CN118329808B - Intelligent identification method and system for textile fibers based on spectrum data - Google Patents
Intelligent identification method and system for textile fibers based on spectrum data Download PDFInfo
- Publication number
- CN118329808B CN118329808B CN202410772935.5A CN202410772935A CN118329808B CN 118329808 B CN118329808 B CN 118329808B CN 202410772935 A CN202410772935 A CN 202410772935A CN 118329808 B CN118329808 B CN 118329808B
- Authority
- CN
- China
- Prior art keywords
- textile fiber
- fiber
- value
- crystallinity
- textile
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Treatment Of Fiber Materials (AREA)
Abstract
The invention discloses an intelligent identification method and system for textile fibers based on spectrum data, and relates to the field of textile fiber identification, wherein the identification method comprises the following steps: performing transformation analysis on the textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelengths based on spectrum characteristics and coefficient correlation techniques, and establishing a prediction model to output the crystallinity of the textile fiber; establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison; performing equalization and segmentation treatment on the textile fiber image, and mapping pixel points of the treated textile fiber image to a state space to evaluate the toughness of the textile fiber; and dynamically analyzing the application state of the textile fiber to evaluate the service life of the textile fiber. The method can obtain the predictive estimation value about the strength and the structure of the fiber, and provide data support for the evaluation of the toughness of the fiber.
Description
Technical Field
The invention relates to the field of textile fiber identification, in particular to an intelligent textile fiber identification method and system based on spectrum data.
Background
Textile fibers are raw materials for spinning and making fabrics, clothing and other textile products, which can be divided into two main categories depending on the source and the manner of manufacture: natural fibers and chemical fibers, wherein the natural fibers are directly derived from animals and plants, including cotton, wool, silk, hemp and the like, and the chemical fibers are fibers synthesized through a chemical processing process, including synthetic fibers, regenerated fibers, inorganic fibers and the like.
Meanwhile, the textile fiber has various characteristics of hygroscopicity, toughness, air permeability and the like, so that in the manufacture and application of textiles, the selection of proper textile fiber is important according to different requirements and purposes, and in the textile industry, the identification and performance evaluation of the textile fiber are also necessary operations for ensuring the quality and the functionality of the textiles.
However, the textile fiber identification method in the prior art cannot effectively evaluate the toughness and the wear resistance of the fiber at the same time, lacks the capability of comprehensively analyzing the microstructure and the macroscopic performance of the fiber, cannot identify the textile fiber by utilizing the advanced analysis technology, and has problems, thus limiting the capability of understanding the fiber characteristics.
Disclosure of Invention
In order to achieve the above purpose, the present invention adopts the following technical scheme:
In a first aspect, the present invention provides an intelligent identification method for textile fibers based on spectral data, the intelligent identification method for textile fibers comprising the steps of:
S1, performing transformation analysis on textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelengths based on spectrum characteristics and coefficient correlation techniques, and establishing a prediction model to output the crystallinity of the textile fiber;
s2, establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison;
s3, performing equalization and segmentation treatment on the textile fiber image, and mapping pixel points of the treated textile fiber image to a state space to evaluate the strength and toughness of the textile fiber;
s4, dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result, and evaluating the service life of the textile fiber according to the application state;
performing equalization and segmentation processing on the textile fiber image, mapping pixels of the processed textile fiber image to a state space, and evaluating the toughness of the textile fiber, wherein the method comprises the following steps of:
s31, performing image denoising processing based on the acquired textile fiber image, and performing enhancement processing on the textile fiber image by utilizing a histogram equalization technology;
S32, separating a fiber region from a background region in the textile fiber image based on an image segmentation technology, and mapping pixel points corresponding to the fiber region to a state space to obtain a prediction estimation value;
S33, defining a matching rule of geometric features and point continuity based on a continuity standard, and generating continuity of a rough matching stage and a fine matching stage verification prediction estimated value based on the matching rule;
s34, judging whether the fiber points in the textile fibers have fracture conditions according to the continuity verification result, if the fracture conditions exist, defining that the toughness of the textile fibers is unqualified, and if the fracture conditions do not exist, defining that the toughness is qualified.
Preferably, the method for performing transformation analysis processing on the textile fiber spectrum data to obtain spectrum characteristics, determining the wavelength of the spectrum characteristics based on the spectrum characteristics and coefficient correlation technology, and establishing a prediction model to analyze the crystallinity of the textile fiber comprises the following steps:
S11, performing slope transformation and continuous system elimination operation based on textile fiber spectrum data to obtain a variation spectrum to realize spectrum conversion, and obtaining spectrum characteristics based on a spectrum conversion result;
S12, performing correlation analysis on textile fiber spectral data, variation spectrums and corresponding morphological characteristics by using a coefficient correlation technology, and analyzing correlation coefficients between the spectral characteristics and absorption peaks;
s13, traversing all the correlation coefficients, selecting the wavelength corresponding to the correlation coefficient equal to the preset value as the spectral characteristic wavelength, and judging the morphological characteristics of the textile fiber based on the spectral characteristic wavelength;
S14, taking the characteristic wavelength as an independent variable, taking the morphological characteristics of the textile fiber as the independent variable, establishing a prediction model to output the peak number of the spectrum of the textile fiber, and analyzing the crystallinity of the textile fiber based on the peak number.
Preferably, the method for determining the distribution condition of the crystallinity of the textile fiber to generate the morphological characteristic value based on the mapping space established by the data analysis technology and evaluating the wear resistance of the textile fiber by similarity comparison comprises the following steps:
S21, preprocessing the crystallinity of the textile fiber based on a data analysis technology, constructing a mapping space by combining a multivariate analysis technology, and analyzing the change frequency of the crystallinity of the textile fiber in the mapping space;
s22, adopting characteristic analysis of the change frequency of the crystallinity of the textile fiber and the distribution change of the design space to obtain a distribution rule of the crystallinity change, and taking the crystallinity corresponding to the upper limit value and the lower limit value of the result of the distribution rule as a comparison value;
s23, analyzing morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result;
And S24, evaluating the wear resistance of the textile fiber based on the similarity result matched with a preset wear resistance grading standard.
Preferably, analyzing the morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result, wherein the method comprises the following steps of:
S231, judging the size of a crystallization area and the size of crystals in the textile fiber based on the crystallinity corresponding to the comparison value, and analyzing the cross-sectional shape of the textile fiber by combining a projection scanning technology;
s232, converting the cross section shape into a single-order data surface by adopting a weighted average method to form a histogram, defining the morphological characteristics of the textile fiber, and constructing a morphological characteristic matrix by combining a morphological analysis method;
s233, respectively solving corresponding density indexes based on morphological characteristics and morphological characteristic matrixes of standard thresholds, and selecting morphological characteristics corresponding to the maximum value of the density indexes as central characteristics;
s234, judging the variable weight Euclidean distance between the central features corresponding to the standard threshold and the comparison value, and solving the similarity between the central features based on the variable weight Euclidean distance;
s235, selecting the minimum similarity value corresponding to the comparison value as a similarity result between the standard threshold and the standard threshold.
Preferably, the method of converting the cross-sectional shape into a single-order data surface by a weighted average method to form a histogram, defining the morphological characteristics of the textile fiber, and constructing a morphological characteristic matrix by combining a morphological analysis method comprises the following steps:
S2321, converting the cross section shape corresponding to the morphological feature matrix into a single-order data surface by adopting a weighted average method, obtaining shape points in the single-order data surface to form a representation graph, and judging the distribution uniformity of the representation graph;
S2322, judging the mean value coefficient and the distribution symmetry degree of the cross-section shape points based on the representation diagram, analyzing the discrete degree of the mean value coefficient of the cross-section shape points, and reflecting the kurtosis of the cross-section shape points according to the discrete degree to obtain the distribution uniformity;
s2323, defining the distribution uniformity, the mean value coefficient of the cross-section shape points, the distribution symmetry degree, the discrete degree and the distribution uniformity of the representation graph as characteristic parameters of the cross-section shape points to obtain morphological characteristics of the textile fibers;
S2324, performing parameter dimension reduction operation on the morphological characteristics by using a morphological analysis method, projecting the dimension-reduced morphological characteristics to a low-dimensional space, and performing correlation elimination processing by using characteristic parameters of cross-section shape points to construct a morphological characteristic matrix.
Preferably, the mean coefficient discrete degree of the cross-sectional shape point is calculated as:
；
Wherein L represents the degree of dispersion of the mean value coefficient, T represents the number of cross-sectional shape points, E represents the distribution uniformity level in the graph, P represents the mean value coefficient of the cross-sectional shape points, and m (E) represents the number of cross-sectional shape points corresponding to the graph satisfying the distribution uniformity level.
Preferably, the mean coefficient of the cross-sectional shape point is calculated as:
；
Wherein P represents the mean value coefficient of the cross-sectional shape points, T represents the number of the cross-sectional shape points, E represents the distribution uniformity level in the graph, and m (E) represents the number of the cross-sectional shape points corresponding to the graph satisfying the distribution uniformity level.
Preferably, the method for separating the fiber region from the background region in the textile fiber image based on the image segmentation technology, and mapping the pixel point corresponding to the fiber region to the state space to obtain the prediction estimation value comprises the following steps:
s321, separating a fiber region from a background region in a textile fiber image based on a region growing mode selected by an image segmentation technology, and extracting fiber length characteristics from the segmented fiber region;
s322, setting a prediction estimated value attribute definition state space based on the fiber length characteristics, and mapping the pixel points corresponding to the fiber length characteristics to the state space;
S323, carrying out sectional mapping of brightness information on the fiber region in a state space to obtain a predicted estimated value of each region point, and optimizing the estimated value of the region point by adopting a guide filtering mode to obtain a final predicted estimated value.
Preferably, defining a matching rule of geometric features and point continuity based on a continuity criterion, and generating a coarse matching stage and a fine matching stage based on the matching rule to verify continuity of predicted estimation values includes the steps of:
S331, selecting two groups of regional point sets according to a continuity standard, decomposing the regional point sets into discrete sets formed by break points, and judging the distance between the break points in the discrete sets by using a short-side distance algorithm to obtain a distance set;
S332, selecting the median of the distance set as a matching value, and judging the difference value between the median of the distance set of the predicted estimated value and the matching value;
S333, defining that the difference value is larger than or smaller than a set value, and indicating that the geometric features are not matched, and defining that the difference value is equal to the set value, indicating that the geometric features are matched;
S334, defining an edge set and an auxiliary edge set according to break points in the two selected discrete sets, and judging the continuity of the break points by utilizing the combination of the edge set and the auxiliary edge to generate a point continuity matching rule;
S335, generating a coarse matching stage and a fine matching stage based on the geometric feature matching rule and the point continuity matching rule, and verifying the continuity of the predicted estimated value by using the matching rule.
In a second aspect, the invention also provides a fiber intelligent identification system for textile based on spectrum data, which comprises a fiber crystallinity prediction module, an abrasion resistance evaluation module, a strength and toughness evaluation module and an identification platform generation module, wherein the fiber crystallinity prediction module, the abrasion resistance evaluation module, the strength and toughness evaluation module and the identification platform generation module are sequentially connected;
the fiber crystallinity prediction module is used for performing transformation analysis processing on the textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelength based on the spectrum characteristics and coefficient correlation technique, and establishing a prediction model to output the textile fiber crystallinity;
The wear resistance evaluation module is used for establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison;
The obdurability evaluation module is used for carrying out equalization and segmentation processing on the textile fiber image, and mapping the pixel points of the processed textile fiber image to a state space to evaluate the obdurability of the textile fiber;
The identification platform generation module is used for dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result and evaluating the service life of the textile fiber according to the application state.
The beneficial effects of the invention are as follows:
1. according to the invention, the crystallinity information of the fiber is extracted based on the spectral characteristic wavelength, and the distribution condition and the wear resistance of the crystallinity of the fiber are judged according to the similarity comparison, so that the quality of the fiber is estimated more accurately, and meanwhile, the image processing technology is applied to help to process and analyze the toughness of the textile fiber rapidly, so that the aim of comprehensively estimating the quality and the service life of the fiber from multiple angles by considering the crystallinity, the wear resistance and the toughness of the fiber is realized.
2. The invention effectively extracts the characteristics of the fiber crystallinity based on the data preprocessing and the multivariate analysis technology, constructs the mapping space, can accurately analyze the change frequency of the fiber crystallinity, takes the crystallinity corresponding to the upper limit value and the lower limit value of the change frequency as a comparison value, is convenient for the subsequent morphological characteristic analysis and similarity comparison, and realizes the intelligent assessment of the fiber crystallinity and the wear resistance.
3. The method accurately separates the fiber area of the textile fiber image from the background area based on the image segmentation technology, ensures that the subsequent treatment is only carried out on the fiber area, and can acquire the prediction estimated value about the fiber strength and the structure by mapping the pixel points of the fiber area to the state space, thereby providing data support for the evaluation of the fiber toughness.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a flowchart of a method for intelligently identifying textile fibers based on spectral data according to an embodiment of the invention;
Fig. 2 is a schematic block diagram of a textile fiber intelligent authentication system based on spectral data according to an embodiment of the present invention.
In the figure:
1. A fiber crystallinity prediction module; 2. a wear resistance evaluation module; 3. a strength and toughness evaluation module; 4. and an authentication platform generation module.
Detailed Description
The invention is further described below with reference to the drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the invention. 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.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms also are intended to include the plural forms, and furthermore, it is to be understood that the terms "comprises" and "comprising" and any variations thereof are intended to cover non-exclusive inclusions, such as, for example, processes, methods, systems, products or devices that comprise a series of steps or units, are not necessarily limited to those steps or units that are expressly listed, but may include other steps or units that are not expressly listed or inherent to such processes, methods, products or devices.
Embodiments of the invention and features of the embodiments may be combined with each other without conflict.
Referring to fig. 1, the invention provides an intelligent identification method for textile fibers based on spectrum data, which comprises the following steps:
s1, performing transformation analysis on textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelength based on spectrum characteristics and coefficient correlation technology, and establishing a prediction model to output the crystallinity of the textile fiber.
In this embodiment, performing transformation analysis processing on textile fiber spectral data to obtain spectral features, determining spectral feature wavelengths based on a correlation technique of the spectral features and coefficients, and establishing a predictive model to analyze the crystallinity of the textile fiber includes the steps of:
s11, performing slope transformation and continuous system elimination operation based on textile fiber spectrum data to obtain a variation spectrum to realize spectrum conversion, and obtaining spectrum characteristics based on a spectrum conversion result.
Specifically, the method for implementing slope transformation and continuous system elimination operation based on textile fiber spectrum data to obtain a variation spectrum to realize spectrum transformation, and obtaining spectrum characteristics based on spectrum transformation results comprises the following steps:
the raw spectral data of the textile fiber is collected using a spectrometer to spectrally scan a fiber sample under a controlled environment, applying mathematical transformations to process the spectral data to highlight trends or patterns in the data, and eliminating background signals or continuum portions from the spectral data.
And analyzing the change spectrum after slope transformation and continuous system elimination, identifying key spectral features such as peak value, peak area, width and the like, performing further mathematical and statistical analysis such as normalization and standardization treatment on the identified spectral features, facilitating subsequent data comparison and analysis, and establishing a feature database according to the acquired spectral features for future fiber identification and classification work.
S12, performing correlation analysis on the textile fiber spectral data, the variation spectrum and the corresponding morphological characteristics by using a coefficient correlation technology, and analyzing the correlation coefficient between the spectral characteristics and the absorption peaks.
Specifically, the correlation analysis is performed on the textile fiber spectrum data, the variation spectrum and the corresponding morphological characteristics by using the coefficient correlation technology, and the analysis of the correlation coefficient between the spectrum characteristics and the absorption peaks comprises the following steps:
feature extraction is performed on the spectral data and the change spectral data based on a feature database, which generally comprises extracting spectral features such as absorption peak positions, peak heights, peak areas and the like, correlating the features with morphological feature data, and calculating correlation coefficients between the spectral features and the morphological features, including pearson correlation coefficients, spearman correlation coefficients and the like, so as to measure the linear correlation degree or the correlation direction between two variables.
S13, traversing all the correlation coefficients, selecting the wavelength corresponding to the correlation coefficient equal to the preset value as the spectral characteristic wavelength, and judging the morphological characteristics of the textile fiber based on the spectral characteristic wavelength.
Specifically, traversing all the correlation coefficients, selecting a wavelength corresponding to a correlation coefficient equal to a preset value as a spectrum characteristic wavelength, and judging the morphological characteristics of the textile fiber based on the spectrum characteristic wavelength, wherein the method comprises the following steps:
Traversing all the correlation coefficients, selecting the wavelength corresponding to the correlation coefficient equal to the preset value as the spectral characteristic wavelength, and performing programming to traverse all the wavelengths and calculating the correlation coefficient between the wavelength and the morphological characteristic.
And determining spectral characteristic wavelengths according to the wavelength of which the correlation coefficient obtained by screening is equal to a preset value, wherein the wavelengths are considered to have obvious correlation with morphological characteristics, so that the wavelength selected as the spectral characteristic wavelength is used for judging the morphological characteristics of the textile fiber based on the selected spectral characteristic wavelength, and the morphological characteristics of the textile fiber can be deduced by analyzing the reflection of spectral data at the spectral characteristic wavelength.
S14, taking the characteristic wavelength as an independent variable, taking the morphological characteristics of the textile fiber as the independent variable, establishing a prediction model to output the peak number of the spectrum of the textile fiber, and analyzing the crystallinity of the textile fiber based on the peak number.
Specifically, taking characteristic wavelength as an independent variable, taking morphological characteristics of textile fibers as the independent variable, establishing a prediction model to output the peak number of a textile fiber spectrum, and analyzing the crystallinity of the textile fibers based on the peak number, wherein the method comprises the following steps of:
the method uses machine learning or statistical modeling method, takes characteristic wavelength as independent variable, takes morphological characteristics (such as the number of absorption peaks) of textile fiber as dependent variable, and establishes a prediction model, which can include linear regression, support vector machine, neural network and the like.
Model training is carried out by utilizing the existing textile fiber spectrum data set, model parameters are optimized by fitting the relation between characteristic wavelength and wave crest number, so that the prediction of wave crest number is realized, and the trained model is verified by using the other textile fiber spectrum data set, so that the prediction performance of the model is estimated.
The method comprises the steps of predicting new textile fiber spectrum data by using a trained prediction model to obtain the corresponding peak number, automatically realizing rapid assessment of fiber crystallinity, analyzing the correlation between the predicted peak number and the crystallinity of the textile fiber according to the relationship between the peak number and the crystallinity, assessing the crystallinity of the textile fiber, formulating corresponding standards or thresholds, and judging the crystallinity level of the fiber according to the prediction result of the peak number.
S2, establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber so as to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison.
In this embodiment, the method for determining the distribution condition of the crystallinity of the textile fiber to generate the morphological feature value based on the mapping space established by the data analysis technology, and evaluating the abrasion resistance of the textile fiber by similarity comparison includes the following steps:
S21, preprocessing the crystallinity of the textile fiber based on a data analysis technology, constructing a mapping space by combining a multivariate analysis technology, and analyzing the change frequency of the crystallinity of the textile fiber in the mapping space.
Specifically, the method for preprocessing the crystallinity of the textile fiber based on the data analysis technology, constructing a mapping space by combining the multivariate analysis technology, and analyzing the change frequency of the crystallinity of the textile fiber in the mapping space comprises the following steps:
And collecting the crystallinity related data of the textile fiber, cleaning, screening and formatting the collected data, applying multi-element analysis technologies such as Principal Component Analysis (PCA), factor analysis and the like, processing the preprocessed data to identify main variables and modes in the data, and reducing the data dimension.
According to the result of the multivariate analysis, a mapping space is constructed, in which each dimension represents a main influencing factor or variable, in which space the crystallinity data of the textile fiber can be effectively represented and analyzed, and in which mapping space the frequency of change of the crystallinity of the textile fiber is analyzed, in particular to calculate the frequency of occurrence of different crystallinity values and the trend of these frequencies over the condition or time.
S22, adopting characteristic analysis of the change frequency of the crystallinity of the textile fiber and the distribution change of the design space to obtain a distribution rule of the crystallinity change, and taking the crystallinity corresponding to the upper limit value and the lower limit value of the result based on the distribution rule as a comparison value.
Specifically, by adopting characteristic analysis of variation frequency of crystallinity of textile fibers and distribution variation of design space, a distribution rule of crystallinity variation is obtained, and crystallinity corresponding to upper and lower limit values of a result based on the distribution rule is used as a comparison value, and the method comprises the following steps:
The method comprises the steps of analyzing overall distribution characteristics of crystallinity by using a multivariate statistical technique including Kernel Density Estimation (KDE) or other distribution fitting methods, mapping crystallinity data to a design space, and particularly using advanced data visualization techniques such as a scatter diagram matrix, a three-dimensional graph and the like to represent the spatial distribution characteristics of the crystallinity.
Analyzing the distribution characteristics of the crystallinity data, determining upper limit and lower limit values of the distribution, wherein the values represent extreme conditions of crystallinity change in a design space, and selecting the crystallinity corresponding to the two extreme values as a comparison value according to the upper limit and the lower limit value of a distribution rule, wherein the comparison value can be used as a reference standard of future samples or used as a reference for evaluating and optimizing the textile fiber production process.
S23, analyzing morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result.
Specifically, analyzing the morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result, wherein the method comprises the following steps of:
S231, judging the size of a crystallization area and the size of crystals in the textile fiber based on the crystallinity corresponding to the comparison value, and analyzing the cross-sectional shape of the textile fiber by combining a projection scanning technology;
S232, converting the cross section shape into a single-order data surface shape histogram by adopting a weighted average method, defining the morphological characteristics of the textile fiber, and constructing a morphological characteristic matrix by combining a morphological analysis method.
The method comprises the following steps of converting a cross section shape into a single-order data surface by a weighted average method, defining textile fiber morphological characteristics, and constructing a morphological characteristic matrix by combining a morphological analysis method:
S2321, converting the cross section shape corresponding to the morphological feature matrix into a single-order data surface by adopting a weighted average method, obtaining shape points in the single-order data surface to form a representation graph, and judging the distribution uniformity of the representation graph;
S2322, judging the mean value coefficient and the distribution symmetry degree of the cross-section shape points based on the representation diagram, analyzing the discrete degree of the mean value coefficient of the cross-section shape points, and reflecting the kurtosis of the cross-section shape points according to the discrete degree to obtain the distribution uniformity;
s2323, defining the distribution uniformity, the mean value coefficient of the cross-section shape points, the distribution symmetry degree, the discrete degree and the distribution uniformity of the representation graph as characteristic parameters of the cross-section shape points to obtain morphological characteristics of the textile fibers;
S2324, performing parameter dimension reduction operation on the morphological characteristics by using a morphological analysis method, projecting the dimension-reduced morphological characteristics to a low-dimensional space, and performing correlation elimination processing by using characteristic parameters of cross-section shape points to construct a morphological characteristic matrix.
Wherein, the calculation formula of the mean value coefficient discrete degree of the cross section shape point is:
；
The calculation formula of the mean value coefficient of the cross-section shape point is as follows:
；
Wherein L represents the degree of dispersion of the mean value coefficient, T represents the number of cross-sectional shape points, E represents the distribution uniformity level in the graph, P represents the mean value coefficient of the cross-sectional shape points, and m (E) represents the number of cross-sectional shape points corresponding to the graph satisfying the distribution uniformity level.
S233, respectively solving corresponding density indexes based on morphological characteristics and morphological characteristic matrixes of standard thresholds, and selecting morphological characteristics corresponding to the maximum value of the density indexes as central characteristics;
s234, judging the variable weight Euclidean distance between the central features corresponding to the standard threshold and the comparison value, and solving the similarity between the central features based on the variable weight Euclidean distance;
s235, selecting the minimum similarity value corresponding to the comparison value as a similarity result between the standard threshold and the standard threshold.
And S24, evaluating the wear resistance of the textile fiber based on the similarity result matched with a preset wear resistance grading standard.
Specifically, the method for evaluating the wear resistance of the textile fiber based on the matching of the similarity result and the preset wear resistance grading standard comprises the following steps:
The method comprises the steps of collecting related abrasion resistance data of textile fibers, including historical abrasion resistance test results, fiber types, use environments and the like, determining or obtaining preset abrasion resistance grade classification standards, wherein the standards are generally classified according to the performances of the textile fibers under specific test conditions, such as defining different abrasion resistance grades in a numerical range of the abrasion resistance test results, and matching the calculated similarity results with the preset abrasion resistance grade standards to evaluate the abrasion resistance of the textile fibers.
In order to facilitate understanding of the technical scheme of the present invention, the following describes the abrasion resistance evaluation of the present invention in practical process in detail.
Step one, collecting and processing spectrum data;
The visible light spectrometer with the resolution of 1nm is adopted to collect the spectrum data of the textile fiber under the conditions that the temperature is controlled at 22 ℃ and the humidity is controlled at 50 percent, the first derivative transformation is used for slope transformation, and the three-order polynomial fitting is adopted to eliminate the baseline so as to realize continuous system elimination.
Extracting spectral characteristics;
the peak heights and peak areas of the absorption peaks at 450nm, 550nm and 650nm were mainly focused on, and a spectral feature database containing various fiber types was established.
Step three, correlation analysis;
Using pearson correlation coefficient to analyze the correlation between the spectral feature and the morphological feature, and presetting the correlation coefficient threshold to be 0.8 (representing strong correlation);
the wavelength with a correlation coefficient of 0.8 or more with the morphological feature was determined assuming 550nm as the critical spectral feature wavelength.
Step four, a crystallinity prediction model;
selecting a linear regression model as a model type, defining 200 samples, wherein each sample contains absorption peak data of spectral characteristic wavelength, setting 50 samples for model verification, and defining the relation between crystallinity and peak number: the crystallinity is improved by 0.5% for every 1 increase in the number of peaks.
Step five, analyzing the crystallinity distribution and evaluating the wear resistance;
(1) Construction and analysis of mapping space: the Principal Component Analysis (PCA) is applied to reduce the data dimension, and a two-dimensional mapping space based on the PCA is constructed to represent the crystallinity change, and the upper limit and the lower limit of the distribution are determined by using a Kernel Density Estimation (KDE) method, which are respectively 95% quantiles and 5% quantiles.
(2) Morphological characteristics and abrasion resistance similarity analysis: and calculating a single-order data surface by using a weighted average method, constructing a morphological feature matrix comprising a mean value coefficient and a distribution symmetry degree, calculating the similarity by using a variable weight Euclidean distance, and obtaining the average similarity of the crystallinity corresponding to the upper limit and the lower limit.
(3) Abrasion resistance rating evaluation: the wear resistance grading is based on standard ASTM D3884 and the similarity result matches the preset wear resistance grading standard.
Therefore, the characteristics of the fiber crystallinity are effectively extracted based on the data preprocessing and the multivariate analysis technology, the change frequency of the fiber crystallinity can be accurately analyzed by constructing a mapping space, and meanwhile, the crystallinity corresponding to the upper limit value and the lower limit value of the change frequency is used as a comparison value, so that the subsequent morphological characteristic analysis and similarity comparison are facilitated, and the intelligent assessment of the fiber crystallinity and the wear resistance is realized.
And S3, performing equalization and segmentation treatment on the textile fiber image, and mapping pixel points of the treated textile fiber image to a state space to evaluate the toughness of the textile fiber.
In this embodiment, performing equalization and segmentation processing on the textile fiber image, and mapping the pixel points of the processed textile fiber image to a state space to evaluate the toughness of the textile fiber comprises the following steps:
S31, performing image denoising processing based on the acquired textile fiber image, and performing enhancement processing on the textile fiber image by using a histogram equalization technology.
Specifically, performing image denoising processing based on the acquired textile fiber image, and performing enhancement processing on the textile fiber image by using a histogram equalization technique includes the steps of:
High resolution camera devices capture textile fibers to obtain high quality textile fiber images, use image processing software or programming libraries (e.g., openCV, MATLAB) to process the images, and apply denoising algorithms, including gaussian blur, median filtering, or bilateral filtering, to remove random noise in the images and preserve edge information.
The histogram equalization is applied to the denoised image by adjusting the histogram of the image, so that the brightness of the image is more uniformly distributed, the visual contrast of the image is enhanced, the processed image is visually checked, and whether the denoising effect and the equalization achieve the expected image quality improvement is verified.
S32, separating a fiber region from a background region in the textile fiber image based on an image segmentation technology, and mapping pixel points corresponding to the fiber region to a state space to obtain a prediction estimated value.
Specifically, the method for separating the fiber region from the background region in the textile fiber image based on the image segmentation technology, and mapping the pixel points corresponding to the fiber region to the state space to obtain the prediction estimation value comprises the following steps:
s321, separating a fiber region from a background region in a textile fiber image based on a region growing mode selected by an image segmentation technology, and extracting fiber length characteristics from the segmented fiber region;
s322, setting a prediction estimated value attribute definition state space based on the fiber length characteristics, and mapping the pixel points corresponding to the fiber length characteristics to the state space;
S323, carrying out sectional mapping of brightness information on the fiber region in a state space to obtain a predicted estimated value of each region point, and optimizing the estimated value of the region point by adopting a guide filtering mode to obtain a final predicted estimated value.
S33, defining a matching rule of geometric features and point continuity based on a continuity standard, and generating continuity of a rough matching stage and a fine matching stage based on the matching rule to verify the continuity of the prediction estimation value.
Specifically, defining a matching rule of geometric features and point continuity based on a continuity standard, and generating continuity of a rough matching stage and a fine matching stage verification prediction estimation value based on the matching rule comprises the following steps:
S331, selecting two groups of regional point sets according to a continuity standard, decomposing the regional point sets into discrete sets formed by break points, and judging the distance between the break points in the discrete sets by using a short-side distance algorithm to obtain a distance set;
S332, selecting the median of the distance set as a matching value, and judging the difference value between the median of the distance set of the predicted estimated value and the matching value;
S333, defining that the difference value is larger than or smaller than a set value, and indicating that the geometric features are not matched, and defining that the difference value is equal to the set value, indicating that the geometric features are matched;
S334, defining an edge set and an auxiliary edge set according to break points in the two selected discrete sets, and judging the continuity of the break points by utilizing the combination of the edge set and the auxiliary edge to generate a point continuity matching rule;
S335, generating a coarse matching stage and a fine matching stage based on the geometric feature matching rule and the point continuity matching rule, and verifying the continuity of the predicted estimated value by using the matching rule.
S34, judging whether the fiber points in the textile fibers have fracture conditions according to the continuity verification result, if the fracture conditions exist, defining that the toughness of the textile fibers is unqualified, and if the fracture conditions do not exist, defining that the toughness is qualified.
In order to facilitate understanding of the technical scheme of the invention, the following describes the evaluation of the toughness of the invention in the actual process in detail.
Step one, image denoising and enhancing treatment;
capturing a textile fiber map under natural illumination by using high-definition camera equipment with resolution of 4096x4096 pixels, applying a Gaussian blur filter, processing by using a core of 5x5 to remove high-frequency noise, and applying histogram equalization to the denoised image to enhance the contrast and detail of the image so as to make the fiber more obvious.
Step two, image segmentation and mapping state space;
Setting an initial seed point threshold as an image gray level mean value, setting an expansion threshold as a gray level value difference not exceeding 10 so as to distinguish fibers from a background, calculating the pixel length of each fiber to extract the fiber length, and defining a state space by taking the fiber length as an X axis and the average brightness of the fibers as a Y axis.
The fiber characteristics are mapped to a defined state space, the strength and toughness level of each region is predicted by using a K-nearest neighbor (KNN) algorithm, smoothing is carried out by adopting a guide filter, and the core size is set to be 10x10.
Step three, continuous analysis and verification;
Selecting two groups of fiber point sets, calculating a distance set of each group of internal points by using Euclidean distance, setting a matching threshold value as a median distance + -2 pixels, comparing whether break points in the two groups of discrete sets are kept continuous by using an edge set method, verifying the connectivity of the points by using a graph theory algorithm, and determining the integrity of the fibers.
Judging the continuity of the fiber, if all fiber points are continuous, judging the toughness of the fiber as qualified, if breakage or large-scale discontinuity is found, marking as unqualified toughness, and comprehensively evaluating the toughness of the fiber according to the result of continuity verification.
Therefore, the fiber area of the textile fiber image is accurately separated from the background area based on the image segmentation technology, the follow-up treatment is ensured to be carried out only on the fiber area, and the predicted estimated value about the fiber strength and the structure can be obtained by mapping the pixel points of the fiber area to the state space, so that data support is provided for the evaluation of the fiber toughness.
S4, dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result, and evaluating the service life of the textile fiber according to the application state.
In this embodiment, dynamically analyzing the application state of the textile fiber based on the evaluation result of the abrasion resistance and the toughness, and evaluating the service life of the textile fiber according to the application state includes the steps of:
the method comprises the steps of analyzing the change trend of wear resistance and toughness data along with time by using statistical software or a data analysis tool, analyzing the correlation of the wear resistance and the toughness, evaluating how the two properties mutually affect the overall performance of the fiber, classifying the fiber into different application state categories according to the test results of the wear resistance and the toughness, and establishing a model to predict the rate and the mode of the performance degradation of the fiber according to the decrease trend of the wear resistance and the toughness.
Based on historical data and performance degradation models, the expected service life of the textile fibers is predicted, and the risks and failure modes of the fibers which can occur under different application states are evaluated.
Referring to fig. 2, the invention further provides a fiber intelligent identification system for textile based on spectral data, which comprises a fiber crystallinity prediction module 1, an abrasion resistance evaluation module 2, a strength and toughness evaluation module 3 and an identification platform generation module 4, wherein the fiber crystallinity prediction module 1, the abrasion resistance evaluation module 2, the strength and toughness evaluation module 3 and the identification platform generation module 4 are sequentially connected;
The fiber crystallinity prediction module 1 is used for performing transformation analysis processing on the textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelengths based on the spectrum characteristics and coefficient correlation techniques, and establishing a prediction model to output the textile fiber crystallinity;
The wear resistance evaluation module 2 is used for establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber so as to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison;
The toughness evaluation module 3 is used for carrying out equalization and segmentation processing on the textile fiber image, and mapping the pixel points of the processed textile fiber image to a state space to evaluate the toughness of the textile fiber;
The identification platform generating module 4 is used for dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result and evaluating the service life of the textile fiber according to the application state.
In summary, by means of the technical scheme, the fiber crystallinity information is extracted based on the spectral characteristic wavelength, and the distribution condition and the wear resistance of the fiber crystallinity are judged according to the similarity comparison, so that the quality of the fiber is estimated more accurately, and meanwhile, the image processing technology is applied to help to process and analyze the toughness of the textile fiber rapidly, so that the aim of comprehensively estimating the quality and the service life of the fiber from multiple angles by considering the fiber crystallinity, the wear resistance and the toughness is fulfilled. The invention effectively extracts the characteristics of the fiber crystallinity based on the data preprocessing and the multivariate analysis technology, constructs the mapping space, can accurately analyze the change frequency of the fiber crystallinity, takes the crystallinity corresponding to the upper limit value and the lower limit value of the change frequency as a comparison value, is convenient for the subsequent morphological characteristic analysis and similarity comparison, and realizes the intelligent assessment of the fiber crystallinity and the wear resistance. The method accurately separates the fiber area of the textile fiber image from the background area based on the image segmentation technology, ensures that the subsequent treatment is only carried out on the fiber area, and can acquire the prediction estimated value about the fiber strength and the structure by mapping the pixel points of the fiber area to the state space, thereby providing data support for the evaluation of the fiber toughness.
Those of ordinary skill in the art will appreciate that the elements of the various examples described in connection with the present embodiments, i.e., the algorithm steps, can be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
While the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it is intended to cover all modifications or variations within the scope of the invention as defined by the claims of the present invention.
Claims (2)
1. The intelligent identification method for the textile fiber based on the spectrum data is characterized by comprising the following steps of:
S1, performing transformation analysis on textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelengths based on spectrum characteristics and coefficient correlation techniques, and establishing a prediction model to output the crystallinity of the textile fiber;
s2, establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison;
s3, performing equalization and segmentation treatment on the textile fiber image, and mapping pixel points of the treated textile fiber image to a state space to evaluate the strength and toughness of the textile fiber;
s4, dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result, and evaluating the service life of the textile fiber according to the application state;
the method for evaluating the toughness of the textile fiber by carrying out equalization and segmentation treatment on the textile fiber image and mapping the pixel points of the treated textile fiber image to a state space comprises the following steps:
s31, performing image denoising processing based on the acquired textile fiber image, and performing enhancement processing on the textile fiber image by utilizing a histogram equalization technology;
S32, separating a fiber region from a background region in the textile fiber image based on an image segmentation technology, and mapping pixel points corresponding to the fiber region to a state space to obtain a prediction estimation value;
S33, defining a matching rule of geometric features and point continuity based on a continuity standard, and generating continuity of a rough matching stage and a fine matching stage verification prediction estimated value based on the matching rule;
s34, judging whether a fiber point in the textile fiber has a fracture condition according to a continuity verification result, if the fiber point has the fracture condition, defining that the toughness of the textile fiber is unqualified, and if the fiber point does not have the fracture condition, defining that the toughness is qualified;
The method for obtaining the spectral characteristics by performing transformation analysis on the textile fiber spectral data, determining the wavelengths of the spectral characteristics based on the spectral characteristics and coefficient correlation technology, and establishing a predictive model to analyze the crystallinity of the textile fiber comprises the following steps:
S11, performing slope transformation and continuous system elimination operation based on textile fiber spectrum data to obtain a variation spectrum to realize spectrum conversion, and obtaining spectrum characteristics based on a spectrum conversion result;
S12, performing correlation analysis on textile fiber spectral data, variation spectrums and corresponding morphological characteristics by using a coefficient correlation technology, and analyzing correlation coefficients between the spectral characteristics and absorption peaks;
s13, traversing all the correlation coefficients, selecting the wavelength corresponding to the correlation coefficient equal to the preset value as the spectral characteristic wavelength, and judging the morphological characteristics of the textile fiber based on the spectral characteristic wavelength;
S14, taking the characteristic wavelength as an independent variable, taking the morphological characteristics of the textile fiber as the independent variable, establishing a prediction model to output the peak number of the spectrum of the textile fiber, and analyzing the crystallinity of the textile fiber based on the peak number;
The method for judging the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value based on the mapping space established by the data analysis technology, and evaluating the wear resistance of the textile fiber by similarity comparison comprises the following steps:
S21, preprocessing the crystallinity of the textile fiber based on a data analysis technology, constructing a mapping space by combining a multivariate analysis technology, and analyzing the change frequency of the crystallinity of the textile fiber in the mapping space;
s22, adopting characteristic analysis of the change frequency of the crystallinity of the textile fiber and the distribution change of the design space to obtain a distribution rule of the crystallinity change, and taking the crystallinity corresponding to the upper limit value and the lower limit value of the result of the distribution rule as a comparison value;
s23, analyzing morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result;
S24, evaluating the wear resistance of the textile fiber based on the similarity result matched with a preset wear resistance grading standard;
Analyzing the morphological characteristics of the textile fiber according to the crystallinity corresponding to the comparison value, and comparing the morphological characteristics with a standard threshold value to obtain a similarity result, wherein the method comprises the following steps of:
S231, judging the size of a crystallization area and the size of crystals in the textile fiber based on the crystallinity corresponding to the comparison value, and analyzing the cross-sectional shape of the textile fiber by combining a projection scanning technology;
s232, converting the cross section shape into a single-order data surface by adopting a weighted average method to form a histogram, defining the morphological characteristics of the textile fiber, and constructing a morphological characteristic matrix by combining a morphological analysis method;
s233, respectively solving corresponding density indexes based on morphological characteristics and morphological characteristic matrixes of standard thresholds, and selecting morphological characteristics corresponding to the maximum value of the density indexes as central characteristics;
s234, judging the variable weight Euclidean distance between the central features corresponding to the standard threshold and the comparison value, and solving the similarity between the central features based on the variable weight Euclidean distance;
s235, selecting a similarity value with the minimum corresponding to the comparison value as a similarity result between the standard threshold and the standard threshold;
The method for converting the cross section shape into a single-order data surface by adopting a weighted average method to form a histogram, defining the morphological characteristics of the textile fiber, and constructing a morphological characteristic matrix by combining a morphological analysis method comprises the following steps of:
S2321, converting the cross section shape corresponding to the morphological feature matrix into a single-order data surface by adopting a weighted average method, obtaining shape points in the single-order data surface to form a representation graph, and judging the distribution uniformity of the representation graph;
S2322, judging the mean value coefficient and the distribution symmetry degree of the cross-section shape points based on the representation diagram, analyzing the discrete degree of the mean value coefficient of the cross-section shape points, and reflecting the kurtosis of the cross-section shape points according to the discrete degree to obtain the distribution uniformity;
s2323, defining the distribution uniformity, the mean value coefficient of the cross-section shape points, the distribution symmetry degree, the discrete degree and the distribution uniformity of the representation graph as characteristic parameters of the cross-section shape points to obtain morphological characteristics of the textile fibers;
S2324, performing parameter dimension reduction operation on the morphological characteristics by using a morphological analysis method, projecting the dimension-reduced morphological characteristics to a low-dimensional space, and performing correlation elimination processing by using characteristic parameters of cross-section shape points to construct a morphological characteristic matrix;
the calculation formula of the mean value coefficient discrete degree of the cross section shape point is as follows:
；
wherein L represents the degree of dispersion of the mean value coefficient, T represents the number of cross-sectional shape points, E represents the distribution uniformity level in the graph, P represents the mean value coefficient of the cross-sectional shape points, and m (E) represents the number of cross-sectional shape points corresponding to the graph satisfying the distribution uniformity level;
The calculation formula of the mean value coefficient of the cross-section shape point is as follows:
；
wherein P represents the mean value coefficient of the cross-sectional shape points, T represents the number of the cross-sectional shape points, E represents the distribution uniformity level in the graph, and m (E) represents the number of the cross-sectional shape points corresponding to the graph satisfying the distribution uniformity level
The method for separating the fiber region from the background region in the textile fiber image based on the image segmentation technology and mapping the pixel points corresponding to the fiber region to the state space to obtain the prediction estimated value comprises the following steps:
s321, separating a fiber region from a background region in a textile fiber image based on a region growing mode selected by an image segmentation technology, and extracting fiber length characteristics from the segmented fiber region;
s322, setting a prediction estimated value attribute definition state space based on the fiber length characteristics, and mapping the pixel points corresponding to the fiber length characteristics to the state space;
S323, carrying out sectional mapping of brightness information on the fiber region in a state space to obtain a predicted estimated value of each region point, and optimizing the estimated value of the region point by adopting a guide filtering mode to obtain a final predicted estimated value;
The method for verifying the continuity of the prediction estimation value based on the continuity standard defines a matching rule of geometric features and point continuity and generates a coarse matching stage and a fine matching stage based on the matching rule comprises the following steps:
S331, selecting two groups of regional point sets according to a continuity standard, decomposing the regional point sets into discrete sets formed by break points, and judging the distance between the break points in the discrete sets by using a short-side distance algorithm to obtain a distance set;
S332, selecting the median of the distance set as a matching value, and judging the difference value between the median of the distance set of the predicted estimated value and the matching value;
S333, defining that the difference value is larger than or smaller than a set value, and indicating that the geometric features are not matched, and defining that the difference value is equal to the set value, indicating that the geometric features are matched;
S334, defining an edge set and an auxiliary edge set according to break points in the two selected discrete sets, and judging the continuity of the break points by utilizing the combination of the edge set and the auxiliary edge to generate a point continuity matching rule;
S335, generating a coarse matching stage and a fine matching stage based on the geometric feature matching rule and the point continuity matching rule, and verifying the continuity of the predicted estimated value by using the matching rule.
2. The intelligent identification system for the textile fiber based on the spectral data is used for realizing the intelligent identification method for the textile fiber according to claim 1, and is characterized by comprising a fiber crystallinity prediction module, an abrasion resistance evaluation module, a strength and toughness evaluation module and an identification platform generation module, wherein the fiber crystallinity prediction module, the abrasion resistance evaluation module, the strength and toughness evaluation module and the identification platform generation module are connected in sequence;
the fiber crystallinity prediction module is used for performing transformation analysis processing on the textile fiber spectrum data to obtain spectrum characteristics, determining spectrum characteristic wavelength based on the spectrum characteristics and coefficient correlation technique, and establishing a prediction model to output the textile fiber crystallinity;
The wear resistance evaluation module is used for establishing a mapping space based on a data analysis technology to judge the distribution condition of the crystallinity of the textile fiber to generate a morphological characteristic value, and evaluating the wear resistance of the textile fiber through similarity comparison;
the obdurability evaluation module is used for carrying out equalization and segmentation processing on the textile fiber image, and mapping the pixel points of the processed textile fiber image to a state space to evaluate the obdurability of the textile fiber;
The identification platform generation module is used for dynamically analyzing the application state of the textile fiber based on the abrasion resistance and toughness evaluation result and evaluating the service life of the textile fiber according to the application state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410772935.5A CN118329808B (en) | 2024-06-17 | 2024-06-17 | Intelligent identification method and system for textile fibers based on spectrum data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410772935.5A CN118329808B (en) | 2024-06-17 | 2024-06-17 | Intelligent identification method and system for textile fibers based on spectrum data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118329808A CN118329808A (en) | 2024-07-12 |
| CN118329808B true CN118329808B (en) | 2024-08-23 |
Family
ID=91766481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410772935.5A Active CN118329808B (en) | 2024-06-17 | 2024-06-17 | Intelligent identification method and system for textile fibers based on spectrum data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118329808B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120388018B (en) * | 2025-06-27 | 2025-08-22 | 常州纺兴精密机械有限公司 | A high-strength and high-modulus polyethylene fiber spinneret microstructure control system and method |
| CN120404644B (en) * | 2025-06-30 | 2025-09-02 | 苏州工学院 | Qualitative online analysis system and method for textile fiber components |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115656096A (en) * | 2022-09-20 | 2023-01-31 | 中山海关技术中心 | Qualitative identification test method for textile fibers |
| CN117705754A (en) * | 2023-11-30 | 2024-03-15 | 浙江大学 | Textile polyester fiber content online detection method based on hyperspectral imaging |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101876633B (en) * | 2009-11-13 | 2011-08-31 | 中国矿业大学 | Terahertz time domain spectroscopy-based textile fiber identification method |
| CN104730004B (en) * | 2015-03-25 | 2017-06-13 | 浙江理工大学 | The discrimination method of the textile fabric based on UV Diffuse Reflectance Spectroscopy |
| CN114088657B (en) * | 2021-10-09 | 2024-05-14 | 池明旻 | Textile fiber component analysis method based on depth encoder/decoder |
| CN117197137B (en) * | 2023-11-07 | 2024-02-09 | 山东大学 | Tissue sample analysis method and system based on hyperspectral images |
| CN118190859B (en) * | 2024-03-18 | 2024-10-01 | 西安工程大学 | Rapid identification method for wool and cashmere fibers |
-
2024
- 2024-06-17 CN CN202410772935.5A patent/CN118329808B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115656096A (en) * | 2022-09-20 | 2023-01-31 | 中山海关技术中心 | Qualitative identification test method for textile fibers |
| CN117705754A (en) * | 2023-11-30 | 2024-03-15 | 浙江大学 | Textile polyester fiber content online detection method based on hyperspectral imaging |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118329808A (en) | 2024-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118329808B (en) | Intelligent identification method and system for textile fibers based on spectrum data | |
| CN119417838B (en) | Semiconductor chip defect automatic detection method and system based on hyperspectral imaging | |
| CN118848670B (en) | Numerical control machine tool wear monitoring method based on deep learning | |
| CN119150175A (en) | Pipeline interior detection method and system based on multi-sensor fusion | |
| CN118429346B (en) | Color fastness evaluation method and system | |
| CN117994236B (en) | Vortex spun broken yarn intelligent detection method and system based on image analysis | |
| CN117974658B (en) | Cable anomaly identification method and system based on image processing | |
| CN109241870B (en) | Coal mine underground personnel identity identification method based on gait identification | |
| CN116754467B (en) | Evaluation method for ageing performance of natural rubber | |
| CN120339270A (en) | A method and system for monitoring the quality of stranded wire production based on machine vision | |
| CN118839142B (en) | Method and device for detecting center hole of wire drawing die | |
| CN118797532A (en) | Multi-modal fusion intelligent detection system for railway track contact network suspension monitoring | |
| CN119832318B (en) | A method and system for evaluating eggshell surface color quality based on color reflection | |
| CN119444670B (en) | A textile quality inspection method integrating multi-scale features | |
| CN119150005A (en) | Intelligent testing method and system for wire harness connection performance | |
| CN119510454A (en) | X-ray based online detection method and device for aramid fiber tape defects | |
| CN112613233A (en) | Algorithm for discovering environmental monitoring abnormal data based on single-classification support vector machine model | |
| CN114119551B (en) | A quantitative analysis method for face image quality | |
| CN114692701A (en) | A multi-image processing and pattern recognition method for electrical equipment defects | |
| Ozseven et al. | Surface defect detection and quantification with image processing methods | |
| CN120801284A (en) | Method for rapidly detecting compound components of traditional Chinese medicine | |
| CN119068373B (en) | Pipeline monitoring and maintaining method based on unmanned aerial vehicle | |
| CN119805153A (en) | Semiconductor testing method and device | |
| CN118229614A (en) | Quantitative statistics and safety diagnosis method of tunnel lining crack apparent disease | |
| CN117372352A (en) | An underwater image quality evaluation method based on multi-stream convolutional neural network |
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 |