CN117523013B

CN117523013B - Color change characteristic analysis method and system for high-moisture-conductivity fabric

Info

Publication number: CN117523013B
Application number: CN202410008569.6A
Authority: CN
Inventors: 卢雨正; 部一彤; 傅佳佳
Original assignee: Jiangnan University
Current assignee: Jiangnan University
Priority date: 2024-01-04
Filing date: 2024-01-04
Publication date: 2024-03-08
Anticipated expiration: 2044-01-04
Also published as: CN117523013A

Abstract

The invention relates to the technical field of color analysis of fabrics, in particular to a color change characteristic analysis method and system of a high-moisture-conductivity fabric. The method comprises the following steps: collecting standard visible spectrum data according to the high-moisture-conductivity fabric sample to generate standard visible spectrum data; carrying out sample humidity-adjusted visible spectrum data acquisition according to the humidity-adjusted high-moisture-conductivity fabric sample so as to establish a visible spectrum data matrix; performing fabric color change analysis on the visible spectrum data matrix according to the reference visible spectrum data to generate a fabric color change matrix; performing color change trend analysis according to the fabric color change matrix to generate a three-dimensional color change trend graph; and extracting features of the three-dimensional color change trend graph to generate color change feature data. The invention realizes the characteristic analysis of the color change of the high-moisture-conductivity fabric under different conditions.

Description

Color change characteristic analysis method and system for high-moisture-conductivity fabric

Technical Field

The invention relates to the technical field of color analysis of fabrics, in particular to a color change characteristic analysis method and system of a high-moisture-conductivity fabric.

Background

With the development of technology, high moisture-conductive fabrics are widely used in various fields, such as sportswear, outdoor equipment and the like. Because the high moisture-conductive fabric may generate different color change effects under different humidity conditions, a systematic and comprehensive color change characteristic analysis method is required. Through the comprehensive application of colorimetry and spectroscopy, the color change of the fabric in different humidity environments can be accurately measured, the optical response rule of the fabric is revealed, and scientific guidance is provided for production and manufacture. The analysis method can optimize the product design, improve the humidity adaptability and the comfort of the product, provide useful information for the application scene of the high-moisture-conductivity fabric and promote the wide application of the high-moisture-conductivity fabric in different fields. However, the conventional analysis method of color change characteristics of the high-moisture-conductivity fabric often analyzes the color change of the high-moisture-conductivity fabric through single humidity difference or illumination difference, so that the analysis effect is poor, the applicable application scene is too single, and the accurate analysis of the color change stability of the high-moisture-conductivity fabric under the conditions of different humidity and illumination cannot be comprehensively considered.

Disclosure of Invention

Based on the above, the present invention provides a method and a system for analyzing color change characteristics of a high moisture-conductive fabric, so as to solve at least one of the above technical problems.

In order to achieve the above purpose, the color change characteristic analysis method of the high moisture-conductive fabric comprises the following steps:

step S1: collecting standard visible spectrum data according to the high-moisture-conductivity fabric sample to generate standard visible spectrum data;

step S2: carrying out sample humidity adjustment treatment on the high-humidity-conductivity fabric sample to obtain a humidity-adjusted high-humidity-conductivity fabric sample; performing target area positioning on the humidity-adjusted high-moisture-conductivity fabric sample to obtain a target area of the humidity-adjusted high-moisture-conductivity fabric sample;

step S3: carrying out sample humidity-adjusted visible spectrum data acquisition according to a target area of the humidity-adjusted high-moisture-conductivity fabric sample so as to establish a visible spectrum data matrix;

step S4: according to the reference visible spectrum data, carrying out fabric color change analysis on the visible spectrum data matrix under different humidity to generate a fabric color change matrix; establishing an optimized color change analysis model based on a decision tree-regression algorithm and a visible spectrum data matrix;

step S5: transmitting a preset humidity data test interval and a preset irradiation spectrum wavelength test interval to an optimized color change analysis model for color change trend analysis, and generating a three-dimensional color change trend graph; extracting features of the three-dimensional color change trend graph to generate color change feature data; and performing color change stability calculation according to the color change characteristic data to generate color change stability data.

According to the method, the optical characteristic reference point of the fabric in a normal state without humidity adjustment can be established by acquiring the reference visible spectrum data. The reference data provides accurate description of the optical characteristics of the fabric in a dry environment, and lays a foundation for subsequent humidity adjustment treatment and color change analysis. Through the collection of the reference visible spectrum data, the light reflection, transmission and absorption conditions of the high-moisture-conductivity fabric under different wavelengths can be captured, a rich and reliable data basis is provided for subsequent comparison and analysis, and the accuracy and reliability of subsequent analysis are ensured. By adjusting the humidity of the sample, different humidity environments possibly encountered in actual use are simulated, more real conditions close to actual application scenes are provided for subsequent analysis, the humidity-adjusted high-conductivity wet fabric sample reflects the performance response of the fabric in the humidity environment, and a necessary basis is provided for comprehensively understanding the color change characteristics of the fabric under the humidity change. Meanwhile, in the process of positioning the target area of the humidity-regulated high-moisture-conductivity fabric sample, the area of interest, such as the most obvious part possibly affected by moisture, can be accurately positioned, so that the accuracy of analysis is improved, and the follow-up visible spectrum data acquisition and color change analysis have the operability and guidance of practical application. The method is characterized in that the method is focused on the data collection of the target areas of the high-moisture-conductivity fabric under different humidity conditions, the optical characteristic change of the fabric under the humidity adjustment state can be more accurately captured, the established visible spectrum data matrix is more accurate and is closely related to the fabric characteristic under the humidity change of interest, the optical response of the high-moisture-conductivity fabric under the humidity condition can be more comprehensively known, a more reliable data basis is provided for subsequent color change analysis, the accuracy and the repeatability of experimental results are ensured, and a reliable basis is provided for further researching the humidity sensitivity of the high-moisture-conductivity fabric. The color change of the high-moisture-conductivity fabric under different humidity conditions can be quantitatively analyzed through the comparison of the spectrum data under different humidity in the visible spectrum data matrix and the reference visible spectrum data, so that a fabric color change matrix is generated, and the difference of the optical characteristics of the fabric under the humidity change is revealed. Based on a decision tree-regression algorithm and a visible spectrum data matrix, an optimized color change analysis model is established, so that the influence mechanism of humidity on the color change of the high-moisture-conductivity fabric can be better understood, and the prediction capability and generalization capability of the model are improved. By combining a decision tree-regression algorithm, the color change trend under different humidity conditions can be more finely depicted, so that the model is more flexible and strong in adaptability, and more accurate guidance is provided for subsequent practical application. The method comprises the steps of transmitting a preset humidity data test interval and a preset irradiation spectrum wavelength test interval to an optimized color change analysis model, so that more comprehensive quantitative analysis of color change of the high-conductivity wet fabric can be realized, and transmitting the preset humidity data test interval and the irradiation spectrum wavelength test interval to the model, so that the model can perform comprehensive test and analysis within a set range, and various humidity conditions and illumination conditions with different wavelengths are covered. The color change trend analysis is carried out through the model to generate a three-dimensional color change trend graph, the overall color change rule of the high-moisture-conductivity fabric under different humidity and illumination conditions can be intuitively observed, the graph is subjected to characteristic extraction to generate color change characteristic data, and the key characteristics of the fabric color response under different humidity and illumination conditions are further analyzed. Finally, the color change stability is calculated according to the color change characteristic data to generate color change stability data, so that visual visualization of the color change data of the fabric under different humidity and illumination is provided, and numerical support is provided for evaluating the color change stability of the high-moisture-conductivity fabric in actual use.

Preferably, step S1 comprises the steps of:

performing light source adjustment on the light source equipment according to a preset irradiation spectrum wavelength to obtain an adjusted light source equipment;

illuminating the high moisture-conductive fabric sample with an adjusting light source device, and collecting standard visible spectrum data of a reflection spectrum formed by the high moisture-conductive fabric sample through a spectrometer to generate standard visible spectrum data, wherein the standard visible spectrum data comprises: and indexing the corresponding visible spectrum data according to the irradiation spectrum wavelength.

According to the invention, the light source equipment is successfully regulated by regulating the light source according to the preset irradiation spectrum wavelength, and stable illumination conditions are provided for subsequent experiments, so that the reliability and the repeatability of experimental results are improved. The high-conductivity wet fabric sample is irradiated by utilizing the adjusting light source equipment, the reflection spectrum formed by the sample is subjected to visible spectrum data acquisition by the spectrometer, reference visible spectrum data are generated, the optical responses of the high-conductivity wet fabric under different wavelengths are accurately recorded, accurate and comprehensive reference data are provided for subsequent comparison and analysis, the high-quality acquisition of the reference visible spectrum data is ensured, and a solid foundation is laid for the accuracy of the whole color change characteristic analysis method.

Preferably, step S2 comprises the steps of:

step S21: carrying out sample humidity adjustment treatment on the high-humidity-conductivity fabric sample to obtain a humidity-adjusted high-humidity-conductivity fabric sample;

step S22: image data acquisition is carried out on the high-moisture-conductivity fabric sample with the humidity adjusted by using monitoring equipment, so that fabric image data are generated;

step S23: performing image edge detection on the fabric image data by using an edge detection algorithm to generate an edge fabric image;

step S24: carrying out image color core region segmentation on the fabric image data according to the edge fabric image to obtain core fabric image data;

step S25: and positioning a target area of the humidity-adjusted high-moisture-conductivity fabric sample according to the core fabric image data so as to obtain the target area of the humidity-adjusted high-moisture-conductivity fabric sample.

According to the invention, the sample humidity adjustment treatment is carried out on the high-humidity-conductivity fabric sample so as to obtain the humidity-adjusted high-humidity-conductivity fabric sample, and the fabric is in a state closer to an actual application scene by simulating the actual use environment under different humidity conditions, so that more real and targeted data are provided for subsequent color change characteristic analysis. The monitoring equipment is utilized to collect image data of the humidity-regulated high-moisture-conductivity fabric sample, so that fabric image data is generated, visual information of the humidity-regulated high-moisture-conductivity fabric can be comprehensively captured through image data collection, sufficient original data is provided for subsequent image processing and analysis, and the image characteristics of the fabric under humidity change can be more accurately and comprehensively known. The edge detection algorithm is utilized to carry out image edge detection on the fabric image data, an edge fabric image is generated, the edge characteristics of the fabric are emphasized, the texture and outline information of the fabric after humidity adjustment can be captured more clearly through the protruding edge, a more accurate basis is provided for subsequent target area positioning, and the accuracy and reliability of analysis are improved. And (3) image color core region segmentation is carried out on the fabric image data according to the edge fabric image, so that core fabric image data is obtained, redundant information of the whole image is effectively reduced by segmenting the core fabric image, and attention is focused on a key region of the fabric. The method is beneficial to improving the efficiency and the precision of subsequent analysis, so that the core characteristics of the humidity-regulated high-moisture-conductivity fabric sample are more focused in the color change characteristic analysis. The target area positioning is carried out on the humidity-regulated high-moisture-conductivity fabric sample according to the image data of the core fabric, so that the interested area, namely the target area of the humidity-regulated high-moisture-conductivity fabric sample, can be accurately positioned, is beneficial to improving the accuracy and pertinence of analysis, ensures that the subsequent humidity sensitivity analysis and color change research are concentrated in a key area, and further comprehensively knows the characteristics of the high-moisture-conductivity fabric under the humidity change.

Preferably, step S3 comprises the steps of:

step S31: acquiring fabric humidity data of a target area of a humidity-adjusted high-moisture-conductivity fabric sample by using a sensor so as to obtain fabric humidity data;

step S32: irradiating a target area of the humidity-adjusted high-moisture-conductivity fabric sample by using an adjusting light source device, and collecting visible spectrum data of sample humidity adjustment on a reflection spectrum formed by the humidity-adjusted high-moisture-conductivity fabric sample through a spectrometer to generate adjusting visible spectrum data;

step S33: establishing a visible spectrum data matrix according to fabric humidity data and adjusting visible spectrum data, wherein the visible spectrum data matrix comprises: and adjusting visible spectrum data corresponding to the transverse index and the longitudinal index of the irradiation spectrum wavelength according to the fabric humidity data.

According to the invention, the sensor is used for acquiring the fabric humidity data of the target area of the humidity-regulated high-humidity-conductivity fabric sample so as to obtain the fabric humidity data, and the sensor is used for acquiring the fabric humidity data, so that the humidity information of the humidity-regulated high-humidity-conductivity fabric sample can be directly and accurately acquired, a key physical property parameter in an experiment is provided, and the deep understanding of the fabric performance under the humidity change is facilitated. The target area of the humidity-adjusted high-conductivity wet fabric sample is irradiated by the adjusting light source equipment, and the visible spectrum data acquisition of sample humidity adjustment is carried out on the reflection spectrum formed by the humidity-adjusted high-conductivity wet fabric sample through the spectrometer so as to generate adjusting visible spectrum data, so that the optical characteristics of the fabric under different humidity conditions can be captured, and a rich data basis is provided for the establishment of a subsequent visible spectrum data matrix. The visible spectrum data matrix is established according to the fabric humidity data and the visible spectrum data is adjusted, the humidity data and the spectrum data are combined to form a comprehensive data matrix, the change of the humidity-adjusted high-moisture-conductivity fabric in the visible spectrum range is reflected, the comprehensive humidity-optical characteristic association is established, and more detailed and comprehensive data support is provided for subsequent color change characteristic analysis.

Preferably, step S4 comprises the steps of:

step S41: performing visible spectrum difference calculation on the visible spectrum data matrix under different humidity according to the standard visible spectrum data to generate a visible spectrum difference matrix;

step S42: performing fabric color change analysis according to the visible spectrum difference matrix to generate a fabric color change matrix;

step S43: establishing a mapping relation of color change analysis by utilizing a decision tree-regression algorithm to obtain an initial color change analysis model;

step S44: and performing model training optimization on the initial color change analysis model by using the fabric color change matrix to generate an optimized color change analysis model.

According to the visible spectrum data matrix, the visible spectrum difference calculation is carried out on the visible spectrum data matrix under different humidity conditions according to the reference visible spectrum data, the visible spectrum difference matrix is generated, the optical characteristic change of the fabric under the humidity change can be quantitatively captured through calculating the visible spectrum difference under the different humidity conditions, the visible spectrum difference matrix provides a clear data visual angle, and a powerful basis is provided for subsequent fabric color change analysis. And carrying out fabric color change analysis according to the visible spectrum difference matrix to generate a fabric color change matrix, so that the color change trend of the fabric under different humidity conditions can be deeply known, detailed experimental data is provided for subsequent modeling, the fabric color change matrix reflects the specific color response of the fabric under humidity change, and a foundation is laid for establishing a color change analysis model. The mapping relation of color change analysis is established by utilizing a decision tree-regression algorithm to obtain an initial color change analysis model, and the association model of color change, humidity and illumination is established by utilizing a machine learning algorithm, so that the color change trend of the fabric under different humidity conditions can be predicted more accurately. The use of decision tree-regression algorithms increases the flexibility and adaptability of the model. Model training optimization is carried out on the initial color change analysis model by using the fabric color change matrix to generate an optimized color change analysis model, and through training optimization on the initial model, the accuracy and generalization capability of the model can be improved, so that the model is better adapted to fabric color change under different humidity environments, the prediction performance of the model can be improved, and more accurate and reliable model support is provided for subsequent color change trend analysis.

Preferably, step S44 includes the steps of:

step S441: dividing the fabric color change matrix into data, and respectively generating a fabric color change training set, a fabric color change verification set and a fabric color change test set;

step S442: transmitting the fabric color change training set to an initial color change analysis model for model training, and generating a training color change analysis model;

step S443: performing model verification evaluation on the training color change analysis model according to the fabric color change verification set to generate model verification evaluation data;

step S444: performing model parameter optimization adjustment on the training color change analysis model according to a Bayesian optimization algorithm and model verification evaluation data to generate an optimized color change analysis model to be tested;

step S445: and performing model test on the optimized color change analysis model to be tested by using the fabric color change test set to generate the optimized color change analysis model.

According to the invention, the fabric color change matrix is subjected to data division to respectively generate the fabric color change training set, the fabric color change verification set and the fabric color change test set, the generalization performance of the model can be effectively evaluated by dividing the data set, the training set is used for parameter learning of the model, the verification set is used for parameter adjustment and verification of the model, and the test set is used for evaluating the performance of the model, so that the robustness and applicability of the model are improved. And transmitting the fabric color change training set to an initial color change analysis model for model training, generating a training color change analysis model, and adjusting model parameters through a training data set, so that the model can better fit the real situation of fabric color change, and the prediction accuracy of the model is improved. And carrying out model verification evaluation on the training color change analysis model according to the fabric color change verification set to generate model verification evaluation data, and obtaining the performance of the model on unseen data through evaluation of the verification set, so as to provide powerful reference for subsequent model optimization. And carrying out model parameter optimization adjustment on the training color change analysis model according to a Bayesian optimization algorithm and model verification evaluation data to generate an optimized color change analysis model to be tested, searching optimal parameters of the model through the optimization algorithm, and further improving the performance and generalization capability of the model. And carrying out model test on the optimized color change analysis model to be tested by using the fabric color change test set to generate the optimized color change analysis model, and comprehensively evaluating the performance of the model and ensuring the robustness and reliability of the model under different conditions through evaluating the test set.

Preferably, step S444 includes the steps of:

step S401: selecting target model verification data according to the model verification evaluation data by using a Bayesian optimization algorithm to obtain target model verification data; performing model parameter adjustment processing on the training color change analysis model according to model target model verification data to generate a parameter-optimized training color change analysis model;

step S402: and (3) iteratively executing the step S401 according to the preset model optimization iteration times, stopping iterative execution when the execution times of the step S401 are not less than the preset model optimization iteration times, and marking the training color change analysis model with optimized parameters as an optimized color change analysis model to be tested.

According to the invention, the model verification evaluation data is selected according to the Bayesian optimization algorithm to obtain the target model verification data, and verification data can be intelligently selected through application of the Bayesian optimization algorithm, so that adjustment of model parameters is more effectively guided, the efficiency of the optimization algorithm and the performance of the model are improved, the training color change analysis model is subjected to model parameter adjustment according to the target model verification data, and a training color change analysis model with optimized parameters is generated, so that the parameter adjustment of the model is finer and personalized to adapt to different data distribution and characteristics. By executing the optimization algorithm for multiple iterations, the optimal solution can be more comprehensively found in the search space of the model parameters, after the iteration is stopped, the obtained optimized color change analysis model to be tested has higher performance and generalization capability, more reliable model support is provided for subsequent fabric color change feature analysis, model parameters and an iteration process which enable model prediction results to be more accurate are found through the Bayesian optimization algorithm, and the parameter adjustment efficiency and final performance of the model are improved.

Preferably, step S5 comprises the steps of:

step S51: transmitting a preset humidity data test interval and a preset irradiation spectrum wavelength test interval to an optimized color change analysis model for color change trend analysis, and generating a three-dimensional color change trend graph;

step S52: gradient calculation is carried out on the three-dimensional color change trend graph, so that trend graph gradient data are obtained;

step S53: extracting color change extreme points of the three-dimensional color change trend graph according to the trend graph gradient data to generate color change extreme value data;

step S54: extracting features of the three-dimensional color change trend graph by using a principal component analysis method to generate color change feature data;

step S55: and performing color change stability calculation on the color change extremum data and the color change characteristic data by using a color change stability calculation formula to generate color change stability data.

According to the invention, the preset humidity data test interval and the preset irradiation spectrum wavelength test interval are transmitted to the optimized color change analysis model for color change trend analysis, a three-dimensional color change trend chart is generated, the optimized color change analysis model can accurately predict color changes under different humidity and spectrum wavelength by transmitting the preset test interval data, and an intuitive three-dimensional color change trend chart is generated, so that a visual basis is provided for further analysis. Gradient calculation is carried out on the three-dimensional color change trend graph, so that trend graph gradient data is obtained, the quantitative analysis of the change rate of color change is facilitated, and through gradient calculation, the change trend information of each point in the three-dimensional color change trend graph can be obtained, so that a data basis is provided for subsequent extremum point extraction and feature extraction. And extracting color change extreme points of the three-dimensional color change trend graph according to the trend graph gradient data to generate color change extreme value data, and determining key characteristics in the color change trend by searching extreme points of the trend graph gradient, thereby being beneficial to identifying peaks and valleys of color change and providing deeper color change information. The three-dimensional color change trend graph is subjected to feature extraction by using a principal component analysis method to generate color change feature data, the most obvious color change feature can be reduced in dimension and extracted by using the principal component analysis method, the data structure is simplified, main information is reserved, and more effective feature representation is provided for subsequent color change stability calculation. The color change stability calculation formula is utilized to calculate the color change stability of the color change extremum data and the color change characteristic data, the color change stability data is generated, the stability of the color change of the fabric under different humidity and spectrum wavelength can be evaluated through calculating the color change stability, important performance indexes are provided for the application of the high moisture-conducting fabric, the information of a color change trend chart is deeply mined, and comprehensive information is provided for the color change characteristics of the high moisture-conducting fabric.

Preferably, the color change stability calculation formula in step S55 is as follows:

；

in the method, in the process of the invention,expressed as colour change stability data>Data quantity expressed as color change extremum data, < +.>Weight information expressed as hue value deviation, +.>Denoted as +.>Hue value variance data of the individual color variation extremum data,/->Weight information expressed as contrast deviation, +.>Denoted as +.>Color contrast variance data of the individual color variation extremum data, < ->Expressed as color change featuresData volume of data->Denoted as +.>Hue value variance data of individual color variation characteristic data,/-, and the like>Denoted as +.>Color contrast variance data of the individual color change feature data.

The invention utilizes a color change stability calculation formula which fully considers the data quantity of color change extremum dataWeight information of hue value deviation>First->Hue value variance data of the individual color variation extremum data +.>Weight information of contrast deviation->First->Color contrast variance data of the individual color variation extremum data +.>Data volume of color change characteristic data +.>First->Hue value variance data of individual color variation characteristic data +.>First->Color contrast variance data of individual color variation characteristic data +. >And interactions between functions to form a functional relationship:

that is to say,the functional relation provides a comprehensive and careful analysis for measuring the stability of the color change of the fabric. The formula takes into account the contribution of hue values and contrast in the overall color change by taking into account the deviation of both. Wherein the weight parameter->And->The relative importance of hue value and contrast in calculation and the calculation result of the normalized functional relation are respectively adjusted. />Focusing on the extreme value data of color change, namely analyzing the color change poles of the fabric under different humidity and different illumination light source intensities, and performing +.>The color change characteristic data, namely the characteristic of the overall color change trend of the fabric, is focused on and analyzed. The functional relation quantifies fluctuations in hue and contrast by taking into account hue value variances and contrast variances, thereby knowing the stability of color changes at different humidities and different illumination light source intensities.

The present disclosure provides a color change feature analysis system of a high moisture-conductive fabric, for performing the color change feature analysis method of a high moisture-conductive fabric, where the color change feature analysis system of a high moisture-conductive fabric includes:

The standard visible spectrum data acquisition module is used for acquiring standard visible spectrum data according to the high-moisture-conductivity fabric sample so as to generate standard visible spectrum data;

the target humidity area positioning module is used for carrying out sample humidity adjustment treatment on the high-humidity-conductivity fabric sample so as to obtain a humidity-adjusted high-humidity-conductivity fabric sample; performing target area positioning on the humidity-adjusted high-moisture-conductivity fabric sample to obtain a target area of the humidity-adjusted high-moisture-conductivity fabric sample;

the visible spectrum matrix establishing module is used for acquiring visible spectrum data of sample humidity adjustment according to a target area of the high-conductivity wet fabric sample with humidity adjustment so as to establish a visible spectrum data matrix;

the optimized color change analysis model building module is used for carrying out fabric color change analysis on the visible spectrum data matrix under different humidity according to the standard visible spectrum data to generate a fabric color change matrix; establishing an optimized color change analysis model based on a decision tree-regression algorithm and a visible spectrum data matrix;

the color change characteristic analysis module is used for transmitting a preset humidity data test interval and a preset irradiation spectrum wavelength test interval to the optimized color change analysis model for color change trend analysis to generate a three-dimensional color change trend graph; extracting features of the three-dimensional color change trend graph to generate color change feature data; and performing color change stability calculation according to the color change characteristic data to generate color change stability data.

The color change characteristic analysis method for the high-moisture-conductivity fabric has the beneficial effects that the analysis of the color change of the high-moisture-conductivity fabric is not carried out through single humidity difference or illumination difference, but the humidity difference which has important influence on the high-moisture-conductivity fabric and the color change characteristic under the illumination difference are considered at the same time, so that the color change characteristic analysis effect on the high-moisture-conductivity fabric is better, the applicable application scene is wider, the accurate analysis of the color change stability of the high-moisture-conductivity fabric under the conditions of different humidity and illumination can be comprehensively considered, and the response degree of the color change of the high-moisture-conductivity fabric is clearly known.

Drawings

FIG. 1 is a schematic flow chart of the steps of a color change characteristic analysis method of a high moisture-conductive fabric;

FIG. 2 is a flowchart illustrating the detailed implementation of step S2 in FIG. 1;

FIG. 3 is a flowchart illustrating the detailed implementation of step S4 in FIG. 1;

FIG. 4 is a flowchart illustrating the detailed implementation of step S5 in FIG. 1;

the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.

Detailed Description

The following is a clear and complete description of the technical method of the present patent in conjunction with the accompanying drawings, and it is evident that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, are intended to fall within the scope of the present invention.

Furthermore, the drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus a repetitive description thereof will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. The functional entities may be implemented in software or in one or more hardware modules or integrated circuits or in different networks and/or processor methods and/or microcontroller methods.

It will be understood that, although the terms "first," "second," etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

In order to achieve the above objective, referring to fig. 1 to 4, the present invention provides a color change feature analysis method of a high moisture-conductive fabric, comprising the following steps:

In the embodiment of the present invention, as described with reference to fig. 1, the step flow diagram of the color change characteristic analysis method of the high moisture-conductive fabric of the present invention is shown, and in the embodiment, the color change characteristic analysis method of the high moisture-conductive fabric includes the following steps:

in the embodiment of the invention, a professional visible spectrum instrument and a preset light source are used for carrying out light source reflection treatment on the high-moisture-conductivity fabric sample, the light source is adjusted to ensure proper irradiation intensity and wavelength coverage of a required optical characteristic area, the high-moisture-conductivity fabric sample is placed at a proper position, the high-moisture-conductivity fabric sample is subjected to reflection spectrum acquisition by using the spectrum instrument, and spectral data of reflection spectrum or transmission in a visible spectrum range is recorded to obtain standard visible spectrum data for subsequent analysis and comparison.

In the embodiment of the invention, the sample humidity adjustment treatment of the high-moisture-conductivity fabric sample can be realized by using professional humidity adjustment equipment or an environment control room. The conditioning apparatus may expose the highly moisture conductive fabric to specific humidity conditions according to the experimental design, ensuring that the sample is adequately conditioned over a range of humidity levels. Target area location is performed on the humidity conditioned high moisture conductive fabric sample, image or data acquisition is performed on the fabric using a monitoring device such as a camera or sensor, and then image processing algorithms, such as edge detection and target detection, are applied to locate and identify the target area of interest in the high moisture conductive fabric sample, obtain the humidity conditioned high moisture conductive fabric sample, and precisely locate the area of interest in the sample for focusing on the color change of the specific area in subsequent analysis.

in the embodiment of the invention, the target area of the humidity-regulated high-conductivity fabric sample is irradiated by the preset light source intensity, the visible spectrum data acquisition is carried out on the target area of the humidity-regulated high-conductivity fabric sample, the spectrum wavelength range and the resolution which are acquired by using the spectrum instrument which is calibrated and regulated in advance are ensured to meet the experimental requirements, the instrument is aligned to the target area so as to acquire the reflection spectrum or the transmission spectrum of the area, the sensor of the instrument can be placed on the target area, the spectrum data received in the whole visible spectrum range is recorded, and the acquired data can comprise the spectrum reactions of the target area under different humidity conditions and different light source intensities. The obtained visible spectrum data are then consolidated into a data matrix, each row of the matrix representing a different humidity condition and each column representing a corresponding visible spectrum data of different light source intensities, such data matrix forming the basis data for analyzing the color change of the fabric, thereby creating a visible spectrum data matrix.

in the embodiment of the invention, fabric color change analysis under different humidity is carried out on the visible spectrum data matrix according to the standard visible spectrum data. For the spectrum data under each humidity condition, the color change of the fabric can be calculated by comparing the difference with the reference spectrum data, the color change of the fabric can be calculated by comparing the difference of different spectrum wavelengths in the visible spectrum, and the color change of the fabric can be calculated according to the difference, so that a fabric color change matrix is obtained, and each node in the matrix represents the fabric color change data under different humidity conditions and different brightness intensities. And then, establishing an optimized color change analysis model based on a decision tree-regression algorithm and a visible spectrum data matrix, wherein the decision tree-regression algorithm can be used for modeling the relationship between humidity and color change, and can learn a complex nonlinear relationship through training the model, so that the prediction precision of the color change is improved. In the model training process, the humidity condition in the existing visible spectrum data matrix, the spectrum wavelength for irradiating the high moisture-conducting fabric and the color change data of the fabric are used as input and output.

In the embodiment of the invention, the preset humidity data test interval and the preset irradiation spectrum wavelength test interval are transmitted to the optimized color change analysis model for color change trend analysis, and the model can predict the color change trend of the fabric under the conditions by inputting the preset humidity data and the irradiation spectrum wavelength data into the model, so that a three-dimensional color change trend graph containing humidity, spectrum wavelength and color change can be obtained. And extracting the characteristics of the three-dimensional color change trend graph. By carrying out gradient calculation, extreme point extraction or other feature extraction methods on the trend graph, key features in the trend graph, such as peaks and valleys of colors, can be captured to form color change feature data. And performing color change stability calculation according to the color change characteristic data to generate color change stability data. By using a proper stability calculation formula, the stability of the color change of the fabric under different humidity and spectrum conditions can be evaluated, the color change performance of the high-moisture-conductivity fabric can be quantified, and an important index is provided for the reliability of the fabric in practical application.

Preferably, step S1 comprises the steps of:

In the embodiment of the invention, the light source device is subjected to light source adjustment according to the preset irradiation spectrum wavelength. For example, the light source device is set to emit light of a specific wavelength, which can be adjusted by a wavelength selector of the light source or a light source intensity adjuster, assuming that a specific wavelength in the visible light range, such as 550 to 800 nm, is selected as the irradiation spectrum wavelength, and the interval is 10 nm or the like. And then, the high-moisture-conductivity fabric sample is irradiated by utilizing the light source adjusting device, the high-moisture-conductivity fabric sample is placed in the irradiation range of the light source, the uniform irradiation of the light source on the surface of the sample is ensured, and the high-moisture-conductivity fabric is irradiated by light with a specific wavelength. And then, carrying out visible spectrum data acquisition on a reflection spectrum formed by the high-moisture-conductivity fabric sample through a spectrometer. The spectrometer measures the reflection spectrum of the high-conductivity wet fabric in the visible light range, records the reflection intensity at each wavelength, and takes 550 nanometers as an example, can obtain reflection spectrum data at the wavelength, and can obtain standard visible spectrum data, wherein the standard visible spectrum data comprises visible spectrum data corresponding to a preset irradiation spectrum wavelength as an index, such as visible spectrum data corresponding to an irradiation spectrum wavelength of 550 nanometers.

Preferably, step S2 comprises the steps of:

As an example of the present invention, referring to fig. 2, a detailed implementation step flow diagram of step S2 in fig. 1 is shown, where step S2 includes:

in the embodiment of the invention, the high-moisture-conductivity fabric sample is placed in a humidity control chamber or equipment, so that the humidity in the environment can be accurately regulated. Professional humidity control devices, such as a hygrometer or a humidity cabinet, are used to control and adjust the humidity level of the environment in which the sample is located. The desired humidity value is set, for example, to 50% rh (relative humidity). Along with the adjustment of the humidity control equipment, the humidity in the environment is monitored and ensured to gradually reach a set value, and after the humidity is stable, the high-moisture-conductivity fabric sample is subjected to humidity adjustment treatment, so that the high-moisture-conductivity fabric sample subjected to humidity adjustment is obtained.

in the embodiment of the invention, after the humidity adjusting device is stable and the target humidity is set, a professional monitoring device, such as a high-resolution camera or an image acquisition device, is used for carrying out real-time image acquisition on the humidity-adjusted high-moisture-conductivity fabric sample. The arrangement and the position of the monitoring equipment can be ensured to comprehensively and accurately capture the image of the high-moisture-conductivity fabric sample, and the proper acquisition frequency is set to ensure continuous monitoring so as to acquire the fabric state change under different humidity conditions. The collected image data can comprise appearance characteristics and texture information of the high moisture-conductive fabric under different humidity conditions, and the image data can be used as a basis for subsequent analysis and used for carrying out detailed study on color change of the high moisture-conductive fabric under different humidity conditions.

in the embodiment of the invention, the specific embodiment of carrying out image edge detection on the fabric image data by utilizing an edge detection algorithm is as follows: first, image data of a humidity-conditioned high moisture-conductive fabric sample acquired by a monitoring device is acquired. Next, a classical edge detection algorithm, such as the Sobel operator or Canny edge detection algorithm, is applied. Taking the Sobel operator as an example, the algorithm emphasizes edge features in the image by performing convolution operation on the image. For each pixel, the Sobel operator calculates the gradient of the Sobel operator in the horizontal and vertical directions respectively, and then obtains the gradient amplitude by means of square sum evolution, thereby obtaining the intensity information of the edge in the image. After the edge detection algorithm is applied, the generated edge fabric image highlights the edge structure of the high-moisture-conductivity fabric sample, and clearer and more outstanding edge information is provided.

in the embodiment of the invention, the edge fabric image is converted from an RGB color space to other color spaces, such as HSV (hue, saturation, brightness) or Lab (brightness, green-red, blue-yellow) and the like. The image is divided into different areas according to the color intensity or the color component threshold value by using the threshold value dividing method, and the selection of the threshold value can be adjusted according to the color characteristic of the high-moisture-conductivity fabric so as to ensure the correct division of the color core area. Morphological operations such as erosion and dilation may be applied to further optimize the segmentation results to obtain core fabric image data.

In the embodiment of the invention, the core fabric image is analyzed through an image processing technology and a target positioning algorithm to accurately position the target area of the humidity-regulated high-moisture-conductivity fabric sample. And contour extraction is carried out on the core fabric image by utilizing a contour detection algorithm in image processing, such as a contour detection function in an OpenCV library, so that the overall shape and structure of the fabric can be recognized. The target area of the highly moisture conductive fabric sample, which may be moisture conditioned, is screened by setting appropriate screening conditions, such as area, aspect ratio, etc., which may be tailored to the specific application and sample characteristics. And (3) positioning the target area of the humidity-regulated high-moisture-conductivity fabric sample by calculating the central coordinate or other reference points of the target area.

Preferably, step S3 comprises the steps of:

In the embodiment of the invention, a sensor suitable for humidity measurement, such as a capacitive humidity sensor, is selected and placed in a target area of a humidity-regulated high-moisture-conductivity fabric sample, and the position of the sensor should cover the whole target area so as to ensure that comprehensive humidity information is acquired. The humidity of the fabric is measured in real time by a sensor, and the humidity data is recorded in a digital form. The output of the sensor may be a voltage, a current or a digital signal. Calibration of the sensor may be required to ensure that the acquired humidity data is accurate and reliable, and may be achieved by calibrating the sensor under known humidity conditions.

An adjusting light source device suitable for illumination experiments is selected, wherein the adjusting light source device sets the light source intensity for the steps, and the irradiation spectrum is ensured to be the same as the spectrum of the steps. And irradiating the adjusted light source equipment on a target area of the humidity-adjusted high-moisture-conductivity fabric sample, so as to ensure that the uniform irradiation of the light source covers the whole target area and obtain comprehensive spectrum information. The spectrum reflected by the target area of the humidity-regulated high-moisture-conductivity fabric sample is collected by using a spectrometer, the spectrometer can record the reflected spectrum intensities at different wavelengths, and the data are stored in a digital form, so that the regulated visible spectrum data are generated. And forming a matrix for adjusting visible spectrum data according to the collected spectrum data, and establishing the relationship between the humidity of the fabric and the irradiation spectrum wavelength and the visible spectrum. Preprocessing the fabric humidity data acquired from the sensor to ensure the accuracy and consistency of the data, wherein possible preprocessing steps comprise abnormal value removal, data normalization or standardization and the like. And taking the pretreated fabric humidity data as a transverse index of the matrix, and taking the recorded irradiation spectrum wavelength as a longitudinal index of the matrix according to the adjustment of visible spectrum data acquisition. Thus, each matrix element corresponds to visible spectrum data at a particular humidity and wavelength. If the fabric moisture data acquisition range is 0% to 100% and the illumination spectrum wavelength range is 400nm to 700nm, then each element of the visible spectrum data matrix will contain reflectance spectrum data in this moisture range and wavelength range. All acquired visible spectrum data are organized into a two-dimensional matrix according to the mode, namely a visible spectrum data matrix, the matrix provides a structured data base for subsequent analysis, and the relationship between the humidity of the fabric and the wavelength of an irradiation spectrum can be intuitively displayed.

Preferably, step S4 comprises the steps of:

As an example of the present invention, referring to fig. 3, a detailed implementation step flow diagram of step S4 in fig. 1 is shown, where step S4 includes:

in the embodiment of the invention, for each humidity condition in the visible spectrum data matrix, the difference between the reference visible spectrum data under each irradiation spectrum data and the visible spectrum of the visible spectrum data under the corresponding humidity is calculated, and the difference, the ratio and the like of the spectrum data can be adopted to calculate, so as to obtain a visible spectrum difference matrix, wherein the visible spectrum difference matrix contains the spectrum change condition of the high moisture-conducting fabric under different humidity.

in the embodiment of the invention, the color change analysis of the fabric is performed by using the difference matrix, and for the visible spectrum difference matrix under each humidity condition, further analysis can be performed by adopting image processing and computer vision technology, which may include filtering, threshold processing and other methods in image processing or converting the frequency of each visible spectrum into color information by using inverse Fourier transform technology, so as to extract key color change information, thereby generating the fabric color change matrix, wherein the color change condition of the fabric under different humidity levels and different irradiation spectrum intensities is recorded.

in the embodiment of the invention, the color change matrix data of the fabric are collected, wherein the data comprise visible spectrum difference matrixes of the fabric under different humidity and different irradiation spectrum intensities and corresponding color change conditions, and each data point consists of multidimensional characteristics, wherein the data comprise elements in the humidity, the irradiation spectrum wavelength and the color change matrixes. And (3) designing model parameters of a decision-regression algorithm, such as maximum depth, the number of split nodes and the like, according to the fabric color change matrix, so as to construct a decision tree-regression model. In this model, humidity and illumination spectral wavelength are used as inputs, while elements in the color change matrix are used as outputs. The decision tree algorithm establishes mapping relation between humidity and irradiation spectrum wavelength and color change by recursively dividing an input space. The decision tree-regression algorithm has the advantage of being capable of processing complex nonlinear relationships and has better interpretation. The model is built by learning and fitting the data to capture the effect of humidity and illumination spectral wavelength on the changes in the color of the fabric to the maximum.

In the embodiment of the invention, the collected fabric color change matrix data is divided into a training set, a verification set and a test set, wherein the training set is used for learning and training a model, the verification set is used for adjusting model parameters, and the test set is used for evaluating the performance of the model. Transmitting the training set to an initial color change analysis model for model training, and gradually optimizing parameters of the model by learning the relation between the color change matrix of the fabric and the actual humidity and the irradiation spectrum wavelength so as to more accurately predict the color change condition of the fabric under different humidity. And carrying out verification and evaluation on the trained model by using a verification set, and carrying out performance evaluation on the model by comparing the difference between the predicted result and the actual observed value of the model on the verification set. The process may include adjusting the super parameters of the model to improve the generalization ability of the model, performing further parameter optimization adjustment on the model by using a bayesian optimization algorithm, and finding an optimal combination of model parameters by iteratively optimizing the model parameters, so that the model is more superior in unseen data. And transmitting the test set to an optimized color change analysis model for testing, and generating a final optimized color change analysis model which can more accurately predict the color change trend of the high moisture-conductive fabric under different humidity conditions.

Preferably, step S44 includes the steps of:

In the embodiment of the invention, the fabric color change matrix is subjected to data division to respectively generate the fabric color change training set, the fabric color change verification set and the fabric color change test set, for example, 70% of the data can be used for training, 15% of the data can be used for verification, and 15% of the data can be used for testing. Transmitting the fabric color change training set to an initial color change analysis model for model training, wherein the initial color change analysis model learns the mapping relation of fabric color change under different humidity and different irradiation spectrum intensities according to the characteristics of a decision tree-regression algorithm, and the ultra-parameters such as splitting standard of the decision tree, tree depth and the like are adjusted by repeatedly learning and adjusting the training set, so that the fitting capacity of the color change mode is gradually improved. And carrying out model verification and evaluation on the training color change analysis model according to the fabric color change verification set, and inputting the prepared fabric color change verification set into the trained color change analysis model. For each sample, the model predicts by using the learned mapping relation, generates a corresponding fabric color change predicted value, compares the predicted value with a real target value of the verification set, and can calculate a predicted performance index of the model on the verification set, such as a mean square error (Mean Squared Error, MSE) or other evaluation indexes, wherein the indexes reflect generalization capability and accuracy of the model on unknown data, and the generated model verification evaluation data comprises various performance indexes of the model on the verification set. And carrying out model parameter optimization adjustment on the training color change analysis model according to the Bayesian optimization algorithm and the model verification evaluation data. An initial parameter space required by the Bayesian optimization algorithm is prepared, and the parameter space comprises the value range of each adjustable parameter of the model, such as a learning rate, a tree depth, the minimum number of samples of the split nodes and the like. An optimization objective function is defined, which takes as input the output data of the model and the data set of the real conditions corresponding to the output data, and returns performance indicators, such as mean square error or other evaluation indicators, on the verification set, and this objective function is to be used as an optimization target of the bayesian optimization algorithm. Searching the optimal parameter combination in the defined parameter space by using a Bayesian optimization algorithm, and gradually optimizing an objective function by continuously exploring and utilizing the parameter space to find the parameter configuration with the best performance of the model on the verification set. In each iteration, the bayesian optimization algorithm selects a new set of parameters for evaluation, and then updates the parameters of the model according to the performance of the new set of parameters on the verification set, and the process is repeated for a plurality of times until a set stopping condition is reached, such as the maximum number of iterations is reached or an optimal parameter combination meeting the performance requirement is found. The method comprises the steps of carrying out model test on an optimized color change analysis model to be tested by using a fabric color change test set, so that the color change trend is predicted by the optimized color change analysis model to be tested, the output of the model is the color change prediction of each sample under the given humidity condition, the model output value generated according to the test set is compared with a real condition value, the predicted result of the model is compared with the color change condition actually observed in the test set, different performance indexes (such as mean square error, accuracy and the like) can be used for evaluating the performance of the model on the test set, when the model output effect is poor, the step S442 is returned to readjust the initial parameters of the model until the model effect reaches the preset model target, and the optimized color change analysis model is generated by the result of the model test.

Preferably, step S444 includes the steps of:

In the embodiment of the invention, an optimization objective function is defined, the function receives the output data of the model and the data set of the real condition corresponding to the output data as input, and returns performance indexes such as mean square error or other evaluation indexes on the verification set, and the objective function is used as an optimization target of a Bayesian optimization algorithm. Searching the optimal parameter combination in the defined parameter space by using a Bayesian optimization algorithm, and gradually optimizing an objective function by continuously exploring and utilizing the parameter space to find the parameter configuration with the best performance of the model on the verification set. According to the preset model optimization iteration number, the preset model optimization iteration number is the model iteration number manually input by people, the design can be performed according to the optimal model iteration number in the historical data, the step S401 is performed in an iteration mode, and in each iteration, parameter adjustment is performed by using different target model verification data so as to gradually improve the model performance. And stopping iterative execution when the number of times of executing the step S401 reaches the preset model optimization iteration number. At this time, a training color change analysis model subjected to multiple optimization adjustment is obtained.

Preferably, step S5 comprises the steps of:

As an example of the present invention, referring to fig. 4, a detailed implementation step flow diagram of step S5 in fig. 1 is shown, where step S5 includes:

in the embodiment of the invention, the analysis range is defined through a preset humidity data test interval and an irradiation spectrum wavelength test interval. For example, assume that the humidity test interval is 0% to 100%, and the irradiation spectrum wavelength test interval is 400 nm to 700 nm. The preset test intervals are transmitted to the optimized color change analysis model, and can be performed through interfaces or input parameters of the model, so that the model is ensured to know the humidity and the spectral range to be analyzed. And then, the model performs color change trend analysis by using the data of the test interval, considers the color change condition of the fabric under the conditions of different humidity and irradiation spectrum wavelength, and generates a three-dimensional color change trend graph.

In the embodiment of the invention, gradient calculation is performed on the generated three-dimensional color change trend graph so as to obtain trend graph gradient data. The purpose of the gradient calculation is to identify the region in the image where the color change is most significant or severe, so as to reveal the color sensitivity of the fabric under different humidity and spectrum conditions, and the gradient of each data point can be calculated by using a mathematical difference operation, and assuming that each data point in the three-dimensional trend chart represents the color value under the specific humidity and irradiation spectrum wavelength conditions, the change rate between adjacent data points, namely the gradient, can be obtained by performing the difference operation on the values. For example, for humidity and the irradiation spectrum wavelength respectivelyAnd->Is +.>And->The gradient between these two points can be obtained by the following calculation: />Wherein the function->Color value expressed as humidity and corresponding to the wavelength of the illumination spectrum,/->Gradient data represented as neighboring points.

in the embodiment of the invention, the zero crossing points of the first derivative are detected on the generated gradient map, namely, the positions of the gradient from positive value to negative value or from negative value to positive value are found, and the zero crossing points represent the positions where color change possibly exists in the gradient map. For the detected zero crossing point, the surrounding area is further analyzed to determine whether the surrounding area is a local extreme point of color change, which can be achieved by checking the change trend and the local minimum/maximum of the gradient, and the position information of the identified color change extreme point and the corresponding gradient value are saved as color change extreme value data.

in the embodiment of the invention, the three-dimensional color change trend graph is subjected to standardization processing to ensure that the influence of the change of different scales on PCA is relatively equal, and the method generally comprises the steps of standardizing data (namely, the data under the conditions of each humidity and irradiation spectrum wavelength) in columns to ensure that the mean value is zero and the variance is one. A Principal Component Analysis (PCA) algorithm is applied to calculate principal components of the three-dimensional color change trend graph, wherein the principal components are a group of orthogonal basis vectors, and the ordering of the principal components represents the contribution degree of the principal components to the total variance. The number of principal components selected to remain depends on the degree of dimension reduction required, and the original data is projected onto the selected principal components, creating a new feature space. The new feature vector is the color change feature data.

In the embodiment of the invention, the color change stability calculation formula is utilized to calculate the color change extreme value data and the color change characteristic data, which may include applying a formula to each sample or time point, taking into account the position and number of extreme points and the change trend of the characteristic data. The calculated result is color change stability data, and the data can be used for comprehensively quantifying the stability of color change under different conditions, so as to help understand the performance of the material under the conditions of humidity and illumination.

；

in the method, in the process of the invention,expressed as colour change stability data>Data quantity expressed as color change extremum data, < +.>Weight information expressed as hue value deviation, +.>Denoted as +.>Hue value variance data of the individual color variation extremum data,/->Weight information expressed as contrast deviation, +.>Denoted as +.>Color contrast variance data of the individual color variation extremum data, < ->Data amount expressed as color change characteristic data, +.>Denoted as +.>Hue value variance data of individual color variation characteristic data,/-, and the like>Denoted as +.>Color contrast variance data of the individual color change feature data.

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The color change characteristic analysis method of the high-moisture-conductivity fabric is characterized by comprising the following steps of:

step S4, including:

step S44, including:

step S445: performing model test on the optimized color change analysis model to be tested by using the fabric color change test set to generate an optimized color change analysis model;

2. The method for analyzing color change characteristics of high moisture conductive fabric according to claim 1, wherein the step S1 comprises the steps of:

3. The method for analyzing color change characteristics of high moisture conductive fabric according to claim 1, wherein the step S2 comprises the steps of:

4. The method for analyzing the color change characteristics of the high moisture-conductive fabric according to claim 2, wherein the step S3 comprises the steps of:

5. The method of analyzing color change characteristics of high moisture conductive fabric according to claim 1, wherein step S444 comprises the steps of:

6. The method for analyzing color change characteristics of high moisture conductive fabric according to claim 1, wherein the step S5 comprises the steps of:

7. The method for analyzing color change characteristics of a highly moisture conductive fabric according to claim 6, wherein the color change stability calculation formula in step S55 is as follows:

；

in the method, in the process of the invention,expressed as colour change stability data>Data quantity expressed as color change extremum data, < +.>Weight information expressed as hue value deviation, +.>Denoted as +.>Hue value variance number of each color change extremum dataAccording to (I)>Weight information expressed as contrast deviation, +.>Denoted as +.>Color contrast variance data of the individual color variation extremum data, < ->Data amount expressed as color change characteristic data, +.>Denoted as +.>Hue value variance data of individual color variation characteristic data,/-, and the like>Denoted as +.>Color contrast variance data of the individual color change feature data.

8. A color change characteristic analysis system of a highly moisture conductive fabric, for performing the color change characteristic analysis method of a highly moisture conductive fabric according to any one of claims 1 to 7, comprising:

the optimized color change analysis model building module is used for carrying out visible spectrum difference calculation on the visible spectrum data matrix under different humidity according to the standard visible spectrum data to generate a visible spectrum difference matrix; performing fabric color change analysis according to the visible spectrum difference matrix to generate a fabric color change matrix; establishing a mapping relation of color change analysis by utilizing a decision tree-regression algorithm to obtain an initial color change analysis model; dividing the fabric color change matrix into data, and respectively generating a fabric color change training set, a fabric color change verification set and a fabric color change test set; transmitting the fabric color change training set to an initial color change analysis model for model training, and generating a training color change analysis model; performing model verification evaluation on the training color change analysis model according to the fabric color change verification set to generate model verification evaluation data; performing model parameter optimization adjustment on the training color change analysis model according to a Bayesian optimization algorithm and model verification evaluation data to generate an optimized color change analysis model to be tested; performing model test on the optimized color change analysis model to be tested by using the fabric color change test set to generate an optimized color change analysis model;