CN113724253A - Spinning carding process quality monitoring method based on image processing - Google Patents
Spinning carding process quality monitoring method based on image processing Download PDFInfo
- Publication number
- CN113724253A CN113724253A CN202111189614.5A CN202111189614A CN113724253A CN 113724253 A CN113724253 A CN 113724253A CN 202111189614 A CN202111189614 A CN 202111189614A CN 113724253 A CN113724253 A CN 113724253A
- Authority
- CN
- China
- Prior art keywords
- pixel
- degree
- image
- distribution
- acquiring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000009960 carding Methods 0.000 title claims abstract description 69
- 238000000034 method Methods 0.000 title claims abstract description 62
- 230000008569 process Effects 0.000 title claims abstract description 31
- 238000012545 processing Methods 0.000 title claims abstract description 22
- 238000009987 spinning Methods 0.000 title claims abstract description 14
- 238000012544 monitoring process Methods 0.000 title claims abstract description 12
- 239000000835 fiber Substances 0.000 claims abstract description 153
- 238000009826 distribution Methods 0.000 claims abstract description 79
- 239000013598 vector Substances 0.000 claims abstract description 75
- 238000004804 winding Methods 0.000 claims abstract description 43
- 230000002159 abnormal effect Effects 0.000 claims abstract description 36
- 238000005452 bending Methods 0.000 claims abstract description 26
- 238000009827 uniform distribution Methods 0.000 claims abstract description 19
- 238000010586 diagram Methods 0.000 claims abstract description 6
- 238000001914 filtration Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 3
- 238000005299 abrasion Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Treatment Of Fiber Materials (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a spinning carding process quality monitoring method based on image processing. The method comprises the following steps: acquiring a gray image of the surface of the fiber to obtain a plurality of edge segments; obtaining the bending degree and the distribution characteristic vector of the edge segment; traversing the gray level image by using a sliding window, calculating the thickness degree of the edge segment where each pixel point is located, and obtaining the long fiber distribution uniformity of the center point of each sliding window according to the thickness degree and the bending degree to form a uniform distribution map; obtaining the winding degree of short fibers at the central point of each sliding window according to the number of long fibers and short fibers in the sliding window and the angle between the long fibers and the short fibers to form a winding distribution diagram; acquiring a uniform distribution map sequence and a winding distribution map sequence in a preset time period, and acquiring abnormal carding degree of each pixel point to form an abnormal distribution map; and determining the abnormal area by using the abnormal distribution map. The embodiment of the invention can accurately monitor the abnormity in the carding process.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a spinning carding process quality monitoring method based on image processing.
Background
In the spinning production flow, the fiber raw material needs to be carded to open the fiber bundle, so that the fiber bundle is separated into parallel and straight single fibers, and fine and short fibers and impurities with strong adhesion are removed to mix the fibers finely and uniformly.
Carding equipment is needed in the carding process, the carding equipment is various, most of the carding equipment utilizes needle teeth to pierce into fiber raw materials, and utilizes relative motion of the fibers and the needle teeth to divide large bundles of fibers into small bundles, so that the fiber bundles are opened, and carding is finished.
During carding, if the parameters of carding equipment are not proper or the equipment is worn, the carding effect is not ideal, the distribution is not uniform, and even the mixed winding among fibers is more disordered. At present, quality monitoring of the carding process is mostly realized through manual observation, and the condition after winding can only be identified through image identification and detection, and unreasonable carding cannot be timely and effectively found in the prior art.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a spinning carding process quality monitoring method based on image processing, which adopts the following technical scheme:
one embodiment of the invention provides a spinning carding process quality monitoring method based on image processing, which comprises the following steps:
acquiring a plurality of gray level images of the fiber surface in the fiber carding process, calculating a pixel characteristic vector of each pixel point in the gray level images, and clustering the pixel characteristic vectors to obtain a plurality of edge segments; the pixel feature vector represents the position and the distribution direction of the corresponding pixel point;
acquiring the principal component direction of each edge segment, and calculating the bending degree of the edge segment and a distribution characteristic vector representing the orientation of the edge segment according to the principal component direction;
traversing pixel points in the gray level image by using a sliding window, calculating the thickness degree of an edge segment where each pixel point is located, and obtaining the long fiber distribution uniformity of the center point of each sliding window according to the thickness degree and the bending degree to form a uniform distribution map;
obtaining the winding degree of the short fibers at the central point of each sliding window according to the number of the long fibers and the short fibers in the sliding window and the angle between the long fibers and the short fibers to form a winding distribution diagram;
acquiring a uniform distribution map sequence and a winding distribution map sequence in a preset time period, and acquiring the carding abnormal degree of each pixel point according to the correlation between the uniform distribution map sequence and the winding distribution map sequence and the distribution change of the two sequences after carding in the time period to form an abnormal distribution map; and determining the abnormal area by using the abnormal distribution map.
Preferably, the step of calculating the pixel feature vector includes:
acquiring a hessian matrix in the neighborhood of each pixel point in the gray level image, and taking a feature vector corresponding to the minimum feature value of the hessian matrix as a structural feature vector of the pixel point;
and combining the structural feature vector with the coordinates of the corresponding pixel point to obtain the pixel feature vector of the pixel point.
Preferably, the method for obtaining the edge segment comprises:
and extracting the edge of the gray image to obtain an edge image, acquiring the pixel characteristic vectors of pixel points belonging to the edge in the edge image, clustering the pixel characteristic vectors, and taking each category as the edge segment.
Preferably, the method for acquiring the distribution feature vector includes:
and acquiring two principal component directions of all pixel points in the edge segment and second eigenvalues corresponding to the two principal component directions, and taking the unit vector of the principal component direction corresponding to the largest second eigenvalue as the distribution eigenvector.
Preferably, the method for obtaining the bending degree comprises:
and obtaining a difference vector of coordinates of each pixel point in the edge segment and the mean value coordinate, calculating an inner product of the difference vector and a unit vector in a principal component direction corresponding to the minimum second characteristic value, and taking the variance of the inner product corresponding to all the pixel points as the bending degree.
Preferably, the step of obtaining the thickness degree includes:
dividing the edge segments into long fibers and short fibers by setting a length threshold;
obtaining a normal vector of the structural feature vector of the pixel points belonging to the long fiber in the sliding window, and forming a pixel sequence by all the pixel points of the straight line where the normal vector is located in the gray level image;
and performing phase reversal processing on the pixel sequence, and then obtaining a plurality of pixel sets by watershed segmentation, wherein the number of the pixel points of the pixel set to which each pixel point belongs is taken as the thickness degree.
Preferably, the step of obtaining the degree of uniformity of the distribution of the long fibers comprises:
calculating the variance of the thickness degree of the pixel points belonging to the long fiber in each sliding window;
obtaining the bending degree of each long fiber in the sliding window, and averaging to obtain an average bending degree;
and calculating the uniformity degree of the long fiber distribution according to the variance and the average bending degree.
Preferably, the step of determining the abnormal region includes:
carrying out mean value filtering on the abnormal distribution map, and carrying out thresholding processing through a preset threshold value to obtain a position image;
and acquiring a region image of the position image by utilizing an on operation, and taking a connected domain in the region image as the abnormal region.
The embodiment of the invention at least has the following beneficial effects:
through obtaining even distribution map and winding distribution map, show the evenly distributed condition of long fiber in carding process and the winding condition of short-staple, obtain the abnormal degree of carding through the fibre evenly distributed condition of different times and the winding condition of short-staple for accurately assess the abnormity in carding process, and then can make the early warning and in time regulate and control production facility before the winding takes place, improved the efficiency that the fibre was carded.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of the steps of a spinning carding process quality monitoring method based on image processing according to an embodiment of the invention.
Detailed Description
In order to further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the method for monitoring the quality of spinning and carding process based on image processing according to the present invention, its specific implementation, structure, features and effects, with reference to the accompanying drawings and preferred embodiments, is provided below. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The specific scheme of the spinning carding process quality monitoring method based on image processing is described in detail below with reference to the accompanying drawings.
Referring to fig. 1, there is shown a flow chart of the steps of a method for monitoring the quality of a spinning carding process based on image processing according to an embodiment of the present invention, the method includes the following steps:
s001, acquiring a plurality of gray level images on the surface of the fiber in the fiber carding process, calculating a pixel characteristic vector of each pixel point in the gray level images, and clustering the pixel characteristic vectors to obtain a plurality of edge segments; the pixel feature vector represents the position and distribution direction of the corresponding pixel point.
The method comprises the following specific steps:
1. and acquiring a gray level image.
The camera overlooks the fibers in the yarn fiber carding working area, and acquires one piece of gray image data every second. In addition, the yarn fibers in the working area are irradiated by a parallel white light source.
The method is characterized in that gray level images of fibers in the carding process are collected, the fibers are parallel and straight and have uniform length and thickness distribution after being carded for many times under an ideal condition, but because actual production equipment and production technology are difficult to achieve the ideal standard, the fibers are partially incompletely straightened, short fibers can exist between long and straight fibers and can be removed by carding equipment, but because the carding equipment has defects or abrasion and control parameters of the equipment are unreasonable, the short fibers cannot be removed, the adhesion of the short fibers can cause the fibers not to be further carded, straightened and uniformly mixed, the noise carding quality is unqualified, the carding time is too long, and the efficiency and the product quality of a subsequent spinning process can be influenced.
2. And acquiring a pixel characteristic vector of each pixel point in the gray level image.
Specifically, a hessian matrix in the neighborhood of each pixel point in the gray level image is obtained, and a feature vector corresponding to the minimum feature value of the hessian matrix is used as a structural feature vector of the pixel point; and combining the structural feature vector with the coordinate of the corresponding pixel point to obtain the pixel feature vector of the pixel point.
For any gray-scale image collected by the camera, the image contains a large number of fibers, the fibers can be distributed in an interweaving way, the lengths and the thicknesses are different, and the lengths and the thicknesses can be distributed unevenly. Because the gray level image has thin and many fiber structures, is easily influenced by noise and may have blurs, the embodiment of the invention enhances the image by using an unsharp masking method based on local gradient and complexity, and highlights detail information of the fiber structures, so that the fibers are distinguished more obviously, and the subsequent analysis of the distribution condition of the fibers is facilitated.
Acquiring each pixel on the enhanced image and a Hessian matrix in the neighborhood, then calculating two characteristic values and two corresponding characteristic vectors of the Hessian matrix, reserving the characteristic vector with the minimum characteristic value as a structural characteristic vector, representing the tangential direction of a pixel point on the edge, namely the direction with the minimum gradient change, as the distribution direction of the structural information of the fiber on the image at the position, wherein the characteristic vector is a two-dimensional unit vector, and each pixel point corresponds to one structural characteristic vector.
And combining the structural feature vector of each pixel and the coordinates of the pixel into a pixel feature vector for representing the position of the pixel and the distribution direction of the structural information at the position of the pixel.
3. And extracting the edge of the gray image to obtain an edge image, acquiring pixel feature vectors of pixel points belonging to the edge in the edge image, clustering the pixel feature vectors, and taking each category as an edge segment.
And extracting the edge of the enhanced image by using a Canny operator to obtain an edge binary image, wherein the pixel with the gray value of 1 on the image is a pixel point on the fiber.
Obtaining all pixels with the gray value of 1 on the edge image, carrying out DBSCAN clustering by using the pixel feature vectors of the pixels, wherein the clustering radius is 3 pixels, obtaining a plurality of categories, each category is a set of some pixels, the pixels in the same category are continuously distributed, and the distribution directions of the adjacent pixels are not different greatly, so that each category represents an edge segment, and the edge segment represents the distribution of fibers.
Step S002 acquires the principal component direction of each edge segment, and calculates the degree of curvature of the edge segment and the distribution feature vector representing the orientation of the edge segment from the principal component direction.
The method comprises the following specific steps:
1. and acquiring two principal component directions of all pixel points in the edge segment and second eigenvalues corresponding to the two principal component directions, and taking the unit vector of the principal component direction corresponding to the largest second eigenvalue as a distribution eigenvector.
And acquiring two principal component directions of all pixel coordinates in each category by using a PCA algorithm, wherein the PCA algorithm can know that each principal component direction is a unit vector and corresponds to a characteristic value.
The principal component direction with the largest eigenvalue is used as a distribution eigenvector for each class representing the orientation of the edge.
2. And obtaining a difference vector of the coordinate of each pixel point in the edge segment and the coordinate of the mean value, calculating an inner product of the difference vector and a unit vector in the principal component direction corresponding to the minimum second characteristic value, and taking the variance of the inner products corresponding to all the pixel points as the bending degree.
And acquiring a principal component direction corresponding to the minimum characteristic value in the category, calculating a difference value between each pixel coordinate in the category and the mean value of all pixel coordinates, then calculating an inner product of a difference vector formed by each difference value and the principal component direction corresponding to the minimum characteristic value, wherein the variance of the inner product of all pixels represents the bending degree of the edge in the category.
When the edge segments represented by a certain class are distributed straight, the bending characteristic is 0; the more curved the edge segment, the greater the degree of curvature.
And S003, traversing pixel points in the gray level image by using the sliding window, calculating the thickness degree of an edge segment where each pixel point is located, and obtaining the long fiber distribution uniformity of the center point of each sliding window according to the thickness degree and the bending degree to form a uniform distribution map.
The method comprises the following specific steps:
1. the edge segments are divided into long fibers and short fibers by setting a length threshold.
Setting a length threshold according to the type of fibers in the actual production process, wherein edge segments larger than the length threshold are called long fibers; edge segments that are less than the length threshold are referred to as short fibers.
2. And obtaining a normal vector of the structural feature vector of the pixel points belonging to the long fiber in the sliding window, and forming a pixel sequence by all the pixel points of the straight line where the normal vector is located in the gray level image.
As an example, the embodiment of the present invention constructs a 13 × 13 sliding window with a sliding window step size of 1.
And acquiring all pixel points belonging to long fibers in a window during sliding each time, acquiring a normal vector of a structural feature vector of the pixel point for the p-th pixel point, and acquiring all pixels on a straight line where the normal vector is located on a gray image to form a pixel sequence.
Each position in the pixel sequence corresponds to a pixel point and also corresponds to a gray value, so that the pixel sequence is also a distribution sequence L1 of gray values.
3. And performing phase reversal processing on the pixel sequence, and then obtaining a plurality of pixel sets by watershed segmentation, wherein the number of the pixel points of the pixel set to which each pixel point belongs is taken as the thickness degree.
Subtracting all gray values in the sequence L1 from 1.0 to obtain a new sequence L2, and finishing the inversion processing of L1; and performing Gaussian blur processing on the sequence L2, and then segmenting the blurred sequence into a plurality of segments by using a watershed segmentation algorithm, wherein each segment is a set of some pixel points, and the number of the pixel points in the set to which the p-th pixel point belongs is obtained and used for representing the thickness degree of the long fiber at the p-th pixel point.
4. Calculating the variance of the thickness degree of the pixel points belonging to the long fiber in each sliding window; obtaining the bending degree of each long fiber in the sliding window, and averaging to obtain the average bending degree; and calculating the uniformity degree of the long fiber distribution according to the variance and the average bending degree.
And calculating the variance x of the thickness degree of all pixel points on the long fiber in the sliding window. Meanwhile, obtaining the mean value y of the bending degrees of all the long fibers in the sliding window, and calculating the distribution uniformity degree of the long fibers in the local area of the central pixel point of the sliding window:
the uniformity includes thickness uniformity and curvature. The larger x and y, the less homogeneous the distribution of the long fibers, i.e. the more inhomogeneous the distribution. Where exp represents an exponential function with a natural constant e as the base.
5. And traversing all the pixel points by sliding the gray image window to obtain the long fiber distribution average degree of each pixel point, and generating a uniform distribution graph of the gray image by taking the long fiber distribution average degree as the pixel value of the pixel point.
And step S004, obtaining the winding degree of the short fibers at the central point of each sliding window according to the number of the long fibers, the number of the short fibers and the angle between the long fibers and the short fibers in the sliding window to form a winding distribution diagram.
The method comprises the following specific steps:
1. and calculating the winding degree of the short fibers at the central point of each sliding window.
And respectively obtaining the quantity of all long fibers and short fibers in the sliding window and the angle between each long fiber and each short fiber, and calculating the winding degree of the short fibers.
As an example, assuming that N1 short fibers appear or partially appear in the sliding window and N2 long fibers appear or partially appear in the sliding window, the winding degree of the short fibers of the pixel point at the center of the sliding window is:
wherein,
showing the cosine similarity of the distribution characteristic vectors of the mth short fiber and the nth long fiber,
representing the absolute value of this cosine similarity.
Absolute value of cosine similarity
The larger the size, the m-th short fiber and the n-th long fiber are specifiedThe fibers are distributed in parallel and straight, and the short fibers do not influence the carding process of the long fibers;
the smaller the size, the non-parallel distribution directions of the mth short fiber and the nth long fiber are indicated, and at this time, the mth long fiber and other long fibers nearby may be wound on the nth long fiber during carding, so that on one hand, the long fibers are no longer straight due to the short fibers, and on the other hand, the short fibers may be wound on part of the long fibers during carding, so that the long fibers cannot be carded smoothly or carded uniformly faster. Thus, it is possible to provide
The smaller the degree of entanglement of the short fibers, the greater the carding interference and negative impact of the short fibers on the local long fibers.
Since the long fibers are distributed more, there are always long fibers in any one sliding window, and in order to avoid the case that the denominator is 0, in the embodiment of the present invention, when N1=0, M = 0.
2. And traversing all the pixel points by the gray image sliding window to obtain the short fiber winding degree of each pixel point, and generating a winding distribution map of the gray image by taking the short fiber winding degree as the pixel value of the pixel point.
Step S005, acquiring a uniform distribution map sequence and a winding distribution map sequence in a preset time period, and acquiring the abnormal carding degree of each pixel point according to the correlation between the uniform distribution map sequence and the winding distribution map sequence and the distribution change of the two sequences subjected to carding in the time period to form an abnormal distribution map; and determining the abnormal area by using the abnormal distribution map.
The method comprises the following specific steps:
1. and acquiring an abnormal distribution map.
And acquiring the uniform distribution map sequence and the winding distribution map sequence in a preset time period before the current time.
The sequence of the uniform distribution map is
(ii) a The winding distribution diagram sequence is
。
Wherein, T represents the length of the sequence,
representing the homogeneity profile at the t-th instant,
showing the winding profile at the t-th instant.
Each pixel point at each moment has a pixel value in the uniform distribution map and the winding distribution map, so that each pixel point also has two corresponding pixel value sequences, namely a pair of sequences and a sum
And performing median filtering on all the image data to obtain filtering results of each pixel point in the uniform distribution map sequence and the winding distribution map sequence.
Taking the p-th pixel point as an example, the filtering result in the histogram sequence is:
(ii) a The result of the filtering in the sequence of the winding profiles is
。
As an example, the filter window size in the embodiment of the present invention is 5 × 5.
Calculating the carding abnormal degree of the p-th pixel point:
wherein,
representing a sequence
And
the correlation coefficient of the pearson (r) is,
i.e. when
When the time is long, the correlation coefficient is not changed,
(ii) a When in use
When the temperature of the water is higher than the set temperature,
and setting 0.
It should be noted that, in the following description,
the larger the value of the long fiber is, the smaller the difference between the uniformity degrees of the long fibers at the 1 st moment and the T th moment is, which indicates that the uniformity degree of the long fibers at the p-th pixel point position is not greatly improved in the carding process in the time period;
the smaller the value of (A), the larger the difference between the uniformity degrees of the long fibers at the 1 st moment and the T th moment, which shows that the carding process in the time period obviously improves the uniformity degree of the long fibers at the p-th pixel point position.
In the same way, the method for preparing the composite material,
the larger the size of the short fibers, the larger the winding degree of the short fibers at the p-th pixel point position in the time period is, and the larger the winding degree of the short fibers at the p-th pixel point position at the T-th time is, so that the carding at the p-th pixel point position at the T-th time is more abnormal.
As an example, in the embodiment of the present invention, T =120 seconds, and the camera sampling frequency is 1 frame/second.
Acquiring the abnormal carding degree of each pixel point, generating an abnormal carding distribution map by using the abnormal carding degree of each pixel point as the pixel value of the pixel point,
2. carrying out mean value filtering on the abnormal distribution map, and carrying out thresholding treatment through a preset threshold value to obtain a position image; and acquiring a region image of the position image by using an on operation, and taking a connected domain in the region image as a region with abnormality.
Carrying out mean value filtering on the abnormal distribution map, carrying out threshold value processing, obtaining a position image, carrying out open operation on the position image, removing an isolated connected domain, obtaining a region image, obtaining a connected domain on the image if the connected domain exists in the region image, wherein the connected domain represents a region in which carding abnormality is concentrated, the winding degree of short fibers is serious and the long fibers are unevenly distributed in the region during carding, and the short fibers cannot be removed during carding, so that the long fibers are influenced to be carded and uniformly mixed, defects or damage and abrasion of carding equipment can be caused, or equipment parameters have problems.
As an example, the window size of the mean filtering in the embodiment of the present invention is 5 × 5.
The process of thresholding is as follows: and setting the gray value of the pixel point with the gray value larger than the preset threshold value as 1, otherwise, setting the gray value as 0. As an example, the preset threshold value is 0.4 in the embodiment of the present invention.
As an example, the open operation template size in the embodiment of the present invention is 7 × 7.
In order to ensure high quality and carding efficiency of yarn carding, equipment needs to be overhauled and production parameters need to be adjusted in time at the current moment so as not to influence yarn quality. If the connected domain does not exist, the yarn fiber quality is better in the carding process.
In summary, in the embodiment of the present invention, a plurality of gray-scale images on the surface of a fiber during a fiber carding process are obtained, a pixel feature vector of each pixel point in the gray-scale images is calculated, and the pixel feature vectors are clustered to obtain a plurality of edge segments; acquiring the principal component direction of each edge segment, and calculating the bending degree of the edge segment and a distribution characteristic vector representing the orientation of the edge segment according to the principal component direction; traversing pixel points in the gray level image by using a sliding window, calculating the thickness degree of an edge segment where each pixel point is located, and obtaining the long fiber distribution uniformity of the center point of each sliding window according to the thickness degree and the bending degree to form a uniform distribution map; obtaining the winding degree of short fibers at the central point of each sliding window according to the number of long fibers and short fibers in the sliding window and the angle between the long fibers and the short fibers to form a winding distribution diagram; acquiring a uniform distribution map sequence and a winding distribution map sequence in a preset time period, and acquiring the carding abnormal degree of each pixel point according to the correlation between the uniform distribution map sequence and the winding distribution map sequence and the distribution change of the two sequences after carding in the time period to form an abnormal distribution map; and determining the abnormal area by using the abnormal distribution map. The embodiment of the invention can accurately monitor the abnormity in the carding process, and further can give out early warning before winding and prevent the spinning carding efficiency from becoming low by timely regulating and controlling production equipment.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (8)
1. The spinning carding process quality monitoring method based on image processing is characterized by comprising the following steps of:
acquiring a plurality of gray level images of the fiber surface in the fiber carding process, calculating a pixel characteristic vector of each pixel point in the gray level images, and clustering the pixel characteristic vectors to obtain a plurality of edge segments; the pixel feature vector represents the position and the distribution direction of the corresponding pixel point;
acquiring the principal component direction of each edge segment, and calculating the bending degree of the edge segment and a distribution characteristic vector representing the orientation of the edge segment according to the principal component direction;
traversing pixel points in the gray level image by using a sliding window, calculating the thickness degree of an edge segment where each pixel point is located, and obtaining the long fiber distribution uniformity of the center point of each sliding window according to the thickness degree and the bending degree to form a uniform distribution map;
obtaining the winding degree of the short fibers at the central point of each sliding window according to the number of the long fibers and the short fibers in the sliding window and the angle between the long fibers and the short fibers to form a winding distribution diagram;
acquiring a uniform distribution map sequence and a winding distribution map sequence in a preset time period, and acquiring the carding abnormal degree of each pixel point according to the correlation between the uniform distribution map sequence and the winding distribution map sequence and the distribution change of the two sequences after carding in the time period to form an abnormal distribution map; and determining the abnormal area by using the abnormal distribution map.
2. The method of claim 1, wherein the step of computing the pixel feature vector comprises:
acquiring a hessian matrix in the neighborhood of each pixel point in the gray level image, and taking a feature vector corresponding to the minimum feature value of the hessian matrix as a structural feature vector of the pixel point;
and combining the structural feature vector with the coordinates of the corresponding pixel point to obtain the pixel feature vector of the pixel point.
3. The method according to claim 1, wherein the edge segment is obtained by:
and extracting the edge of the gray image to obtain an edge image, acquiring the pixel characteristic vectors of pixel points belonging to the edge in the edge image, clustering the pixel characteristic vectors, and taking each category as the edge segment.
4. The method according to claim 1, wherein the distributed feature vector is obtained by:
and acquiring two principal component directions of all pixel points in the edge segment and second eigenvalues corresponding to the two principal component directions, and taking the unit vector of the principal component direction corresponding to the largest second eigenvalue as the distribution eigenvector.
5. The method of claim 4, wherein the degree of bending is obtained by:
and obtaining a difference vector of coordinates of each pixel point in the edge segment and the mean value coordinate, calculating an inner product of the difference vector and a unit vector in a principal component direction corresponding to the minimum second characteristic value, and taking the variance of the inner product corresponding to all the pixel points as the bending degree.
6. The method according to claim 2, wherein the step of obtaining the thickness degree comprises:
dividing the edge segments into long fibers and short fibers by setting a length threshold;
obtaining a normal vector of the structural feature vector of the pixel points belonging to the long fiber in the sliding window, and forming a pixel sequence by all the pixel points of the straight line where the normal vector is located in the gray level image;
and performing phase reversal processing on the pixel sequence, and then obtaining a plurality of pixel sets by watershed segmentation, wherein the number of the pixel points of the pixel set to which each pixel point belongs is taken as the thickness degree.
7. The method of claim 6, wherein the step of obtaining the uniformity of the distribution of the long fibers comprises:
calculating the variance of the thickness degree of the pixel points belonging to the long fiber in each sliding window;
obtaining the bending degree of each long fiber in the sliding window, and averaging to obtain an average bending degree;
and calculating the uniformity degree of the long fiber distribution according to the variance and the average bending degree.
8. The method according to claim 1, wherein the step of determining the region where the abnormality occurs comprises:
carrying out mean value filtering on the abnormal distribution map, and carrying out thresholding processing through a preset threshold value to obtain a position image;
and acquiring a region image of the position image by utilizing an on operation, and taking a connected domain in the region image as the abnormal region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111189614.5A CN113724253A (en) | 2021-10-13 | 2021-10-13 | Spinning carding process quality monitoring method based on image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111189614.5A CN113724253A (en) | 2021-10-13 | 2021-10-13 | Spinning carding process quality monitoring method based on image processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113724253A true CN113724253A (en) | 2021-11-30 |
Family
ID=78685831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111189614.5A Withdrawn CN113724253A (en) | 2021-10-13 | 2021-10-13 | Spinning carding process quality monitoring method based on image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113724253A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114283144A (en) * | 2022-03-06 | 2022-04-05 | 山东金有粮脱皮制粉设备有限公司 | Intelligent control method for stable operation of corncob crusher based on image recognition |
| CN114445348A (en) * | 2021-12-31 | 2022-05-06 | 扬州中卓泵业有限公司 | New material water pump defect detection method and system based on optical means |
| CN114708224A (en) * | 2022-03-31 | 2022-07-05 | 南通金斯曼纺织有限公司 | Artificial intelligence-based textile texture quality assessment method and system |
| CN114757901A (en) * | 2022-04-01 | 2022-07-15 | 海门市恒昌织带有限公司 | Textile carding system based on computer vision |
| CN115115615A (en) * | 2022-07-26 | 2022-09-27 | 南通好心情家用纺织品有限公司 | Textile fabric quality evaluation method and system based on image recognition |
| CN115496751B (en) * | 2022-11-16 | 2023-03-24 | 威海捷诺曼自动化股份有限公司 | Winding detection method of fiber winding machine |
| CN117994271A (en) * | 2024-04-07 | 2024-05-07 | 无锡文康科技有限公司 | Human brain image segmentation method based on watershed algorithm |
-
2021
- 2021-10-13 CN CN202111189614.5A patent/CN113724253A/en not_active Withdrawn
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114445348A (en) * | 2021-12-31 | 2022-05-06 | 扬州中卓泵业有限公司 | New material water pump defect detection method and system based on optical means |
| CN114445348B (en) * | 2021-12-31 | 2022-11-29 | 扬州中卓泵业有限公司 | New material water pump defect detection method and system based on optical means |
| CN114283144A (en) * | 2022-03-06 | 2022-04-05 | 山东金有粮脱皮制粉设备有限公司 | Intelligent control method for stable operation of corncob crusher based on image recognition |
| CN114283144B (en) * | 2022-03-06 | 2022-05-17 | 山东金有粮脱皮制粉设备有限公司 | Intelligent control method for stable operation of corncob crusher based on image recognition |
| CN114708224A (en) * | 2022-03-31 | 2022-07-05 | 南通金斯曼纺织有限公司 | Artificial intelligence-based textile texture quality assessment method and system |
| CN114708224B (en) * | 2022-03-31 | 2023-06-23 | 吴江市双泽纺织有限公司 | Textile texture quality assessment method and system based on artificial intelligence |
| CN114757901A (en) * | 2022-04-01 | 2022-07-15 | 海门市恒昌织带有限公司 | Textile carding system based on computer vision |
| CN115115615A (en) * | 2022-07-26 | 2022-09-27 | 南通好心情家用纺织品有限公司 | Textile fabric quality evaluation method and system based on image recognition |
| CN115115615B (en) * | 2022-07-26 | 2022-12-13 | 南通好心情家用纺织品有限公司 | Textile fabric quality evaluation method and system based on image recognition |
| CN115496751B (en) * | 2022-11-16 | 2023-03-24 | 威海捷诺曼自动化股份有限公司 | Winding detection method of fiber winding machine |
| CN117994271A (en) * | 2024-04-07 | 2024-05-07 | 无锡文康科技有限公司 | Human brain image segmentation method based on watershed algorithm |
| CN117994271B (en) * | 2024-04-07 | 2024-05-28 | 无锡文康科技有限公司 | Human brain image segmentation method based on watershed algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113724253A (en) | Spinning carding process quality monitoring method based on image processing | |
| CN114862862A (en) | Pump body cold shut defect identification method and system based on image processing | |
| CN116777941B (en) | Profile contour detection method and system based on machine vision | |
| CN114842007B (en) | Textile wear defect detection method based on image processing | |
| CN106683099A (en) | Product surface defect detection method | |
| Ying et al. | A study on the method of image pre-processing for recognition of crop diseases | |
| CN113610852B (en) | Yarn drafting quality monitoring method based on image processing | |
| Zhang et al. | A fast segmentation method for high-resolution color images of foreign fibers in cotton | |
| CN114841982B (en) | Air conditioner filter element quality monitoring method based on image processing | |
| CN114998321B (en) | Textile material surface hairiness degree identification method based on optical means | |
| Bodmarova et al. | Suitability analysis of techniques for flaw detection in textiles using texture analysis | |
| CN117745724B (en) | Stone polishing processing defect region segmentation method based on visual analysis | |
| Sharma et al. | Classification of breast tumors as benign and malignant using textural feature descriptor | |
| CN114565607B (en) | Fabric defect image segmentation method based on neural network | |
| Patil et al. | Automatic detection of microaneurysms in retinal fundus images using modified high boost filtering, line detectors and OC-SVM | |
| CN109919939B (en) | Yarn defect detection method and device based on genetic algorithm | |
| CN116563276A (en) | Chemical fiber filament online defect detection method and detection system | |
| Zhang et al. | An investigation of ramie fiber cross-section image analysis methodology based on edge-enhanced image fusion | |
| Wan et al. | A method for analyzing on-line video images of crystallization at high-solid concentrations | |
| TW201211348A (en) | Yarn detecting device and method thereof | |
| Krishna et al. | Unsupervised automated retinal vessel extraction framework using enhanced filtering and hessian based method with hysteresis thresholding | |
| Savakar et al. | Hidden Markov model for identification of different marks on human body in forensic perspective | |
| Sabino et al. | Automatic leukemia diagnosis | |
| Rippel et al. | Automated Segmentation of Profiled Fibers in cross-sectional Micrographs for Quality Control | |
| Yue et al. | Denoising and segmentation of digital feather image using mean shift algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211130 |
|
| WW01 | Invention patent application withdrawn after publication |