is set to within 20 columns or 20 rows from the edge, only the centerline segments whose start and end points are both between column number 21 to 620 and row number 21 to 460 are retained. If a centerline start point is in column 18 or the end point is in row 470, it is removed. Only the centerline segments whose two ends are both within the safe zone are marked as compliant path direction segments, and their corresponding frame number or period number is recorded. This yields a set of centerline segment numbers that are within the motion restriction range, have complete projection, and whose direction satisfies the symmetrical angle formation condition, forming the restricted interval advancement centerline set.

[0133] The trajectory configuration generation submodule advances the centerline set according to the restricted interval, filters the coordinate segments in continuous frames that have consistent centerline direction and whose boundaries are not touched, extracts the advancement path between the start and end points of the segment, and connects it with the target point coordinates to form a path chain, thereby obtaining the corrected travel trajectory configuration.

[0134] Extracting the centerline segments from all centerline numbering segments to determine if there are centerline segments with consistent direction labels between consecutive image frames, checking frame by frame whether the direction labels are the same in consecutive numbering segments. If the direction is consistent for four consecutive frames in a segment and does not change, then the frame range is recorded as a consistent segment. Further reading the centerline start and end coordinates of each frame in the consistent segment, connecting the start and end coordinates to form a propulsion path segment, and checking whether the boundary of the path segment within the image frame has been reached. If the start or end point is not in the image edge area, then the path segment is retained as a stable propulsion segment. Then reading the target point relocation coordinates in the current image frame, connecting the end coordinates of the propulsion path segment with the target point to establish a complete path structure link, connecting point by point to form a path chain extending from the previous cycle to the current cycle, extracting all coordinate segments with connectivity, direction consistency, and boundary validity as the output of the corrected path structure, and constructing the corrected travel trajectory configuration.

[0135] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0136] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0137] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0138] Those skilled in the art will recognize that the units and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0139] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0140] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0141] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0142] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0143] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0144] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A machine vision-based fabric seam alignment control system, characterized in that, The system includes: The grayscale peak recognition module acquires image frames of the fabric turning process, identifies the continuous position of grayscale peaks in the image, compares the stability of the spacing between adjacent peaks with the change of grayscale gradient slope, determines whether the symmetry is stable, and obtains the effective segment identifier of the pattern. The grayscale peak recognition module includes: The grayscale sequence extraction submodule acquires image frames of the fabric turning process, extracts the grayscale distribution of each column in the Y-axis direction of the image, calculates the maximum grayscale amplitude for each column of pixel grayscale set, locates the grayscale amplitude change arrangement trend in consecutive columns, and generates a grayscale amplitude trend set between columns. The grayscale trajectory construction submodule collects the changing direction between the differences between each group of columns based on the grayscale amplitude trend set between columns, arranges them in order according to the column order, groups the continuous segments of the changing direction, and integrates all direction segment sequences to construct the column order direction transfer trajectory group. The structural symmetry judgment submodule extracts the interval span between continuous wave peaks in the column sequence based on the column sequence direction transfer trajectory group, collects the directional change length of the rising and falling segments on both sides of the wave peak, calculates the column offset in the symmetrical direction, and combines the relationship between the directional arrangement and the column sequence span in the structural symmetry segment to determine whether the symmetry condition is met and obtains the valid segment identifier of the pattern. The seam edge recognition module constructs the trajectory of gray-level difference between adjacent pixels based on the effective segment identifier of the pattern, locates the key position points of two abrupt changes, analyzes the gray-level trend and relative structural features between the two abrupt changes, and obtains the set of structural abrupt change feature points. The image segment alignment judgment module, based on the set of structural mutation feature points, calls the vertical coordinates of the mutation points in two images, compares whether their positional differences fall within the inter-segment offset range, detects whether the mutation direction remains consistent in the continuous image cycle, determines the image segment with stable structural direction and controllable positional changes, and obtains matching image segment data. The tension direction correction module collects the tension direction vector based on the region coordinates in the matching image segment data, constructs the angle sequence formed by the vector and the line connecting the seam target point to the center of the fabric, analyzes whether the angle reverses in the image sequence, and obtains the target point repositioning result.

2. The machine vision-based fabric seam alignment control system according to claim 1, characterized in that, The effective segment identifier of the pattern includes the peak position index, the boundary spacing interval, and the gray-scale symmetrical segment. The structural mutation feature point set includes the mutation point coordinate group, the mutation point direction information, and the gray-scale difference change label. The matchable image segment data includes the inter-segment alignment coordinate group, the structural direction correspondence, and the image periodic distribution label. The target point relocation result includes the tension principal vector indicator, the target point selection number, and the angle change sequence.

3. The machine vision-based fabric seam alignment control system according to claim 1, characterized in that, The seam edge recognition module includes: The column sequence construction submodule extracts the gray value sequence of each column in the corresponding image area according to the effective segment identifier of the pattern, obtains the gray value arrangement set of the corresponding row number in each column according to the column order, and pairs the gray values ​​under the same row number in consecutive columns by index to form a column sequence array to obtain the column pixel gray value arrangement set. The gradient trajectory construction submodule calls the column-oriented pixel grayscale arrangement set, establishes a difference set for grayscale between adjacent pixels, obtains the changing direction of grayscale difference of continuous pixels, records the start and end row numbers of the changing segment according to the direction of grayscale difference, connects them in order to form a directional arrangement chain, and obtains pixel grayscale difference trend data. Based on the pixel grayscale difference trend data, the structural feature filtering submodule identifies abrupt boundary where grayscale direction changes, records its vertical coordinates and corresponding grayscale transformation direction, extracts grayscale arrangement segments between two adjacent abrupt boundary, determines whether the arrangement direction is continuous and consistent, and filters stable structures whose arrangement has not reversed, thereby obtaining a set of structural abrupt feature points.

4. The machine vision-based fabric seam alignment control system according to claim 1, characterized in that, The image segment alignment determination module includes: The coordinate difference detection submodule extracts the vertical coordinates of the corresponding column order in the two images based on the structural abrupt feature point set, calculates the vertical coordinate spacing of the corresponding points under the premise that the column order numbers are consistent, determines whether the coordinate spacing falls within the inter-segment offset allowable range, and obtains the column position offset range. The direction change judgment submodule calls the column position offset interval, extracts the gray-scale gradient direction marker of the corresponding change point in each column in the continuous image period, records the gradient direction change results of the same column order in each period, identifies the position number of the point where the direction changes, and counts the column number segments with continuous and consistent directions to obtain a list of direction stable segments. The structural segment matching and filtering submodule matches the column sequence combination with the smallest column offset and unchanged direction according to the list of directional stable segments. It determines whether the abrupt change point in the combination satisfies the dual conditions of consistent direction and continuous offset within the period, determines the controllable image segment, and obtains the matching image segment data.

5. The machine vision-based fabric seam alignment control system according to claim 4, characterized in that, The determination of whether the mutation point in the combination satisfies the dual conditions of consistent direction and continuous offset within the period is based on the following formula: ； Calculate continuous offset feature values ​​to determine controllable image segments, and obtain matching image segment data. Representing the The offset continuity feature value used to determine the offset continuity in the combination of column segments Representing the In the sequence combination of the nth column segment, the first The column offset of each abrupt change point in a periodic segment is represented by the horizontal pixel value of the image. This indicates the offset of the preceding segment. Representing the In the sequence combination of the nth column segment, the first The image orientation angle of each periodic abrupt change point is obtained from the gradient direction of the neighborhood of the abrupt change point in the image. The direction and angle of its preceding image segment. To prevent positive constants with a denominator of zero, This indicates the number of calculable effective periodic segments in the sequence combination.

6. The machine vision-based fabric seam alignment control system according to claim 1, characterized in that, The tension direction correction module includes: The direction vector extraction submodule collects the start and end positions of the tension direction of the coordinate points based on the region coordinates in the matching image segment data, calculates the straight line direction between the start and end coordinates, and calibrates the projection direction of the vector in the image plane to obtain the tension direction projection trajectory. The angle sequence construction submodule calls the tension direction projection trajectory to obtain the angle between each tension vector and the line connecting the corresponding seam target point to the center of the fabric, calculates the direction of change of the angle, sorts it according to the image frame number, determines whether the angle direction changes in consecutive frames, marks the position of the direction change, and generates the angle turning change distribution. The point relocation submodule filters vector segments whose directions have not reversed based on the distribution of angle changes, matches coordinate points in the backup points that are adjacent in the included angle direction and have the same extension direction, calls the extension direction coordinates of the backup points for cross comparison, locates the points with overlapping directions in the current cycle, and obtains the target point relocation result.

7. The machine vision-based fabric seam alignment control system according to claim 6, characterized in that, The coordinates of the matching backup points that are adjacent in the included angle direction and have the same extension direction are selected using the formula: ； Calculate the difference values ​​of adjacent angles, call the extended direction coordinates of the backup points for cross-comparison, and locate the points where directions coincide in the current cycle. Represents the current candidate point The difference value between the angle direction and the backup point is adjacent to the value. Indicate candidate points The direction of the included angle, This represents the average value of the included angles of all the backup points. Indicate candidate points The extended distance, This represents the set of extended distances for each of the backup points. Indicate candidate points The rate of change in its extension direction, This represents the average rate of change of all backup points in their extension direction.

8. The machine vision-based fabric seam alignment control system according to claim 1, characterized in that, The system also includes: Based on the target point repositioning result, the alignment trajectory adjustment module calls the execution path direction and movement record of the previous cycle, calculates the centerline of the angle between the main tension direction and the original path direction, determines whether the centerline is located within the movement boundary interval, limits the path advancement direction, sets the adjustment trajectory along the centerline direction, and obtains the corrected travel trajectory configuration. The modified travel trajectory configuration includes adjusting the path vector, the direction of the centerline angle, and the path boundary parameter group.

9. The machine vision-based fabric seam alignment control system according to claim 8, characterized in that, The alignment trajectory adjustment module includes: The path parameter extraction submodule extracts the image coordinates of the corresponding point in the current frame based on the target point relocation result, calls the execution path direction data and movement distance under the same column number in the previous period, calculates the angle difference between the tension direction and the original path direction, and obtains tension and path angle comparison data. The centerline direction determination submodule calls the tension and path angle comparison data, calculates the symmetrical angle centerline of each pair of tension directions and original path directions according to the period number, detects whether the starting point and ending point of the centerline in the motion coordinate area are within the boundary interval, and marks the centerline segment number that meets the motion constraints to obtain the restricted interval advancement centerline set. The trajectory configuration generation submodule advances the centerline set according to the restricted interval, filters the coordinate segments in continuous frames that have consistent centerline direction and whose boundaries are not touched, extracts the advancement path between the start and end points of the segment, and connects it with the target point coordinates to form a path chain, thereby obtaining the corrected travel trajectory configuration.

Images