CN111968166B - Precise weaving method based on non-rigid body weaving image registration technology - Google Patents
Precise weaving method based on non-rigid body weaving image registration technology Download PDFInfo
- Publication number
- CN111968166B CN111968166B CN202010843275.7A CN202010843275A CN111968166B CN 111968166 B CN111968166 B CN 111968166B CN 202010843275 A CN202010843275 A CN 202010843275A CN 111968166 B CN111968166 B CN 111968166B
- Authority
- CN
- China
- Prior art keywords
- image
- contour
- point set
- point
- coarse
- 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.)
- Active
Links
- 238000009941 weaving Methods 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000005516 engineering process Methods 0.000 title claims abstract description 18
- 238000009987 spinning Methods 0.000 claims abstract description 42
- 239000004753 textile Substances 0.000 claims abstract description 38
- 238000013507 mapping Methods 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims abstract description 8
- 230000000877 morphologic effect Effects 0.000 claims abstract description 6
- 238000001035 drying Methods 0.000 claims abstract description 3
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000009826 distribution Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 238000005260 corrosion Methods 0.000 claims description 3
- 230000007797 corrosion Effects 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 3
- 238000009827 uniform distribution Methods 0.000 claims description 3
- 238000009940 knitting Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 8
- 239000004744 fabric Substances 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 2
- 241000251468 Actinopterygii Species 0.000 description 1
- 241000272814 Anser sp. Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000007596 consolidation process Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- -1 modeling Substances 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30124—Fabrics; Textile; Paper
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Woven Fabrics (AREA)
- Treatment Of Fiber Materials (AREA)
Abstract
The invention discloses an accurate spinning method based on a non-rigid body spinning image registration technology, which comprises the following steps: firstly, obtaining a coarse weaving pattern; contour extraction is carried out on the model pattern and the coarse weaving pattern to obtain a model image contour containing noise and a coarse weaving image contour containing noise; respectively carrying out drying treatment on the two contours by using morphological treatment to obtain a contour point set of the model image and a contour point set of the coarse weaving image; clustering the two point sets respectively to obtain a matching mapping relation between the class center of the model image contour and the class center of the coarse weaving image contour, and further obtaining an energy equation of non-rigid point set registration; optimizing the energy equation of non-rigid point set registration to obtain deformation parameters f (x) of contour points of each coarse woven image; and optimizing the textile parameters of the coarse weaving pattern according to the point set of the model image outline, the point set of the coarse weaving image outline and the deformation parameter f (x), so as to obtain accurate textile parameters, and carrying out accurate spinning.
Description
Technical Field
The invention belongs to the technical field of textile technology, and relates to an accurate textile method based on a non-rigid textile image registration technology.
Background
The economic benefits of fancy yarns and the influence of fancy yarns on the textile industry have been confirmed, and the traditional fancy yarn research at present is mainly focused on the characteristics of the yarns such as fiber raw materials, modeling, colors, structures and the like, and almost no product combining the fancy yarn structure with a precise pattern exists. The use of a precisely spun effect yarn to directly form a precisely designed pattern on the fabric face is known as precision weaving. It can maximize the utility value of fancy yarn.
Image registration is the process of matching, overlaying, and spatially aligning two or more images. Image registration can be described as a point set registration problem by feature point extraction of the target shape or original image. Although the point set registration algorithm is well applied in the fields of computer vision, pattern recognition, medicine, remote sensing image analysis and the like, the point set registration algorithm is rarely popularized in the industrial fields of textile and the like. In particular, the matching research on textile patterns is still blank. The registration technology is utilized to solve the relation problem between the pattern of the textile template and the deformed pattern, and further the precise textile of the fancy yarn is realized. Textile pattern registration techniques differ from general rigid registration in that the deformation of the pattern cannot be described simply by an integral translation, rotation, and dimensional change. Because different yarn materials have different elasticity, the method is different from common non-rigid point set registration; therefore, the existing registration of non-rigid point sets is not suitable for matching textile patterns, and the traditional intermediate processes of printing, jacquard and embroidery can lead to low production efficiency.
Disclosure of Invention
The invention aims to provide an accurate spinning method based on a non-rigid body spinning image registration technology, which solves the problem of low production efficiency in the prior art.
The technical scheme adopted by the invention is that the accurate spinning method based on the non-rigid body spinning image registration technology comprises the following steps:
step 1, designing a model pattern, and performing a coarse weaving link according to the model pattern to obtain a coarse weaving pattern;
step 2, carrying out contour extraction on the model pattern and the coarse weaving pattern to obtain a noisy model image contour and a noisy coarse weaving image contour;
step 3, respectively carrying out de-drying treatment on the noisy model image contour and the noisy coarse weaving image contour by using morphological treatment to obtain a model image contour point set and a coarse weaving image contour point set;
step 4, clustering the contour point sets of the model image and the contour point sets of the coarse weaving image respectively, calculating the class center of each class point set, and obtaining a matching mapping relation between the class center of the contour of the model image and the class center of the contour of the coarse weaving image, so as to obtain an energy equation for registering the non-rigid point sets;
step 5, optimizing an energy equation of non-rigid point set registration by using a maximum expected algorithm to obtain deformation parameters f (x) of contour points of each coarse woven image;
and 6, optimizing the textile parameters of the coarse weaving pattern according to the point set of the model image outline, the point set of the coarse weaving image outline and the deformation parameter f (x), so as to obtain accurate textile parameters, and carrying out accurate spinning.
The invention is also characterized in that:
the step 2 is specifically as follows: performing binarization treatment on the model pattern and the coarse weaving pattern, and then performing edge extraction on the model pattern and the coarse weaving pattern by using a gradient operator to obtain a noisy model image contour and a noisy coarse weaving image contour.
The step 3 is specifically as follows: and respectively performing expansion, corrosion and skeleton operation on the noisy model image contour and the noisy coarse weaving image contour by morphological treatment to obtain a model image contour point set and a coarse weaving image contour point set.
The step 4 specifically comprises the following steps:
step 4.1, assuming that the point set of the model image contour is x= { X i 1.ltoreq.i.ltoreq.m }, the set of points Y= { Y of the coarse weave image contour j 1.ltoreq.j.ltoreq.n }, clustering the contour point sets of the model image and the contour point sets of the coarse weave image by adopting a neighborhood algorithm, and calculating the class center of each class of point set, which is X respectively c ＝{x ck ,1≤k≤m c }，Y c ＝{y cl ,1≤l≤n c Assume class point y cl Relative to the point set X c Corresponding point x in (2) ck Obeying Gaussian distribution, the class point matching mapping relation is as follows:
and then obtaining a point class matching mapping relation:
in the above formula, p= [ P ] ij ]Representing a point matching matrix, P c ＝[q kl ]The class-matching matrix is represented as such,represents the kth class, < +.in Point set X>Representing a first class in the point set Y;
and 4.2, obtaining an energy equation of non-rigid point set registration according to the point matching mapping relation, wherein the energy equation comprises the following steps:
in the above formula, P represents a matching matrix, when P ij When=1, point x is represented i And y j Is the corresponding point, otherwise p ij =0; f (|θ) represents a deformation function, θ is a deformation parameter, and g is a regularization term for constraining the deformation parameter θ.
The step 5 specifically comprises the following steps:
step 5.1, assume contour Point y j Relative to the corresponding point X in the point set X i Obeying gaussian distribution, let z= { Z j } 1≤j≤n Representing y j Assignment of Gaussian distribution to point set X, then point matching matrixThe outlier likelihood function without the corresponding matching point is denoted as P (y j |z j ＝m+1)＝U(y j I a), wherein U (·) represents a uniform distribution of parameter a;the mixing density function is expressed as:
in the above-mentioned method, the step of,is a gaussian mixture weight, θ= { σ 2 Gamma, f are unknown parameters to be solved;
the same principle is obtained:
step 5.2, the formula (5) is brought into the formula (1) to obtain a class matching matrix:
and finally, calculating a point matching matrix by using a formula (2).
Step 5.3, calculating deformation parameters by utilizing the regenerated nuclear space theory, and updating the parameter sigma 2 Positions of gamma and heart-like contour points;
and 5.4, iterating the steps 5.2-5.3 until the energy process of the non-rigid point set registration is converged, and obtaining the deformation parameter f (x) of each coarse woven image contour point.
The step 5.3 specifically comprises the following steps:
step 5.3.1, first construct Gram matrixBeta is a normal number, then the solution of the deformation function is expressed as:
in the above formula, c satisfies c (Kdiag (1) T P)+λσ 2 I)＝YP；
Step 53.2 update parameters γ=1-N P /n，The locations of the heart-like contour points are then updated.
The step 6 specifically comprises the following steps:
step 6.1, supposing flat knitting time t 1 Spinning time t of knots 2 The parameter relation between the contour point set X of the model image and the contour point set Y of the coarse weaving image is as follows:
Y＝F(X，t 1 ,Ω 1 ，Ω 2 ) (8)；
in the above, Ω 1 Is the rotation speed of the front roller omega 2 The rotation speed of the hollow ingot;
and 2, obtaining an optimization equation of accurate textile parameters, wherein the optimization equation comprises the following steps:
in the above formula, d is an error distance measure;
solving the formula (9) to obtain accurate spinning parameters, and performing accurate spinning.
The beneficial effects of the invention are as follows:
according to the precise spinning method based on the non-rigid body spinning image registration technology, the deformation parameters of the coarse weaving pattern are obtained by establishing the matching mapping relation between the model pattern and the coarse weaving pattern, the spinning parameters of the coarse weaving pattern are optimized, the precise spinning pattern is obtained, and the rapid and efficient registration method from the template drawing to the spinning process and the spinning pattern generation is realized; the accurate spinning of the single fancy yarn in the common loom can be realized according to different patterns and requirements; the intermediate technological processes of printing, jacquard, embroidery and the like can be reduced, the production efficiency is improved, and further the added value of the product is improved.
Drawings
FIG. 1 is a process flow diagram of a precise textile method based on non-rigid textile image registration techniques of the present invention;
FIG. 2a is a schematic diagram of a pattern of a model in a precision weave method based on a non-rigid weave image registration technique according to the invention;
FIG. 2b is a schematic representation of a coarse weave pattern in a precise weave method based on a non-rigid weave image registration technique according to the invention;
FIG. 2c is a schematic illustration of a fine weave pattern in a fine weave method based on a non-rigid weave image registration technique according to the invention;
FIG. 3a is a schematic diagram of a tower-shaped pattern in a precision weaving method based on a non-rigid woven image registration technique according to the present invention;
FIG. 3b is a schematic diagram of a method for accurate textile registration based on non-rigid textile image according to the present invention after binarization of the tower-shaped pattern;
FIG. 3c is a schematic view of a precise textile method based on non-rigid textile image registration technique after extraction of the edges of the tower-shaped pattern;
FIG. 4a is a schematic diagram of a fish pattern in an accurate weave method based on non-rigid weave image registration techniques according to the invention;
FIG. 4b is a schematic diagram of a mapping relationship of fish-shaped patterns in an accurate spinning method based on a non-rigid body spinning image registration technique according to the present invention;
FIG. 4c is a schematic diagram showing the deformation of the fish-shaped pattern in an accurate spinning method based on the non-rigid body spinning image registration technique of the present invention;
fig. 4d is a schematic diagram of a fine woven fish-shaped pattern in a precise weaving method based on a non-rigid woven image registration technique according to the present invention.
Detailed Description
The invention will be described in detail below with reference to the drawings and the detailed description.
An accurate spinning method based on non-rigid body spinning image registration technology, as shown in figure 1, specifically comprises the following steps:
step 1, designing a model pattern, performing spinning process calculation according to the model pattern, and starting a coarse weaving process to obtain a coarse weaving pattern;
taking a knitted fabric pattern as an example, a pattern is formed on the cloth surface of the knitted fabric by utilizing a knot segment of one knot yarn, and the pattern is directly woven on the knitted fabric. FIG. 2a is a designed wild goose tower model pattern, and FIG. 2b is a rough weaving pattern effect; the side length of each square lattice is 1mm (the size of the lattice can be changed), the black small lattice represents the knot portion (knot lattice) of the knot line, the white small lattice represents the flat line portion (flat line lattice) of the knot line, and the knot length and the flat line length of the actual weaving pattern in the right side pattern are calculated from the number of left side lattices as one example. The calculation formula is shown as follows:
flat line length (cm) =number of flat line lattices×3×0.1
Junction length (cm) =number of junction lattices×3×0.1
In the coarse spinning process, when a computer control system is used for controlling a machine to spin yarn, the time t and the front roller rotating speed omega need to be input into process management software of the control system 1 Speed of rotation omega of hollow ingot 2 Three parameters. Therefore, the time and the rotation speed respectively corresponding to the junction part and the flat line part of the designed junction line are required.
(a) Setting of rotating speed: comprising the front roller rotating speed omega 1 And the rotation speed omega 2 of the hollow ingot
In the experiment, when spinning a flat line segment, the rotating speed of a front roller is 700, and the rotating speed of a hollow spindle is 800; in the spinning sub-part, because the spinning sub-part is a self-consolidation sub-line, each knot corresponds to three sections of process parameters, wherein the parameters of the first section are the front roller rotating speed 100, the hollow spindle rotating speed 1200, the parameters of the second section are the front roller rotating speed-50, the hollow spindle rotating speed 1200, and the parameters of the third section are the front roller rotating speed 100 and the hollow spindle rotating speed 1200. (the above parameters are obtained only empirically and can be adjusted according to actual conditions.)
(b) Setting time t: including flat segment part time and knot part time, t=t 1 +t 2 ；
The input front roller rotating speed and the actual roller rotating speed of the hollow spindle twisting machine have a proportional relation, and the actual rotating speed of the roller=0.011×the corresponding input motor rotating speed on software is obtained through measurement.
Flat line segment time: t is t 1 Length of flat lineActual rotation speed of roller= (number of flat-line lattices×0.3)/(0.011×motor rotation speed corresponding to input on software).
The knot is divided into three sections, so the corresponding three sections adopt the following calculation method:
junction part time t 2 Actual rotation speed of the knot length/roller= (number of knots×0.3)/(0.011×motor rotation speed corresponding to input on software).
Step 2, performing binarization treatment on the model pattern and the coarse weaving pattern, and then performing edge extraction on the model pattern and the coarse weaving pattern by using a gradient operator to obtain a noisy model image contour and a noisy coarse weaving image contour, as shown in figures 3a-3 c;
step 3, performing expansion, corrosion and skeleton operation on the noisy model image contour and the noisy coarse weaving image contour by morphological treatment respectively to obtain a model image contour point set and a coarse weaving image contour point set; taking a fish-shaped pattern as an example, as shown in fig. 4 a;
step 4, clustering the contour point sets of the model image and the contour point sets of the coarse weaving image respectively, calculating the class center of each class point set, and obtaining a matching mapping relation between the class center of the contour of the model image and the class center of the contour of the coarse weaving image, as shown in fig. 4b, so as to obtain an energy equation for registering the non-rigid point sets;
step 4.1, assuming that the point set of the model image contour is x= { X i 1.ltoreq.i.ltoreq.m }, the set of points Y= { Y of the coarse weave image contour j 1.ltoreq.j.ltoreq.n }, clustering the contour point sets of the model image and the contour point sets of the coarse weave image by adopting a neighborhood algorithm, and calculating the class center of each class of point set, which is X respectively c ＝{x ck ,1≤k≤m c }，Y c ＝{y cl ,1≤l≤n c Assume class point y cl Relative to the point set X c Corresponding point x in (2) ck Obeying Gaussian distribution, the class point matching mapping relation is as follows:
and then obtaining a point class matching mapping relation:
in the above formula, p= [ P ] ij ]Representing a point matching matrix, P c ＝[q kl ]The class-matching matrix is represented as such,represents the kth class, < +.in Point set X>Representing a first class in the point set Y;
and 4.2, obtaining an energy equation of non-rigid point set registration according to the point matching mapping relation, wherein the energy equation comprises the following steps:
in the above formula, P represents a matching matrix, when P ij When=1, point x is represented i And y j Is the corresponding point, otherwise p ij =0; f (|θ) represents a deformation function, θ is a deformation parameter, and g is a regularization term for constraining the deformation parameter θ.
Step 5, optimizing the energy equation of non-rigid point set registration by using a maximum expected algorithm to obtain deformation parameters f (x) of contour points of each coarse woven image, as shown in fig. 4 b;
the step 5 specifically comprises the following steps:
step 5.1, assume contour Point y j Relative to the corresponding point X in the point set X i Obeying gaussian distribution, let z= { Z j } 1≤j≤n Representing y j Assignment of Gaussian distribution to point set X, then point matching matrixThe outlier likelihood function without the corresponding matching point is denoted as P (y j |z j ＝m+1)＝U(y j I a), wherein U (·) is a tableA uniform distribution showing parameter a; the mixing density function is expressed as:
in the above-mentioned method, the step of,is a gaussian mixture weight, θ= { σ 2 Gamma, f are unknown parameters to be solved; the same principle is obtained:
step 5.2, bringing equation (5) into equation (1), the class-matching matrix can be expressed as:
and finally, calculating a point matching matrix by using a formula (2).
Step 5.3, calculating deformation parameters by utilizing the regenerated nuclear space theory, and updating the parameter sigma 2 Positions of gamma and heart-like contour points;
the step 5.3 specifically comprises the following steps:
step 5.3.1, first construct Gram matrixBeta is a normal number, then the solution of the deformation function is expressed as:
in the above formula, c satisfies c (Kdiag (1) T P)+λσ 2 I)＝YP；
Step 5.3.2, update parameter γ=1-N P /n，The locations of the heart-like contour points are then updated.
And 5.4, iterating the steps 5.2-5.3 until the energy process of the non-rigid point set registration is converged, and obtaining the deformation parameter f (x) of each coarse woven image contour point.
And 6, optimizing the weaving parameters of the coarse weaving pattern according to the point set of the model image outline, the point set of the coarse weaving image outline and the deformation parameter f (x) to obtain accurate weaving parameters, and performing accurate weaving to obtain the accurate weaving pattern, as shown in fig. 4 d.
Step 6.1, the picture width size of the common pattern is fixed, namely the size and the fixing of the knot length and the flat line length. Therefore, it can be assumed that t 1 、t 2 Sum of time (t) 1 +t 2 ) When constant. At this time, the parameter relationship between the contour point set X of the model image and the contour point set Y of the coarse weave image is:
Y＝F(X，t 1 ,Ω 1 ,Ω 2 ) (8)；
in the above, t 1 For flat yarn weaving time, omega 1 Is the rotation speed of the front roller omega 2 The rotation speed of the hollow ingot;
and 6.2, obtaining an optimization equation of accurate textile parameters, wherein the optimization equation is as follows:
in the above formula, d is an error distance measure, such as euclidean distance. In general, the F in the formula (8) can be solved by a nonlinear fitting method to obtain an approximate solution, so that the analytical solution in the formula (9) can be iteratively solved by adopting a convex optimization method to obtain accurate textile parameters; after the precise weaving, a precise weaving pattern is obtained, and fig. 2c shows the effect of the precise weaving pattern.
Through the mode, the accurate spinning method based on the non-rigid body spinning image registration technology, disclosed by the invention, obtains the deformation parameters of the coarse weaving pattern by establishing the matching mapping relation between the model pattern and the coarse weaving pattern, optimizes the spinning parameters of the coarse weaving pattern to obtain the accurate spinning pattern, and realizes a rapid and efficient registration method from template drawing to spinning process and spinning pattern generation; the accurate spinning of the single fancy yarn in the common loom can be realized according to different patterns and requirements; the intermediate technological processes of printing, jacquard, embroidery and the like can be reduced, the production efficiency is improved, and further the added value of the product is improved.
Claims (7)
1. An accurate spinning method based on a non-rigid body spinning image registration technology is characterized by comprising the following steps:
step 1, designing a model pattern, and performing a coarse weaving link according to the model pattern to obtain a coarse weaving pattern;
step 2, carrying out contour extraction on the model pattern and the coarse weaving pattern to obtain a noisy model image contour and a noisy coarse weaving image contour;
step 3, respectively carrying out drying treatment on the noisy model image contour and the noisy coarse weaving image contour by morphological treatment to obtain a model image contour point set and a coarse weaving image contour point set;
step 4, clustering the model image contour point sets and the coarse weaving image contour point sets respectively, calculating class centers of each class of point set, and obtaining a matching mapping relation between the class centers of the model image contour and the class centers of the coarse weaving image contour, so as to obtain an energy equation of non-rigid point set registration;
step 5, optimizing the energy equation of the non-rigid point set registration by using a maximum expected algorithm to obtain deformation parameters f (x) of contour points of each coarse woven image;
and 6, optimizing the textile parameters of the coarse weaving pattern according to the point set of the model image outline, the point set of the coarse weaving image outline and the deformation parameter f (x), so as to obtain accurate textile parameters, and carrying out accurate spinning.
2. The precise textile method based on the non-rigid textile image registration technology according to claim 1, wherein the step 2 is specifically: performing binarization treatment on the model pattern and the coarse weaving pattern, and then performing edge extraction on the model pattern and the coarse weaving pattern by utilizing a gradient operator to obtain a noisy model image contour and a noisy coarse weaving image contour.
3. The precise textile method based on the non-rigid textile image registration technology according to claim 1, wherein the step 3 is specifically: and respectively performing expansion, corrosion and skeleton operation on the noisy model image contour and the noisy coarse weaving image contour by morphological treatment to obtain a model image contour point set and a coarse weaving image contour point set.
4. The precise textile method based on the non-rigid textile image registration technology according to claim 1, wherein the step 4 specifically comprises the following steps:
step 4.1, assuming that the point set of the model image contour is x= { X i 1.ltoreq.i.ltoreq.m }, the set of points Y= { Y of the coarse weave image contour j 1.ltoreq.j.ltoreq.n }, clustering the contour point sets of the model image and the contour point sets of the coarse weave image by adopting a neighborhood algorithm, and calculating the class center of each class of point set, which is X respectively c ＝{x ck ,1≤k≤m c }，Y c ＝{y cl ,1≤l≤n c Assume class point y cl Relative to the point set X c Corresponding point x in (2) ck Obeying Gaussian distribution, the class point matching mapping relation is as follows:
and then obtaining a point class matching mapping relation:
in the above formula, p= [ P ] ij ]Representing a point matching matrix, P c ＝[q kl ]The class-matching matrix is represented as such,represents the kth class in the point set X,representing a first class in the point set Y;
and 4.2, obtaining an energy equation of non-rigid point set registration according to the point matching mapping relation, wherein the energy equation comprises the following steps:
in the above formula, P represents a matching matrix, when P ij When=1, point x is represented i And y j Is the corresponding point, otherwise p ij =0; f (|θ) represents a deformation function, θ is a deformation parameter, and g is a regularization term for constraining the deformation parameter θ.
5. The precise textile method based on the non-rigid textile image registration technology according to claim 4, wherein the step 5 specifically comprises the following steps:
step 5.1, assume contour Point y j Relative to the corresponding point X in the point set X i Obeying gaussian distribution, let z= { Z j } 1≤j≤n Representing y j Assignment of Gaussian distribution to point set X, then point matching matrixThe outlier likelihood function without the corresponding matching point is denoted as P (y j |z j ＝m+1)＝U(y j I a), wherein U (·) represents a uniform distribution of parameter a; the mixing density function is expressed as:
in the above-mentioned method, the step of,is a gaussian mixture weight, θ= { σ 2 Gamma, f are unknown parameters to be solved;
the same principle is obtained:
step 5.2, the formula (5) is brought into the formula (1) to obtain a class matching matrix:
finally, calculating a point matching matrix by using a formula (2);
step 5.3, calculating deformation parameters by utilizing the regenerated nuclear space theory, and updating the parameter sigma 2 Positions of gamma and heart-like contour points;
and 5.4, iterating the steps 5.2-5.3 until the energy process of the non-rigid point set registration converges, and obtaining a deformation parameter f (x) of each coarse weaving image contour point.
6. The precise textile method based on the non-rigid textile image registration technology according to claim 5, wherein the step 5.3 specifically comprises the following steps:
step 5.3.1, first construct Gram matrixBeta is a normal number, then the solution of the deformation function is expressed as:
in the above formula, c satisfies c (Kdiag (1) T P)+λσ 2 I)＝YP；
Step 5.3.2, update parameter γ=1-N P /n，The locations of the heart-like contour points are then updated.
7. The precise textile method based on the non-rigid textile image registration technology according to claim 1, wherein the step 6 specifically comprises the following steps:
step 6.1, supposing flat knitting time t 1 Spinning time t of knots 2 The parameter relation between the contour point set X of the model image and the contour point set Y of the coarse weaving image is as follows:
Y＝F(X，t 1 ，Ω 1 ，Ω 2 )(8)；
in the above, Ω 1 Is the rotation speed of the front roller omega 2 The rotation speed of the hollow ingot;
and 2, obtaining an optimization equation of accurate textile parameters, wherein the optimization equation comprises the following steps:
in the above formula, d is an error distance measure;
solving the formula (9) to obtain accurate spinning parameters, and performing accurate spinning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010843275.7A CN111968166B (en) | 2020-08-20 | 2020-08-20 | Precise weaving method based on non-rigid body weaving image registration technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010843275.7A CN111968166B (en) | 2020-08-20 | 2020-08-20 | Precise weaving method based on non-rigid body weaving image registration technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111968166A CN111968166A (en) | 2020-11-20 |
| CN111968166B true CN111968166B (en) | 2023-09-29 |
Family
ID=73389224
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010843275.7A Active CN111968166B (en) | 2020-08-20 | 2020-08-20 | Precise weaving method based on non-rigid body weaving image registration technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111968166B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113470084B (en) * | 2021-05-18 | 2024-01-30 | 西安电子科技大学 | Point set registration method based on outline rough matching |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102262781A (en) * | 2011-05-11 | 2011-11-30 | 浙江工业大学 | Method for registration of ink-jet printing texture image based on unit decomposition optical flow field |
| CN106485741A (en) * | 2016-10-19 | 2017-03-08 | 哈尔滨工业大学 | A kind of method of the non-firm point set registration retaining partial structurtes |
| CN107545584A (en) * | 2017-04-28 | 2018-01-05 | 上海联影医疗科技有限公司 | The method, apparatus and its system of area-of-interest are positioned in medical image |
| CN108711168A (en) * | 2018-06-04 | 2018-10-26 | 中北大学 | Non-rigid multimodal medical image registration method based on ZMLD Yu GC discrete optimizations |
| CN110363800A (en) * | 2019-06-19 | 2019-10-22 | 西安交通大学 | A kind of accurate rigid registration method blended based on point set data and characteristic information |
| CN110415281A (en) * | 2019-07-30 | 2019-11-05 | 西安交通大学深圳研究院 | A kind of point set rigid registration method based on Loam curvature weighting |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090239025A1 (en) * | 2008-03-04 | 2009-09-24 | High Voltage Graphics, Inc. | Flocked articles having a woven graphic design insert and methods of making the same |
| CN105389774B (en) * | 2014-09-05 | 2019-03-01 | 华为技术有限公司 | The method and apparatus for being aligned image |
| US9721358B2 (en) * | 2015-08-13 | 2017-08-01 | Excelsius Medical Co., Ltd. | Method, system, and non-transitory computer readable medium for video-based circular object localization |
| US11080816B2 (en) * | 2019-01-28 | 2021-08-03 | Ying Ji | Image measuring and registering method |
-
2020
- 2020-08-20 CN CN202010843275.7A patent/CN111968166B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102262781A (en) * | 2011-05-11 | 2011-11-30 | 浙江工业大学 | Method for registration of ink-jet printing texture image based on unit decomposition optical flow field |
| CN106485741A (en) * | 2016-10-19 | 2017-03-08 | 哈尔滨工业大学 | A kind of method of the non-firm point set registration retaining partial structurtes |
| CN107545584A (en) * | 2017-04-28 | 2018-01-05 | 上海联影医疗科技有限公司 | The method, apparatus and its system of area-of-interest are positioned in medical image |
| CN108711168A (en) * | 2018-06-04 | 2018-10-26 | 中北大学 | Non-rigid multimodal medical image registration method based on ZMLD Yu GC discrete optimizations |
| CN110363800A (en) * | 2019-06-19 | 2019-10-22 | 西安交通大学 | A kind of accurate rigid registration method blended based on point set data and characteristic information |
| CN110415281A (en) * | 2019-07-30 | 2019-11-05 | 西安交通大学深圳研究院 | A kind of point set rigid registration method based on Loam curvature weighting |
Non-Patent Citations (2)
| Title |
|---|
| 用于特征点配准的快速聚类凸集投影算法;连玮;梁彦;潘泉;程咏梅;张洪才;;自动化学报(第03期);全文 * |
| 采用多角度成像技术的纺织品三维轮廓重建算法;孟想;辛斌杰;李佳平;;纺织学报(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111968166A (en) | 2020-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103106645B (en) | A kind of recognition method for woven fabric structure | |
| CN101859335B (en) | Computer-aided crewel embroidery production method | |
| CN111968166B (en) | Precise weaving method based on non-rigid body weaving image registration technology | |
| CN106778881A (en) | Digital printing method and device | |
| CN109993755B (en) | Jacquard fabric image weave structure segmentation method | |
| CN105550660A (en) | Woven fabric weave structure type identification method | |
| CN107966444A (en) | Textile flaw detection method based on template | |
| CN106875459B (en) | Self-adaptive equalization method for color jacquard weave structure based on image segmentation | |
| CN114622322A (en) | System and method for quickly generating brocade makeup satin fabric pattern | |
| CN114386295A (en) | Textile computer simulation method based on color separation and color change of colored spun yarns | |
| CN109785283A (en) | A kind of textural characteristics matching process and device for fabric segmentation | |
| CN111353247B (en) | Method for identifying and reconstructing mesoscopic components of plain weave structure of ceramic matrix composite | |
| CN110895707B (en) | Method for judging depth of clothes type in washing machine under strong shielding condition | |
| CN109543733B (en) | Quick production method of yarn-dyed fabric based on cloud platform | |
| CN116778174A (en) | Open-width type single facer control method and system | |
| CN109591472B (en) | Digital ink-jet printing method for warp-knitted vamp based on vision | |
| CN111709429A (en) | Method for identifying structural parameters of woven fabric based on convolutional neural network | |
| CN106934846A (en) | A kind of cloth image processing method and system | |
| CN106600629B (en) | A kind of light stream estimation method towards Large Scale Motion | |
| CN113604949A (en) | Manufacturing method of glass fiber reinforced plastic structure based on knitted fabric | |
| CN109918783B (en) | Intelligent clothing design system | |
| CN115341323B (en) | Preparation method of double-sided woven-printed combined fabric | |
| CN112966637A (en) | Fabric texture classification automatic identification method based on deep learning | |
| CN115100481B (en) | Textile qualitative classification method based on artificial intelligence | |
| CN109377489A (en) | A kind of analysis method and analysis system of the organization construction of woven |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |