CN102523366B - Automatic fabric weaving style analyzing system and method - Google Patents
Automatic fabric weaving style analyzing system and method Download PDFInfo
- Publication number
- CN102523366B CN102523366B CN201110415317.8A CN201110415317A CN102523366B CN 102523366 B CN102523366 B CN 102523366B CN 201110415317 A CN201110415317 A CN 201110415317A CN 102523366 B CN102523366 B CN 102523366B
- Authority
- CN
- China
- Prior art keywords
- fabric
- image
- primitive
- information
- color
- 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.)
- Expired - Fee Related
Links
Landscapes
- Treatment Of Fiber Materials (AREA)
Abstract
The invention discloses an automatic fabric weaving style analyzing system, comprising a computer image processing device and two imaging devices. The two imaging devices are positioned at two sides of a fabric plane to obtain a double-surface image of the fabric, and the double-surface image is processed and analyzed by a computer image processing device so as to obtain a fabric weaving style. Lenses of the two imaging devices are positioned on the two sides of the fabric, take the fabric plane as an intermediate symmetrical surface and are opposite in direction. The imaging devices adopt the CCD (Couple-Charged Device) lenses which are identical in performances and specifications. A fabric is fixed by a sample fixture which consists of a transparent double-surface panel and a fixing frame, wherein one side of the double-surface panel is combined by a hinge, so that the double-surface panel can be opened and closed, a fabric sample to be detected is put at the proper position flatly, two pair of positioning pieces are arranged on the sample fixture, and a detecting area can be positioned by movement of the positioning pieces. The automatic fabric weaving style analyzing system can generate a complete fabric weaving style mold in real time.
Description
Technical field
The invention belongs to field of fabric producing technology, particularly a kind of automatic fabric weaving style analyzing system and method.
Background technology
Weaving cotton cloth by warp thread and weft yarn is weave patterns weaving according to some predefined structures, and weave patterns can represent by two dimension (plane) weave diagram.Weave patterns is basic structure information and the quality control of each braiding of weaving part in Fabric Design during fabric generates.But, weave pattern is always not visible in this workflow, and in final buyer and client's hand, is completely invisible.Based on the consideration of fabric product reverse-engineering, be necessary weave pattern to analyze, and set up a rapid-action system take braid as prototype.This work at present remains based on craft and the simple tool of some and completes, when working expenditure, loaded down with trivial details, and arduous, and its quality is also difficult to control.
At present area of computer aided Dobby Weave Design technology is comparatively ripe, but fabric analysis to fabric sample and classification model construction method are still in the exploratory stage.Use real time imagery and computer image processing technology, from textile image, extract characteristic parameter and automatically identify fabric tissue node, become new study hotspot and difficult point.Existing many scholars have proposed some effective algorithms both at home and abroad.From image processing method angle, a kind of is directly in the spatial domain of image, to adopt the means such as gray analysis, edge enhancing, thresholding, histogram equalization, filtering, with decision node type.As the people such as Kang utilize the transmitted light images of fabric, locating yarn position is come to as gray analysis in yarn gap, obtain fineness and density through weft yarn; The ratio of the line of apsides of the ellipticity projection presenting by longitude and latitude intertwined point in reflected light image is determined interlacing point type.Another kind method is by fast Fourier transform, two-dimensional spectrum analysis, wavelet transformation etc., and image is calculated again, analyzed from spatial domain is converted to frequency domain.Traditional fabric construction analytical method is owing to only adopting the one-sided image visual field information of fabric, tends to cause the image information deficiency obtained, especially, in the time that fabric has more complicated pattern, is difficult to weave pattern to carry out accurate analysis.
Summary of the invention
The object of this invention is to provide a kind of automatic fabric weaving style analyzing system and method, cannot carry out to automatic fabric weaving style the problem of accurate analysis to solve available technology adopting list imaging device.
Technical scheme of the present invention is, a kind of automatic fabric weaving style analyzing system, this system comprises Computer Image Processing device and 2 imaging devices, 2 described imaging devices are positioned at the both sides of described fabric plane, obtain the dual-side image of described fabric, by the Treatment Analysis of Computer Image Processing device, obtain the weave pattern of fabric.
The camera lens of 2 described imaging devices is positioned at the both sides of described fabric, and take fabric plane as the middle plane of symmetry, direction is relative.
Described imaging device camera lens adopts CCD camera lens, and specification is in full accord.
Described fabric is fixed by sample clamp, sample clamp is made up of transparent two surface plate and fixed mount, wherein, combined by a hinge on one side, can open and close, detected fabric sample is entirely put into correct position wherein, also has two pairs of spacers above sample clamp, by running fix sheet, can realize surveyed area location.
A kind of automatic fabric weaving style automatic analysis method, is characterized in that, comprises the following steps:
A1, adopt 2 imaging devices to carry out image acquisition to region corresponding to fabric positive and negative, by Computer Image Processing device, the fabric two-face image obtaining is comprised the preliminary treatment to two-sidedly positioning of obtaining image, registration, removal noise processed and shear treatment of the location that utilizes fixture.
A2, utilize the brightness information in image warp, latitude direction, textile image is carried out to grid to be divided automatically, by analyzing the colouring information of the grid image divided correspondence position aspect positive and negative two, determine the alternate position of yarn, thereby determine node location, isolate the structural motif of fabric;
A3, adopts automatic scan mode to carry out color set division to the primitive of grid, then determines representative color aggregate information, makes primitive distribute the most rational color;
A4, analyzes the primitive of the basic cycling element of fabric, and primitive is carried out to color and structural information coding, and sets up a primitive storage organization message file that comprises the information such as node type and color, for reconstruct fabric patterns structural model;
A5, the establishment styles store structural information file of setting up fabric deposits database in, for the quality of fabric is evaluated.
The method of definite employing color cluster of the color of primitive realizes, comprises,
The textile image gathering is represented with characteristic color, just can describe in right amount by a feature the primitive image-region of analysis image,
C
f＝[S(x
1)，S(x
2)，...S(x
k)]
T
Wherein, S (x
i) representation feature concentrate, the number of i feature shared pixel in this image area, C
fthe vector representing just can reflect the difference existing between them;
Definition texture similarity
Its value is larger, the color characteristics in two regions is more similar, according to fabric color feature, choose the image array window of 1.5～2 times being of a size of through weft yarn elementary cell, color property vector in calculation window, as the sample characteristics of fabric, adopt the method for scanning to the each pixel in image, calculate its characteristic vector in a certain size the field window of setting and the similarity measure with sample characteristics, obtain the new images of a width about similarity measure, according to the similarity of image, obtain a more completely primitive target area roughly by the method for threshold value, and distribute its characteristic color.
The method that realizes primitive location is, utilizes relevant and brightness information, and it is to utilize projection algorithm that textile image is carried out to grid divides automatically, and its gray value is mapped to two independently one dimension waveforms,
G in formula
k(j) be the gray value of k two field picture j row, G
k(i, j) is the grey scale pixel value of (i, j) position on k two field picture, and G
k(i) be the capable gray value of k two field picture i, utilize drop shadow curve's ripple kurtosis to realize after the Preliminary division of grid, recycling is relevant to be adjusted dividing with monochrome information etc., realizes accurate gridding.
The primitive of fabric patterns model adopts domain to represent, its structure comprises status field, color gamut, intertwined point domain information.
Further, in the process of image lattice, weft position parameter deterministic process comprises the following steps:
B1, first, from the result of the spread parameter analysis of the preliminary gridding of weft yarn, sets up weft yarn cell picture;
B2, carries out coefficient correlation analysis to weft yarn cell picture, finds out weft yarn period;
B3, by same-phase weft yarn unit, sets up similar weft yarn cell picture;
B4, obtains the warp-wise brightness of similar weft yarn cell picture, analyzes to distinguish through interlacing point region and latitude interlacing point region by warp-wise luminance signal;
If mathematic expectaion and the variance of stochastic variable X, Y all exist, the coefficient R of X and Y is:
In formula: the mathematic expectaion that E (X) is X, the mathematic expectaion that E (Y) is Y, the variance that D (X) is X, the variance that D (Y) is Y, R represents the linear dependence between X, Y, and, in the time that variable X increases, variable Y has the trend that reduces (R < 0) by linearity increase (R > 0) or linearity, for variable X (n), its auto-correlation function is R (T):
R(τ)＝E[X(n)·X(n+τ)]
The weft yarn cell picture being partitioned into by means of Weft Arrangement Parameters analysis, and set up coordinate system on weft yarn cell picture, x axle is parallel to weft yarn, y axle is parallel to warp thread, the coordinate of arbitrary pixel is (x, y), and brightness value is f (x, y), warp-wise brightness (average of each row pixel brightness) is:
Serial number according to weft yarn unit in fabric is respectively 1,2,3,4,5,6 ... these weft yarn cell pictures are asked to coefficient correlation between two, in the time that any two weft yarn cell pictures are done to warp-wise brightness correlation analysis, if coefficient correlation is greater than 0.5, show that these two weft yarns have the identical rule that interweaves, by the 1st, 2 ... consecutive operations analysis and the differentiation of root weft yarn and follow-up weft yarn coefficient correlation, obtain weft yarn period, and then draw same-phase weft yarn unit
Weft yarn cell picture is divided into behind interlacing point region and latitude interlacing point region, can determine warp thread spread parameter by the variation of analyzing interlacing point region, similar weft yarn cell picture to each weft yarn place in a weft yarn Weaving Cycle carries out signal processing, obtain warp-wise brightness square-wave signal, draw latitude interlacing point region according to square-wave signal.
Further, determine warp thread location parameter with reference to weft position parameter deterministic process, and then textile image is carried out to gridding obtain preliminary automatic fabric weaving style.
The present invention adopts the dual-side image information of fabric to contribute to increase the amount of information of textile image.One-sided image only comprises the state information that interweaves of half between warp thread and weft yarn in fact, and the geometry that thread segment is complete cannot observe from single face, only has the fabric two-face visual field can obtain complete interweaving information.
The present invention proposes a kind of CCD imaging system, and utilizes two CCD systems, realizes the direct picture to fabric and corresponding reverse side mirror image simultaneously and gathers and analyzing and processing.First gather the positive and negative two sides image information of fabric simultaneously, by the monochrome information feature on analysis fabric direction of warp and weft and the colouring information of two sides node, determine the alternate position of yarn, sample image is carried out to primitive separation and color coding.Utilize the attributive character parameter of primitive own and its arrangement mode to realize pattern the interlacing point type of basic cycling element is analyzed, and by the neighbor information of yarn, it is carried out to correction process and obtain accurate result.Finally set up a fabric analysis model that simultaneously comprises interlacing point type and colouring information.
The present invention can generate complete automatic fabric weaving style model in real time, can effectively explain weave pattern and the weaving quality of fabric, contribute to fabric mechanism parameter automatic analysis, and realize the reverse structural remodeling of weave patterns, lay the foundation for setting up automatic fabric weaving style building database system and evaluation system.
Accompanying drawing explanation
Two CCD imaging system schematic diagrames in Fig. 1 embodiment of the present invention.
Appearance of fabrics image digitazation schematic flow sheet in Fig. 2 embodiment of the present invention.
The gridding schematic diagram of textile image in Fig. 3 embodiment of the present invention.
The color coding table of the front-back two-sided image of fabric in Fig. 4 embodiment of the present invention.
Fabric construction coding schedule in Fig. 5 embodiment of the present invention.
Fabric patterns illustraton of model in Fig. 6 embodiment of the present invention.
Embodiment
Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiment are only not used in and limit the scope of the invention for the present invention is described.In addition should be understood that those skilled in the art can make various changes or modifications the present invention after having read the content of the present invention's instruction, these equivalent form of values fall within the application's appended claims limited range equally.
Hardware device of the present invention comprises CCD imaging front, Computerized image processing system.Two CCD imaging systems described in it is characterized in that are carried out double-face imaging to fabric, as shown in Figure 1.Wherein 1.CCD, 2. spacer, 3. fixture, 4. detects sample.
Described CCD imaging front is connected with Computerized image processing system, for gathering the dual-side image of fabric; Described Computerized image processing system, for the image collecting is carried out to analyzing and processing, finally carries out mathematical modeling to the weave pattern of fabric.Its device comprises following functions: CCD imaging front: can be stable obtain fabric tow sides image, and be sent in Computerized image processing system and process in real time.
The method of dual-side image of obtaining fabric is as follows: the detection zone of (1) selected sample, and adopt spacer location.Adopt spacer to position the surveyed area of sample; (2) dual-side image of the automatic acquisition sample by machine vision; (3) by image processing, eliminate due to the caused distortion of difference due to position in positive and negative imaging process, so that it is strictly corresponding to carry out the arrangement of mirror image primitive, and according to witness marker, intercept the image-region of studying; (4), for the analysis to image is correct, adopt edge pruning technology, the impact that the boundary information of removal of images is analyzed textile image.Adopt two-sided sampled picture analysis and mutually proofread and correct, being conducive to the accuracy of yarn location, strengthening identification and the positioning precision of primitive.Its network analysis flow process, as shown in Figure 2.
The one side visual field of only having considered fabric from traditional fabric construction analytical method is different, adopt two-sided sampling analysis, can overcome in an one-sided image due to information deficiency, realize weave pattern is carried out to accurate analysis, be particularly conducive to the information characteristics of the complicated pattern fabric primitive of accurate analysis and location.One-sided image in fact only comprises the state information that interweaves of half between warp thread and weft yarn, and the complete information of yarn braiding is difficult to observe and obtain from single fabric side; Adopt the dual-side image information of fabric corresponding region to verify and error correction for the state information that interweaves of the fabric of analysis.
Because the color of textile image is to enrich very much, in order to realize separation and the location of picture element, image need to be carried out to gridding.The basis of image lattice is color and the strength information that utilizes fabric face image yarn, but because the colouring information of imaging surface is very abundant, need to adopt the algorithm of color cluster, is collected as the number of colours identical with composition yarn variety.In fact, different from collocation of colour with colour correction, we do not need to extract the consistency that an actual reflectance spectrum guarantees that color is assembled.Conventionally characterize color with original rgb value.The present invention directly utilizes the brightness information in image warp, latitude direction, textile image is carried out to grid and automatically divide.
It is to utilize projection algorithm that textile image is carried out to automatic division of grid, and its gray value is mapped to two independently one dimension waveforms, and its formula is:
G in formula
k(j) be the gray value of k two field picture j row.G
k(i, j) is the grey scale pixel value of (i, j) position on k two field picture.And G
k(i) be the capable gray value of k two field picture i.Utilize drop shadow curve's ripple kurtosis to realize after the Preliminary division of grid.
In to the process of image lattice, precisely to cut apart in order to realize grid, on textile image, the accurate location of warp, weft threads position is the prerequisite of fabric tissue identification.Weft position parameter deterministic process can be divided into the following step:
1) first from the result of the spread parameter analysis of the preliminary gridding of weft yarn, set up weft yarn cell picture;
2) weft yarn cell picture is carried out to coefficient correlation analysis, find out weft yarn period;
3), by same-phase weft yarn unit, set up similar weft yarn cell picture;
4) obtain the warp-wise brightness of similar weft yarn cell picture, analyze to distinguish through interlacing point region and latitude interlacing point region by warp-wise luminance signal;
If mathematic expectaion and the variance of stochastic variable X, Y all exist, the coefficient R of X and Y is:
In formula: the mathematic expectaion that E (X) is X, the mathematic expectaion that E (Y) is Y, the variance that D (X) is X, the variance that D (Y) is Y.
R represents the linear dependence between X, Y, and, in the time that variable X increases, variable Y has the trend that reduces (R < 0) by linearity increase (R > 0) or linearity.For variable X (n), its auto-correlation function is R (T):
R(τ)＝E[X(n)·X(n+τ)]
The weft yarn cell picture being partitioned into by means of Weft Arrangement Parameters analysis, and set up coordinate system on weft yarn cell picture, x axle is parallel to weft yarn, y axle is parallel to warp thread, the coordinate of arbitrary pixel is (x, y), and brightness value is f (x, y), warp-wise brightness (average of each row pixel brightness) is:
Serial number according to weft yarn unit in fabric is respectively 1,2,3,4,5,6 ..., these weft yarn cell pictures are asked to coefficient correlation between two.In the time that any two weft yarn cell pictures are done to warp-wise brightness correlation analysis, if coefficient correlation is greater than 0.5, show that these two weft yarns have the identical rule that interweaves, by to the 1st, 2 ... consecutive operations analysis and the differentiation of root weft yarn and follow-up weft yarn coefficient correlation, just can obtain weft yarn period, and then can draw same-phase weft yarn unit.
Weft yarn cell picture is divided into behind interlacing point region and latitude interlacing point region, can determine warp thread spread parameter by the variation of analyzing interlacing point region.First the similar weft yarn cell picture at each weft yarn place in a weft yarn Weaving Cycle is carried out to signal processing, obtain warp-wise brightness square-wave signal, draw latitude interlacing point region according to square-wave signal.In like manner can determine warp thread location parameter.And then textile image is carried out to gridding can obtain preliminary automatic fabric weaving style.As shown in Figure 3.
Primitive color alignment mode is a key character of automatic fabric weaving style, but in fabric, due to the weaving manner difference of yarn, the color distribution that has formed primitive is also different, by the method for color cluster, contributes to realize the separation of primitive.Propose the method for color characteristic collection herein, determine the color of primitive.Color characteristic collection comprise color number, corresponding braiding sample yarn color kind number.
Determine the characteristic color set of fabric face primitive, we adopt two principles, are first the principles of maximum frequency, are secondly the compatible principles of color.The principle of maximum frequency has guaranteed that the feature of the N in color set has represented the distribution of the yarn color in braiding sample image, and the Compatibility Principle of color requires each feature in color space, to have distribution relatively uniformly.Can adopt the color model in compatible spheroid space, each feature has the spheroid space that a radius is R, and in image, color whole merger in this space are this feature, and each feature is to each other apart from being greater than R.Obviously, compatible scope is larger, and R is larger, and the quantity of feature is just fewer, and we can and select the value of R by the method for experiment according to the sample fabric of research.Value by R just can be carried out concrete definite characteristic color set according to algorithm.After image represents with characteristic color, just can describe in right amount by a feature a certain image-region.
C
f＝[S(x
1)，S(x
2)，...S(x
k)]
T
Wherein: S (x
i) representation feature concentrate, the number of I feature shared pixel in this image area.C
fthe vector representing just can reflect the difference existing between them.So can define a texture similarity, that is:
Its value is larger, and the color characteristics in two regions is more similar.According to fabric color feature, choose a certain size rectangular window, color property vector in calculation window, as the sample characteristics of fabric, then to the each pixel in image, calculate its characteristic vector in a certain size field window and the similarity measure with sample characteristics, obtain like this new images of a width about similarity measure.Obviously,, according to the similarity of image, can obtain a more completely primitive target area roughly by the method for threshold value.
According to primitive color, in fact realize tentatively cutting apart of textile image primitive.Fig. 4 has shown the color coding table of the front-back two-sided image of fabric.But, this first requirement of cutting apart that is also difficult to meet its primitive of cutting apart.Consider the impact of yarn fine hair, different study samples, the size and shape of primitive is not identical, and color, shape and the arrangement mode etc. of cloth fabric face primitive image, has just formed the weave pattern of fabric.The primitive that textile image marks off should be generally more regular rectangular-shaped, but due to the impact of picture noise, isolated primitive often reaches desirable effect.
In order to be accurately partitioned into primitive, first adopt the chain representation based on zone boundary, extract the information from objective pattern of primitive image. the existing calculating that is beneficial to morphological feature of this chain representation, is also conducive to save memory space.Realize image is carried out to edge tracking with chain code, can obtain as the geometric characteristic of a series of tuberculosis cells such as girth, area, width.Then we go out the arrangement mode of primitive according to Preliminary division, by the method for statistics, primitive are realized to grid and cut apart.
Primitive is the base unit that forms weave pattern.Primitive figure is an X-Y scheme, and its arrangement mode and attribute are to form weaving textile pattern.In fact, corresponded respectively to warp thread and the weft yarn of fabric by the columns and rows of primitive figure.Divide primitive grid, its number equals the fabric unit quantity along warp thread or weft direction.From fabric color hum pattern, we can, by the colouring information of fabric is carried out to binary coding, can obtain interlacing point type.In ideal situation, the intensity at edge has two states: 0 and 1.Intertwined point can be divided into 8 types by the various combination of four limit intensity.Here, no matter be warp thread or the intertwined point of weft yarn, can Further Division be all four subtypes.But under actual conditions, the intensity at edge is not absolute 1 or 0, but relative high or low.The classification of intertwined point is a typical K type partition problem, and it has four input data and eight output states.
In fact the weave pattern information that, only fabric model comprises is difficult to explain the classification of real fabric.Every kind of fabric sample also has independent attribute, as actual geometric configuration and the colouring information etc. of yarn.There is identical weave pattern, but the different sample of its attribute still can obtain different classification results.In fabric classification process, be necessary the independent attribute of these primitives to be dissolved in fabric classification model.For example, primitive sorting algorithm can incorporate the attribute parameter information of each primitive itself.In order to set up complete mapping relations between fabric sample and disaggregated model, we adopt coding form to expand primitive set of graphs to one to have detail and repeatably fabric unit model of complete structure.Wherein the disaggregated model of fabric adopts a coding schedule to represent, wherein primitive information of each byte representation.Its structure as shown in Figure 5.
On the basis of primitive figure, fabric classification model is proposed, to describe the individual attribute of weave pattern and fabric simultaneously.Take primitive as the 2D of basic research unit network mode, the perfect disaggregated model of describing fabric of energy, wherein, the arrangement mode of primitive is described the weave pattern of fabric, and the parameter of primitive itself can be described its individual attribute.Because the yarn of braiding cloth has limited color category, the present invention proposes similar Bitmap Structure mode, represents in mode in message structure, and we have distributed 4 bits, represent altogether 16 kinds of colors.Each disaggregated model not only has a sorting code number table, also needs the data structure of an another attached palette.According to colouring information code and palette information in structural table, just can determine the color of primitive.
The shape information of yarn depends on the geological information of relevant primitive.By to primitive rasterizing, the length and width Information Availability of its grid represents the shape of different yarns, thereby obtains the shape information of primitive.
In addition, fabric same position, positive and negative primitive is not identical, still, due to the regularity of weaving textile, utilizes the two-sided information of the two-sided primitive of its same position, can realize modelling verification and the correction of automatic fabric weaving style.
Obtain the colouring information of fabric two-face image, by being encoded, colouring information obtains the color pattern of fabric, colouring information is carried out secondary coding and is obtained the intertwined point type of fabric, utilize the intertwined point type that colouring information obtains to carry out error correction and perfect to the preliminary weave pattern of fabric, finally obtain accurate weave pattern, this weave pattern has comprised fabric color pattern and intertwined point type accurately, and with this, structure of fabric is carried out to modeling, obtain simple, directly perceived, clear and definite fabric construction pattern.And by modeling and database manipulation, realize the output of high accuracy fabric model, and then can instead push away the real weave pattern of fabric, the fabric of particular types can be checked fast and be evaluated.If Fig. 6 is the fabric patterns model of setting up.
In implementation process, adopt following equipment:
(1) CCD imaging system, CCD model is HV1303UM, major parameter: the CMOS CCD that resolution is 1280*1024; Optical dimensions is 1/1.8 "; The highest 1,300,000 pixels; The highest horizontal resolution is 1280; Digital-to-analogue conversion precision is 10bit; The high s/n ratio (AGC OFF) of 45dB; Can open/close automatic gain control (AGC), digital gain multiple is: * 2, * 1, * 0.5, * 0.25; For the light source of 550nm, its sensitivity is: 2.1V/Lux-s; Can be from the black-to-white level weighing apparatus correcting mode of motion tracking (ATW)/manually setting; Can accept two kinds of power supply supplies of 24V AC and 12V DC.
(2) computer system Lenovo Qitian M7300:CPU, Intel Duo i3 550, double-core/tetra-thread, frequency 3.2GHz L2 cache 2 × 256KB, three grades of buffer memory 4MB; Board chip set Intel H57, mainboard memory size is 2GB; Type of memory is DDR31333; Hard-disk capacity 5006B, rotating speed 7200 turns, SATAII; Video card type is solely aobvious AMD Radeon HD4350, video memory capacity 512MB, and display sizes is 17 inches.
System is as follows to the analytic process of fabric: will be detected fabric and pack fixture into, and according to the imaging pattern of prior setting, get the positive and negative dual-side image of fabric.Image is carried out after affine transformation rectification, to experimental image shear, after the preliminary treatment such as medium filtering, by image is carried out to Gray Projection, tentatively realize image lattice, at employing colouring information and range information, grid is adjusted, realize the precision that grid is divided, finally obtain the minimal structure unit-primitive of fabric.Extract the colouring information of fabric primitive by the method for color cluster, then primitive marginal information is analyzed, obtain the intertwined point type of fabric.Utilize scan mode to carry out color and intertwined point class coding to primitive and obtain the model of fabric that comprises the features such as intertwined point type and colouring information, model data information is imported in database, analyze and evaluate for weaving quality and performance to fabric.
Claims (3)
1. an automatic fabric weaving style automatic analysis method, is characterized in that, comprises the following steps:
A1, adopt 2 imaging devices to carry out image acquisition to region corresponding to fabric positive and negative, by Computer Image Processing device, the fabric two-face image obtaining is comprised the preliminary treatment to two-sidedly positioning of obtaining image, registration, removal noise processed and shear treatment of the location that utilizes fixture;
A2, utilize the brightness information in image warp, latitude direction, textile image is carried out to grid to be divided automatically, by analyzing the colouring information of the grid image divided correspondence position aspect positive and negative two, determine the alternate position of yarn, thereby determine node location, isolate the structural motif of fabric;
A3, adopts automatic scan mode to carry out color set division to the primitive of grid, then determines representative color aggregate information, makes primitive distribute the most rational color;
A4, analyzes the primitive of the basic cycling element of fabric, and primitive is carried out to color and structural information coding, and sets up a primitive storage organization message file that comprises the information such as node type and color, for reconstruct fabric patterns structural model;
A5, the establishment styles store structural information file of setting up fabric deposits database in, for the quality of fabric is evaluated.
2. automatic fabric weaving style automatic analysis method as claimed in claim 1, is characterized in that, the method that realizes primitive location is, utilize relevant and brightness information, it is to utilize projection algorithm that textile image is carried out to automatic division of grid, and its gray value is mapped to two independently one dimension waveforms
G in formula
k(j) be the gray value of k two field picture j row, G
k(i, j) is the grey scale pixel value of (i, j) position on k two field picture, and G
k(i) be the capable gray value of k two field picture i, utilize drop shadow curve's ripple kurtosis to realize after the Preliminary division of grid, recycling is relevant to be adjusted dividing with monochrome information etc., realizes accurate gridding.
3. automatic fabric weaving style automatic analysis method as claimed in claim 1, is characterized in that, the primitive of fabric patterns model adopts domain to represent, its structure comprises status field, color gamut, intertwined point domain information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110415317.8A CN102523366B (en) | 2011-12-13 | 2011-12-13 | Automatic fabric weaving style analyzing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110415317.8A CN102523366B (en) | 2011-12-13 | 2011-12-13 | Automatic fabric weaving style analyzing system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102523366A CN102523366A (en) | 2012-06-27 |
| CN102523366B true CN102523366B (en) | 2014-07-09 |
Family
ID=46294151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110415317.8A Expired - Fee Related CN102523366B (en) | 2011-12-13 | 2011-12-13 | Automatic fabric weaving style analyzing system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102523366B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103176420B (en) * | 2013-03-26 | 2015-06-03 | 东华大学 | Physical yarn woven-pattern digital modeling device and method |
| CN103674952B (en) * | 2013-12-09 | 2017-04-19 | 上海工程技术大学 | Bilateral texture and color acquisition and analysis method for circular slice sample |
| CN106198542B (en) * | 2016-07-05 | 2017-10-20 | 江南大学 | A kind of knitted fabric industrial analysis method based on smart mobile phone |
| CN106485288B (en) * | 2016-12-21 | 2023-11-28 | 上海工程技术大学 | Automatic identification method for colored fabric tissue |
| CN109993755B (en) * | 2019-04-02 | 2021-01-08 | 浙江大学 | Jacquard fabric image weave structure segmentation method |
| CN111177810B (en) | 2019-12-31 | 2021-11-09 | 南京玻璃纤维研究设计院有限公司 | Method and device for generating texture pattern of preform, electronic device and storage medium |
| CN111177809B (en) * | 2019-12-31 | 2021-09-21 | 南京玻璃纤维研究设计院有限公司 | Texture map generation method and device, electronic equipment and readable storage medium |
| CN112070723B (en) * | 2020-08-14 | 2023-11-28 | 盐城工业职业技术学院 | Automatic identification method for plain woven fabric density |
| CN111932686B (en) * | 2020-09-09 | 2021-01-01 | 南昌虚拟现实研究院股份有限公司 | Mapping relation determining method and device, readable storage medium and computer equipment |
| CN117593285A (en) * | 2023-12-14 | 2024-02-23 | 江苏恒兆电缆有限公司 | Quality detection system and method for flexible mineral insulation flexible fireproof cable |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1534294A (en) * | 2002-06-06 | 2004-10-06 | 香港理工大学 | Textile surface analysis method and its system |
| CN1712888A (en) * | 2004-06-23 | 2005-12-28 | 香港理工大学 | Reconstruction system and method for sheet three-dimensional surface of flexible body |
| CN1844550A (en) * | 2006-01-26 | 2006-10-11 | 香港理工大学 | Textile and yarn analysis system based on two-side scanning technology |
| CN102393178A (en) * | 2011-06-21 | 2012-03-28 | 上海工程技术大学 | Digital imaging and analyzing device for textures and colors of surfaces on both sides of sheet material |
-
2011
- 2011-12-13 CN CN201110415317.8A patent/CN102523366B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1534294A (en) * | 2002-06-06 | 2004-10-06 | 香港理工大学 | Textile surface analysis method and its system |
| CN1712888A (en) * | 2004-06-23 | 2005-12-28 | 香港理工大学 | Reconstruction system and method for sheet three-dimensional surface of flexible body |
| CN1844550A (en) * | 2006-01-26 | 2006-10-11 | 香港理工大学 | Textile and yarn analysis system based on two-side scanning technology |
| CN102393178A (en) * | 2011-06-21 | 2012-03-28 | 上海工程技术大学 | Digital imaging and analyzing device for textures and colors of surfaces on both sides of sheet material |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102523366A (en) | 2012-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102523366B (en) | Automatic fabric weaving style analyzing system and method | |
| CN1844550B (en) | Textile and yarn analysis system based on two-side scanning technology | |
| Siew et al. | Texture measures for carpet wear assessment | |
| CN103759662B (en) | A kind of textile yarn diameter dynamic rapid measurement device and method | |
| CN103176420B (en) | Physical yarn woven-pattern digital modeling device and method | |
| CN106485288B (en) | Automatic identification method for colored fabric tissue | |
| CN109977886A (en) | Shelf vacancy rate calculation method and device, electronic equipment, storage medium | |
| Zhang et al. | A review of fabric identification based on image analysis technology | |
| CN103196917A (en) | CCD linear array camera-based online rolled sheet material surface flaw detection system and detection method thereof | |
| US20110148897A1 (en) | Apparatus and methods for processing images | |
| CN106023098B (en) | Image mending method based on the more dictionary learnings of tensor structure and sparse coding | |
| Dror et al. | Statistics of real-world illumination | |
| CN109035196A (en) | Image local fuzzy detection method based on conspicuousness | |
| CN109211918B (en) | Fabric bow weft detection method based on weft trend | |
| EP2269163A2 (en) | System and method for illumination invariant image segmentation | |
| CN109376787A (en) | Manifold learning network and computer visual image collection classification method based on it | |
| CN103471974B (en) | A kind of image method measures the method for fabric theoretic porosity | |
| Zhang et al. | Automatic inspection of yarn-dyed fabric density by mathematical statistics of sub-images | |
| Jia et al. | Fabric defect inspection based on lattice segmentation and lattice templates | |
| CN109594319A (en) | A kind of pck count intelligent detection device and method | |
| CN104899592B (en) | A kind of road semiautomatic extraction method and system based on circular shuttering | |
| CN106875459A (en) | A kind of colored figured texture weave structure adaptive equalization methods based on image segmentation | |
| CN103778431B (en) | Medical image characteristic extracting and identifying method based on two-directional grid complexity measurement | |
| CN111161228B (en) | Button surface defect detection method based on transfer learning | |
| Sahasrabudhe et al. | Structured spatial domain image and data comparison metrics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140709 Termination date: 20191213 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |