Method for generating prediction models for the perception of tactile roughness of fibers based on physical properties, method and apparatus for objectively evaluating the tactile roughness perception of target fiber for analysis

Abstract

The present invention relates to a method for generating a model for predicting the perception of tactile roughness of a fiber. The method for generating a model for predicting the perception of tactile roughness of a fiber according to the present invention comprises: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting valid parameters from a plurality of parameters constituting the physical characteristic information; and a step of generating a model for predicting the perception of tactile roughness by training a plurality of regression analysis models using the perception grade information and the valid parameters, and evaluating the performance of each of the trained plurality of regression analysis models.

Description

Technology Field

[0001] The present invention relates to a method for generating a model for predicting the perception of tactile roughness of a fiber, a method for objectively evaluating the perception of tactile roughness of a fiber to be analyzed, and an apparatus. More specifically, the invention relates to a method for generating a model for predicting the perception of tactile roughness of a fiber that can predict human roughness perception of various fibers based on the physical characteristics of the fiber, and an apparatus for objectively evaluating the perception of tactile roughness of a fiber using the generated model for predicting the perception of tactile roughness of a fiber. Background Technology

[0002] Generally, human tactile perception occurs through mechanoreceptors in the skin. This tactile perception refers to the process by which humans detect the characteristics of an object's surface and perceive tactile properties such as roughness, smoothness, softness, hardness, coldness, warmth, elasticity, stiffness, stickiness, and dryness through mechanoreceptors distributed in the skin. Such tactile perception capabilities play a key role in the evaluation of textile products.

[0003] Recently, expert evaluation methods have been widely used in the textile industry to quantify human tactile perception. These expert evaluation methods involve assigning scores based on specific criteria or expressing them in concrete language. While these methods have the advantage of enabling in-depth evaluations by utilizing the experience and senses of experts, they have the limitation that it is difficult to generalize evaluation results because the outcomes may vary depending on the individual evaluator's experience and tendencies.

[0004] In order to overcome the limitations of the aforementioned expert evaluation method, various studies have been conducted to link the physical properties of fibers, such as tensile strength, with human tactile perception.

[0005] For example, the Kawabata Evaluation System (KES-F) is known to be capable of measuring physical properties of fibers, such as significant tensile strength, shear, bending, compression, roughness, and surface characteristics. However, while the Kawabata System has the advantage of being able to quantify the physical properties of fibers, it has the disadvantage of being difficult to quantify tactile roughness, such as softness, which is actually perceived by humans. In particular, since the relationship between the physical properties of fibers and perceptual tactile sensation is known to exhibit non-linear characteristics, there is a problem in that it is difficult to predict human perception of roughness using only data that simply measures physical properties, as is the case with the Kawabata System; therefore, research is needed on methods to complement this. Prior art literature

[0006] Korean Registered Patent No. 10-0907760 "Emotion-based Fabric Indexing Automation System" The problem to be solved

[0007] The present invention aims to provide technical content regarding a tactile roughness perception prediction model that measures various physical properties of fibers, such as surface roughness, tensile strength, friction, and hardness, and based on this, can quantify human tactile roughness perception, which varies greatly from person to person, into a generalized form through objective evaluation. means of solving the problem

[0008] A method for generating a tactile roughness perception prediction model of a fiber according to an embodiment may include: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting valid parameters from a plurality of parameters constituting the physical characteristic information; and a step of generating a tactile roughness perception prediction model by training a plurality of regression analysis models using the perception grade information and the valid parameters, and evaluating the performance of each of the trained plurality of regression analysis models.

[0009] According to one embodiment, the plurality of fiber samples may each be collected from at least one of natural fibers and synthetic fibers.

[0010] According to one embodiment, the physical feature information may include at least one parameter among the average roughness (Ra), root mean roughness (Rq), maximum height roughness (Rz), average profile element width (RSm), asymmetry (Rsk), kurtosis (Rku), root mean slope (Rsq), maximum load (ML), elongation at break (EB), static friction coefficient (SF), kinetic friction coefficient (KF), and indentation hardness (SH) of each of the plurality of fiber samples.

[0011] According to one embodiment, the perception grade information can be calculated by a method comprising: a step of classifying the plurality of fiber samples into a plurality of grades according to physical roughness; a step of generating a fiber sample book by physical roughness for the plurality of fiber samples by ranking at least two types of fiber samples classified into the same grade among the plurality of grades according to physical roughness; and a step of performing a subjective evaluation of the tactile roughness of each of the plurality of fiber samples included in the fiber sample book from a plurality of testers, and by performing the step of performing the subjective evaluation at least twice or more, the average value of the subjective evaluation values ​​for the fiber samples can be calculated as the perception grade information.

[0012] According to one embodiment, the step of selecting the valid parameters may involve analyzing the multicollinearity of the physical feature information to calculate the parameter-specific variance expansion factor (VIF) for a plurality of parameters included in the physical feature information, removing parameters whose parameter-specific variance expansion factor exceeds a preset reference value, and selecting parameters whose variance expansion factor is within the preset reference value as the valid parameters.

[0013] According to one embodiment, the effective parameters may include at least one of the breaking elongation (EB), maximum load (ML), asymmetry (Rsk), indentation hardness (SH), root mean slope (Rsq), average profile element width (RSm), kurtosis (Rku), average roughness (Ra), and kinetic friction coefficient (KF) of each of the plurality of fiber samples.

[0014] According to one embodiment, the plurality of regression analysis models may include at least one analysis model among a Gaussian process regression model, a linear regression model, a support vector regression model, a random forest model, a gradient boosting model, and a neural network regression model.

[0015] According to one embodiment, the step of selecting the tactile roughness perception prediction model may be to evaluate the performance of the plurality of regression analysis models using 5-fold cross-validation and select the regression analysis model with the highest coefficient of determination as the tactile roughness perception prediction model.

[0016] According to one embodiment, the tactile roughness perception prediction model may include a Gaussian process regression model as the optimal model.

[0017] A method for evaluating the perception of tactile roughness of a fiber according to an embodiment may include: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting a first valid parameter from a plurality of parameters constituting the physical characteristic information; a step of training a plurality of regression analysis models using the perception grade information and the first valid parameter, and selecting a tactile roughness perception prediction model by evaluating the performance of each of the trained plurality of regression analysis models; a step of collecting a second valid parameter for a fiber to be analyzed; and a step of applying the second valid parameter to the tactile roughness perception prediction model to calculate a tactile roughness perception value of the fiber to be analyzed.

[0018] Meanwhile, the tactile roughness perception analysis device of a fiber according to an embodiment may include an analysis module that, when valid parameters related to the physical characteristics of a fiber to be analyzed are input, applies the valid parameters to a tactile roughness perception prediction model to generate a tactile roughness perception value of the fiber to be analyzed, wherein the tactile roughness perception model may be generated by a method comprising: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting valid parameters from a plurality of parameters constituting the physical characteristic information; and a step of training a plurality of regression analysis models using the perception grade information and the valid parameters, and evaluating the performance of each of the trained plurality of regression analysis models to select a tactile roughness perception prediction model.

[0019] According to one embodiment, the tactile roughness perception analysis device of a fiber according to the embodiment may further include an information collection module that collects and transmits valid parameters regarding the physical characteristics of the fiber to be analyzed. Effects of the invention

[0020] The method for generating a tactile roughness perception prediction model according to the embodiment can generate a tactile roughness perception prediction model capable of objective and generalized tactile evaluation in the textile industry by using data measuring physical properties such as surface roughness, tensile strength, friction, and hardness of the fiber to objectify and generalize human tactile perception, which varies greatly from person to person.

[0021] The tactile roughness perception prediction model generated by the above method can complement existing evaluation methods that relied on the personal and subjective evaluations of experts, and can provide universal, objective, and generalized evaluation criteria in the quality control and evaluation process of textile products.

[0022] In addition, the above-mentioned tactile roughness perception prediction model enables the consideration of consumers' tactile preferences from the design stage of textile products, thereby enhancing efficiency in establishing the product development process. Brief explanation of the drawing

[0023] FIG. 1 is a process diagram illustrating a method for generating a tactile roughness perception prediction model of a fiber based on the physical properties of the fiber according to an embodiment. FIG. 2 is a process diagram showing an objective tactile roughness perception evaluation method of a fiber to be analyzed according to an embodiment. FIG. 3 is a schematic diagram showing an objective tactile roughness evaluation device for a fiber to be analyzed according to an embodiment. Figure 4 is an image of a fiber sample book by physical roughness made using a plurality of fiber samples utilized to classify the roughness of fibers by grade according to an embodiment. Figure 5 is the result of performing a dispersion expansion coefficient analysis considering multicollinearity for individual physical feature information according to an embodiment. Figure 6 is the result of evaluating the performance of a plurality of regression analysis models on the correlation between the physical properties of the fiber and the objective tactile roughness perception evaluated by a tactile roughness perception prediction model according to an embodiment. Specific details for implementing the invention

[0024] Specific structural or functional descriptions of embodiments according to the concept of the present invention disclosed herein are provided merely for the purpose of explaining embodiments according to the concept of the present invention, and embodiments according to the concept of the present invention may be implemented in various forms and are not limited to the embodiments described herein.

[0025] Embodiments according to the concept of the present invention may be subject to various modifications and may take various forms; therefore, embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes modifications, equivalents, or substitutions that fall within the spirit and scope of the present invention.

[0026] Terms such as "first" or "second" may be used to describe various components, but said components should not be limited by said terms. For the sole purpose of distinguishing one component from another, for example, without departing from the scope of rights according to the concept of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0027] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. Conversely, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between. Expressions describing the relationships between components, such as "between," "exactly between," or "directly adjacent to," should be interpreted in the same way.

[0028] The terms used herein are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as “comprising” or “having” are intended to specify the existence of the described features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0029] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.

[0031] Hereinafter, embodiments will be described in detail with reference to the attached drawings. However, the scope of the patent application is not limited or restricted by these embodiments. Identical reference numerals in each drawing indicate identical components.

[0033] In general, roughness in textile products is a key factor determining consumers' purchasing decisions and satisfaction, and it is the attribute that has the most direct impact on the perception of product quality.

[0034] The method for generating a prediction model for the perception of tactile roughness of a fiber based on the physical properties of the fiber according to the embodiment can generate a prediction model capable of evaluating the sensations provided by the fiber. Even without consumers directly touching and evaluating the fiber, the product supplier can provide reference information for indicating the fiber product as a quantified value or grade.

[0035] In particular, the objective tactile roughness perception evaluation method of a fiber subject to analysis according to the embodiment, unlike the existing subjective evaluation method in which a person evaluates the roughness level while touching a fiber sample, utilizes a tactile roughness perception prediction model that can objectively evaluate the tactile roughness perception of a fiber subject to analysis, thereby enabling the tactile roughness perception of an unknown fiber requiring analysis to be calculated as a quantitative value.

[0037] FIG. 1 is a process diagram illustrating a method for generating a tactile roughness perception prediction model of a fiber based on the physical properties of the fiber according to an embodiment.

[0038] Referring to FIG. 1, a method for generating a tactile roughness perception prediction model of a fiber according to an embodiment may include: a step of collecting physical properties data for each of a plurality of fiber samples (S110); a step of collecting perception grade information for tactile roughness of each of the plurality of fiber samples (S120); a step of selecting valid parameters from a plurality of parameters constituting the physical properties data (S130); and a step of selecting a tactile roughness perception prediction model (S140).

[0039] First, in the step of collecting physical characteristic information (S110), physical characteristic information for each of the multiple fiber samples can be collected by evaluating the physical characteristics of the fiber samples.

[0040] Specifically, multiple fiber samples may each exhibit unique physical characteristics depending on the material used, processing method, weaving method, etc. Each physical characteristic of the fiber samples can be utilized as an indicator representing the texture, wearability, and comfort of the fiber samples.

[0041] The plurality of fiber samples may include at least one of natural fibers and synthetic fibers. The plurality of fiber samples may be fabric.

[0042] The above physical characteristic information may include a plurality of parameters. The above physical characteristic information may include at least one of the following parameters for an individual fiber sample: average roughness (Ra), root mean square roughness (Rq), maximum height of profile (Rz), mean width of profile elements (RSm), skewness (Rsk), kurtosis (Rku), root mean square slope (Rsq), maximum force (ML), breaking elongation (EB), static friction (SF), kinetic friction (KF), and shore hardness (SH).

[0043] The above parameters can be collected by evaluating each fiber sample using various conventional methods utilized in the textile industry.

[0044] Next, in the step of collecting perception grade information regarding tactile roughness (S120), perception grade information can be collected by performing an individual evaluation on the tactile roughness of each of the plurality of fiber samples.

[0045] According to one embodiment, in this step, a plurality of testers may be selected, and the selected plurality of testers may touch each of the fiber samples or rub the fiber samples against parts of the plurality of testers' bodies, and then the perceived sensation may be classified by grade based on standard perception grade information to collect perception grade information. In addition, in this step, a signal regarding tactile roughness transmitted from the brain connected to the plurality of testers may be recognized, and the intensity of the signal may be classified by grade to generate and collect perception grade information.

[0046] More specifically, in this step, a plurality of fiber samples are classified into a plurality of grades according to physical roughness, and at least two types of fiber samples classified into the same grade among the plurality of grades are ranked according to physical roughness to generate a fiber sample book by physical roughness for the plurality of fiber samples. At this time, the grades of the fiber samples may be classified into grades 1 to 5, but the grades of the fiber samples may be classified selectively as needed.

[0047] In addition, subjective evaluations of the tactile roughness of each of the multiple fiber samples included in the fiber sample book are performed by multiple testers to calculate subjective evaluation values ​​for each of the multiple fiber samples. Furthermore, the step of performing the subjective evaluation is performed at least twice to derive the average value of the subjective evaluation values ​​for each fiber sample, which can then be calculated as the perception grade information.

[0048] Alternatively, in this step, each fiber sample may be touched or the fiber samples may be rubbed using a tactile identification device comprising a robot hand equipped with at least one of a tactile sensor, a roughness sensor, a pressure sensor, and a friction sensor, and then the recognized signal may be classified by reference grade to collect recognition grade information.

[0049] The above perception grade information may be the average value of subjective evaluation values ​​collected by performing a process at least once in which a fiber sample is touched or a fiber sample is rubbed against a part of a plurality of testers' bodies, and then a subjective evaluation of the tactile roughness of each of the plurality of fiber samples is performed based on the perceived sensation. That is, in this step, the tactile roughness may be evaluated repeatedly multiple times using a tester or a tactile identification device, and the average value of the evaluated tactile roughness may be generated as perception grade information.

[0050] Next, in the step of selecting valid parameters (S130), valid parameters that are evaluated as suitable for tactile roughness evaluation are selected from among the plurality of parameters constituting the physical feature information.

[0051] In this step, physical characteristic information collected from the multiple fiber samples is analyzed using multicollinearity to calculate the Variance Inflation Factor (VIF) for each parameter. Then, valid parameters are selected based on preset reference values.

[0052] Specifically, in this step, a variance expansion coefficient is calculated for multiple parameters of the physical feature information using multicollinearity, and parameters exceeding a preset threshold value among the calculated variance expansion coefficients for each parameter are removed. Then, parameters whose variance expansion coefficient is less than or equal to the preset threshold value can be selected as valid parameters. The preset threshold value may be 10, but is not limited thereto and can be selectively adjusted.

[0053] That is, multicollinearity analysis is performed so that when the correlation between parameters among the multiple parameters included in the physical feature information is too high, the regression estimation becomes unstable and it can make it difficult to predict the perception of tactile roughness, so parameters with multicollinearity smaller than a preset threshold value are calculated as valid parameters.

[0054] At this time, the coefficient of variance expansion can be calculated using the coefficient of determination by performing regression analysis on each parameter using all other parameters. When the coefficient of variance expansion is 1, the parameter has no correlation with other parameters, and if it exceeds a preset threshold value, multicollinearity is high, which may lead to inaccurate results in the regression analysis.

[0055] For example, the effective parameters may include at least one of the following: elongation at break (EB), maximum load (ML), asymmetry (Rsk), indentation hardness (SH), root mean slope (Rsq), average profile element width (RSm), kurtosis (Rku), average roughness (Ra), and kinetic friction coefficient (KF) of each of the plurality of fiber samples.

[0056] Next, the step of selecting a tactile roughness perception prediction model (S140) involves selecting a tactile roughness perception prediction model using machine learning based on a dataset containing the perception grade information and the valid parameters.

[0057] Specifically, in this step, multiple regression analysis models are trained using the above-mentioned perception grade information and the above-mentioned valid parameters. Then, the performance of each of the trained multiple regression analysis models is evaluated. Then, the regression analysis model confirmed to have the highest coefficient of determination as a result of evaluating the performance of the regression analysis models is determined as the optimal regression analysis model and selected as the tactile roughness perception prediction model.

[0058] In this step, multiple regression analysis models are trained using the above cognitive rating information and the above valid parameters. At this time, the multiple regression analysis models may include at least one analysis model among Gaussian process regression, linear regression, support vector regression, random forest, gradient boosting, and neural network regression, and may additionally include various analysis models without being limited thereto.

[0059] Then, evaluate the performance of each trained regression model, and among the regression models evaluated, the coefficient of determination (R²) 2 The regression analysis model with the highest value can be calculated as the optimal regression analysis model. In this step, the performance of each regression model can be evaluated through 5-fold cross-validation.

[0060] In this step, when the effective parameters are elongation at break (EB), maximum load (ML), asymmetry (Rsk), indentation hardness (SH), root mean slope (Rsq), average profile element width (RSm), kurtosis (Rku), average roughness (Ra), and kinetic friction coefficient (KF), the Gaussian process regression model is identified as the optimal regression analysis model and can be selected as the prediction model for the perception of tactile roughness of the fiber.

[0061] The above-described tactile roughness perception prediction model of a fiber can collect physical feature information for an unknown fiber sample to be analyzed, generate tactile roughness perception information for the fiber sample to be analyzed using the collected physical feature information, and transmit the generated information to an administrator or management server.

[0062] The above-mentioned tactile roughness perception prediction model of the fiber can be utilized to collect information on the physical characteristics of the fiber and generate objective information on the tactile roughness of various fibers.

[0064] Meanwhile, FIG. 2 is a process diagram showing a method for evaluating the perception of tactile roughness of a fiber to be analyzed according to an embodiment.

[0065] Referring to FIG. 2, a method for evaluating the perception of tactile roughness of a fiber to be analyzed according to an embodiment may include: a step of collecting physical characteristic information for each of a plurality of fiber samples (S210); a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples (S220); a step of selecting a first valid parameter from a plurality of parameters constituting the physical characteristic information (S230); a step of selecting a tactile roughness perception prediction model (S240); a step of collecting a second valid parameter for the fiber to be analyzed (S250); and a step of calculating the tactile roughness perception value of the fiber to be analyzed (S260).

[0066] The above steps S210, S220, S230, and S240 are identical to the method for generating the tactile roughness perception prediction model described above, so a detailed explanation thereof will be omitted.

[0067] The step of collecting the second effective parameter (S250) collects the second effective parameter for the fiber to be analyzed. The second effective parameter can be collected by evaluating a physical characteristic identical to the first effective parameter among the physical characteristics of the fiber to be analyzed.

[0068] In the step (S260) of calculating the tactile roughness perception value, the tactile roughness perception value of the fiber to be analyzed can be calculated by applying the second valid parameter to the tactile roughness perception prediction model. The tactile roughness perception value is calculated by the prediction model and can be defined as a value that expresses the perception grade of the sensation perceived when a user touches or rubs the fiber to be analyzed as an objective numerical value.

[0069] According to one embodiment, the method for generating a prediction model of tactile roughness perception of a fiber based on the physical characteristics and the method for evaluating the tactile roughness perception of a fiber to be analyzed may be implemented in the form of a computer-readable recording medium that records a program for executing each of the methods separately on a computer.

[0070] The above program may include code coded in various computer languages ​​that can be executed by a processor of a computer or server. The code may include code such as functions that define features necessary to predict sleep disorders, and control code capable of controlling them.

[0071] In addition, the recording medium and memory on which the above program is stored are each media readable by a computer or server, such as ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical disk, USB memory, SD card, micro SD card, etc., but are not accepted therefrom.

[0072] In addition, the method for generating a prediction model for the perception of tactile roughness of a fiber and the method for evaluating the perception of tactile roughness of a fiber to be analyzed are preferably implemented as a single hardware device, but depending on the need, they may be implemented as an embedded device that is housed in an existing hardware device or as an application that is downloaded and installed in the form of software.

[0074] Meanwhile, FIG. 3 is a schematic diagram showing a device for evaluating the tactile roughness of a fiber according to an embodiment.

[0075] Referring to FIG. 3, the tactile roughness perception evaluation device of a fiber according to an embodiment has a structure equipped with an analysis module (10) including the tactile roughness perception prediction model.

[0076] When valid parameters related to the physical characteristics of the fiber to be analyzed are input into the analysis module (10), the analysis module (10) can generate a tactile roughness perception value of the fiber to be analyzed by applying the valid parameters to the tactile roughness perception prediction model. The tactile roughness perception value can be displayed by dividing the tactile roughness perception of the fiber according to the grade value.

[0077] The above analysis module (10) may be implemented in the form of a memory in which a program for evaluating the tactile roughness perception of a fiber to be analyzed is stored. Additionally, the memory for configuring the above analysis module (10) may store input / output information and may store a number of applications or information for applications, operations, and commands that are run on the evaluation device. At least some of these applications may be downloaded from an external server via wireless communication.

[0078] The above tactile roughness perception prediction model can quantitatively calculate the tactile roughness perception value felt by a user for the fiber under analysis by using physical characteristic information of the fiber under analysis or valid parameters extracted from the physical characteristic information. The above tactile roughness perception prediction model may be one generated using the method described above.

[0079] Additionally, the tactile roughness perception evaluation device of a fiber according to the embodiment may further include an information collection module (30) for collecting valid parameters from the fiber to be analyzed. The information collection module (30) can be implemented using various conventional analysis devices used to evaluate the physical characteristics of the fiber.

[0080] And, the tactile roughness perception evaluation device of a fiber according to the embodiment may further include a terminal module (50) that communicates with the analysis module.

[0081] The terminal module (50) receives the tactile roughness perception value of the fiber to be analyzed calculated by the analysis module (10), and outputs the received tactile roughness perception value so that the user can intuitively perceive the tactile roughness perception value of the fiber to be analyzed.

[0082] The above terminal module (50) can be implemented using a computer system such as a desktop, laptop, smartphone, handheld PC, etc.

[0083] The analysis module (10) and the terminal module (50) can communicate with each other through a communication network. The above communication network may include, but is limited to, backbone networks such as X.25 networks, Frame Relay networks, ATM networks, MPLS (Multi-Protocol Label Switching) networks, and GMPLS (Generalized Multi-Protocol Label Switching) networks, and subscriber networks such as FTTH (Fiber To The Home), ADSL (Asymmetric Digital Subscriber Line), cable networks, Zigbee, Bluetooth, Wireless LAN (IEEE 802.11b, IEEE 802.11a, IEEE 802.11g, IEEE 802.11n), Wireless Hart (ISO / IEC 62591-1), ISA 100.11a (ISO / IEC 62734), CoAP (Constrained Application Protocol), MQTT (Message Queuing Telemetry Transport), WIBro (Wireless Broadband), WiMAX, 3G, HSDPA (High Speed ​​Downlink Packet Access), 4G, 5G, 6G, etc. This is not the case. For example, the above communication network may be an internet network or a mobile communication network, and may be a wireless communication or wired communication method.

[0084] Additionally, although the terminal module (50) is depicted as being separate from the analysis module (10), it is not limited thereto and may be implemented as a single device as needed.

[0086] The method for generating a tactile roughness perception prediction model according to the embodiment described above can generate a tactile roughness perception prediction model capable of objective and generalized tactile evaluation in the textile industry by using data measuring physical properties such as surface roughness, tensile strength, friction, and hardness of the fiber to objectify and generalize human tactile perception, which varies greatly from person to person.

[0087] In addition, the generated tactile roughness perception prediction model can complement existing evaluation methods that relied on the personal evaluation of experts, thereby providing universal and generalized evaluation criteria in the quality control and evaluation process of textile products.

[0088] In addition, the above-mentioned tactile roughness perception prediction model enables the consideration of consumers' tactile preferences from the design stage of textile products, thereby enhancing efficiency in establishing the product development process.

[0089] In addition, the method and apparatus for evaluating the perception of tactile roughness of a fiber subject to analysis according to the embodiment are expected to contribute to the improvement of quality and product development of various fiber products, such as clothing, bedding, and furniture fibers, by utilizing them as objective roughness evaluation tools in the quality control and development process of fiber products.

[0091] The present invention will be explained in more detail below with reference to examples.

[0092] The presented embodiments are merely specific examples of the invention and are not intended to limit the technical scope of the invention.

[0093] < Examples >

[0094] (1) Materials

[0095] FIG. 4 is an image of a fiber sample book by physical roughness produced using multiple fiber samples utilized to classify the roughness of fibers by grade according to an embodiment.

[0096] Referring to Fig. 4, to classify fiber samples by roughness grade, 50 types of synthetic fibers composed of materials widely used in daily life, such as polyester, spandex, and nylon, were used as fiber samples. All fiber samples were prepared in the same size.

[0097] The above fiber sample book by physical roughness was produced in the following manner.

[0098] First, multiple fiber samples were prepared and classified into multiple grades based on physical roughness. Next, fiber samples classified into the same grade were ranked according to their physical roughness and arranged within the corresponding grades in order of rank to generate a fiber sample book categorized by physical roughness for the multiple fiber samples.

[0099] (2) Measurement of physical properties

[0100] To increase the reliability of the measurement and minimize the error of individual measurements, all physical property measurements were repeated 10 times for each fiber sample, and the average of the measurements was used.

[0101] Twelve physical characteristics corresponding to four categories were measured as follows.

[0102] i) Surface roughness

[0103] The surface roughness of the fiber was measured using an Alpha-Step IQ (KLA Tencor, USA) instrument. The sampling rate was set to 200 Hz, the scan distance to 5,000 μm, and the scan speed to 100 μm / sec. From the measured roughness profile, features were extracted including the average roughness (Ra), which represents the average absolute value of the height calculated from the roughness curve according to ISO 4287; the root mean square roughness (Rq), which represents the square root of the mean square of the height deviation of the roughness curve; the maximum height of profile (Rz), which represents the height difference between the highest peak and the lowest valley; the mean width of profile elements (RSm), which represents the average spacing between peaks of the roughness curve; the skewness (Rsk), which represents the asymmetry of the roughness distribution; the kurtosis (Rku), which represents the sharpness of the roughness distribution; and the root mean square slope (Rsq), which represents the average slope of the surface profile.

[0104] ii) Tensile strength

[0105] The tensile strength of the fibers was measured using a tensile testing machine (Instron, USA). The tensile strength of the fibers was calculated by measuring the maximum force (ML) and breaking elongation (EB) under conditions of a tensile speed of 100 mm / min.

[0106] iii) coefficient of friction

[0107] The friction coefficient of the fiber was measured using a friction coefficient tester (QMESYS, South Korea). The friction coefficient of the fiber was calculated by measuring the static friction (SF) and kinetic friction (KF) respectively under conditions of a load cell of 1 kgf, a moving speed of 200 mm / min, and a frictional weight of 200 g.

[0108] iv) Hardness

[0109] The hardness of the fibers was measured using a rubber durometer (REED Instruments, USA). The hardness characteristics of the fiber samples were calculated by fixing the durometer to a base and measuring the shore hardness (SH) at various points.

[0110] (3) Perform cognitive experiment

[0111] To generalize the perception of tactile roughness of fiber samples, a tactile roughness perception experiment was conducted on 50 fiber samples with 37 subjects. To reduce perceptual differences among subjects, the area for tactile sensation was limited to the terminal segment of the right index finger, and the direction of movement of the index finger was instructed to move only to the left and right of the fiber samples.

[0112] The tester was instructed to divide the presented samples into five levels according to their degree of roughness, with higher levels indicating rougher samples.

[0113] (4) Data Preprocessing

[0114] Among the 12 measured physical characteristics, a Variance Inflation Factor (VIF) analysis was performed considering multicollinearity, and the results are shown in Table 1 below. The VIF is an indicator that quantitatively evaluates the correlation between independent parameters; since high correlations between independent parameters can lead to unstable estimation of regression coefficients and degrade the predictive performance of the model, this analysis was performed to enhance the accuracy and stability of the prediction model. Generally, a VIF value exceeding 10 indicates that the corresponding parameter has a high correlation with other parameters.

[0115]

[0116] As a result of the VIF analysis, it was confirmed that the elongation at break (EB), maximum load (ML), asymmetry (Rsk), shore hardness (SH), root mean slope (Rsq), average profile element width (RSm), and kurtosis (Rku) showed acceptable VIF values.

[0117] However, the static friction coefficient (SF), kinetic friction coefficient (KF), maximum height roughness (Rz), average roughness (Ra), and root mean roughness (Rq) were found to exhibit severe multicollinearity.

[0118] Additional correlation analysis was performed on parameters with VIF values ​​exceeding 10, and the results are shown in Figure 5.

[0119] As shown in Figure 5, there was a high correlation between surface roughness-related indices such as maximum height roughness (Rz), average roughness (Ra), and root mean roughness (Rq), and it was also confirmed that there was a high correlation between friction-related indices such as static friction coefficient (SF) and kinetic friction coefficient (KF).

[0120] Based on the analysis results above, among the parameters with multicollinearity, the average roughness (Ra), which is most widely used in previous studies, and the kinetic friction coefficient (KF), which is directly related to the rubbing motion during actual tactile evaluation, were selected as representative features.

[0121] The nine finally selected features were normalized to standard scores (z-scores) and used for model development. Specifically, elongation at break (EB), maximum load (ML), asymmetry (Rsk), shore hardness (SH), root mean slope (Rsq), average profile element width (RSm), kurtosis (Rku), average roughness (Ra), and kinetic friction coefficient (KF) were finally selected.

[0122] (5) Model development and verification

[0123] The relationship between the physical properties of fibers and perceptual roughness was analyzed using six types of regression analysis models. The regression analysis models used were the linear regression model, the Gaussian process regression model, the support vector regression model, the random forest model, the gradient boosting model, and the neural network regression model.

[0124] The performance of each regression analysis model was evaluated through 5-fold cross-validation, and the coefficient of determination (R²) was used to calculate the predictive performance of the regression analysis model.

[0125] The coefficient of determination above ranges from 0 to 1 and serves as an indicator of how well the model explains the variance of the data; a value closer to 1 signifies higher prediction accuracy. The average of the results of five cross-validations was used as the final performance metric for each regression analysis model.

[0126] Using the above regression analysis models, the relationship between the physical properties of the fiber and the perception of tactile roughness was evaluated, and the results are shown in Figure 6. As a result, as shown in Figure 6, the Gaussian process regression analysis model was found to have the highest performance with a coefficient of determination of 0.71.

[0127] Through the results described above, it was confirmed that by measuring various physical properties of the fiber, such as surface roughness, tensile strength, friction, and hardness, and applying a tactile roughness perception prediction model to the measured physical properties, human tactile roughness perception can be quantitatively predicted based on the physical properties of the fiber.

[0128] In addition, the tactile roughness perception prediction model according to the embodiment suggested that objective and consistent evaluation is possible, moving away from the existing method that relied on the personal evaluation of experts.

[0129] Furthermore, it was determined that evaluation is possible even when other physical properties are collected in addition to the aforementioned physical properties (surface roughness, tensile strength, friction, and hardness), and thus it was concluded that it can be universally applied to a wider variety of fiber properties.

[0130] In addition, it was determined that the tactile roughness perception prediction model according to the example can be applied not only to the roughness of the fiber but also to the prediction of other tactile characteristics such as softness and elasticity.

[0131] As described above, the tactile roughness perception prediction model according to the embodiment is expected to be practically utilized in the textile industry to objectively evaluate and predict the tactile characteristics of products.

[0133] The device described above may be implemented as a hardware component, a software component, and / or a combination of a hardware component and a software component. For example, the device and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. In addition, other processing configurations, such as parallel processors, are also possible.

[0134] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0135] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0136] Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can make various modifications and variations from the description above. For example, suitable results can be achieved even if the described techniques are performed in a different order than described, and / or the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0137] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 A method for generating a tactile roughness perception prediction model of a fiber, comprising: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting valid parameters from a plurality of parameters constituting the physical characteristic information; and a step of training a plurality of regression analysis models using the perception grade information and the valid parameters, and evaluating the performance of each of the trained plurality of regression analysis models to generate a tactile roughness perception prediction model. Claim 2 In claim 1, the plurality of fiber samples are each collected from at least one of natural fibers and synthetic fibers. A method for generating a tactile roughness perception prediction model of fibers. Claim 3 A method for generating a tactile roughness perception prediction model of a fiber, wherein the physical feature information comprises at least one parameter among the average roughness (Ra), root mean roughness (Rq), maximum height roughness (Rz), average profile element width (RSm), asymmetry (Rsk), kurtosis (Rku), root mean slope (Rsq), maximum load (ML), elongation at break (EB), static friction coefficient (SF), kinetic friction coefficient (KF), and indentation hardness (SH) of each of the plurality of fiber samples. Claim 4 A method for generating a prediction model of tactile roughness perception of fibers according to claim 1, wherein the perception grade information comprises: a step of classifying the plurality of fiber samples into a plurality of grades according to physical roughness; a step of generating a fiber sample book by physical roughness for the plurality of fiber samples by assigning ranks according to physical roughness to at least two types of fiber samples classified into the same grade among the plurality of grades; and a step of performing a subjective evaluation of the tactile roughness of each of the plurality of fiber samples included in the fiber sample book by a plurality of testers, thereby calculating a subjective evaluation value for each of the plurality of fiber samples, and performing the step of performing the subjective evaluation at least twice or more to calculate the average value of the subjective evaluation values ​​for each of the fiber samples as the perception grade information. Claim 5 A method for generating a tactile roughness perception prediction model of a fiber, wherein the step of selecting valid parameters comprises analyzing the multicollinearity of the physical feature information to calculate the parameter-specific variance expansion coefficient (VIF) for a plurality of parameters included in the physical feature information, removing parameters whose parameter-specific variance expansion coefficient exceeds a preset reference value, and selecting parameters whose variance expansion coefficient is within the preset reference value as valid parameters. Claim 6 A method for generating a tactile roughness perception prediction model of a fiber, wherein the effective parameters include at least one of the breaking elongation (EB), maximum load (ML), asymmetry (Rsk), indentation hardness (SH), root mean slope (Rsq), average profile element width (RSm), kurtosis (Rku), average roughness (Ra), and kinetic friction coefficient (KF) of each of the plurality of fiber samples. Claim 7 A method for generating a prediction model for the perception of tactile roughness of a fiber, wherein the plurality of regression analysis models comprises at least one of a Gaussian process regression model, a linear regression model, a support vector regression model, a random forest model, a gradient boosting model, and a neural network regression model. Claim 8 A method for generating a tactile roughness perception prediction model of a fiber, wherein the step of selecting the tactile roughness perception prediction model is to evaluate the performance of the plurality of regression analysis models using 5-fold cross-validation and select the regression analysis model with the highest coefficient of determination as the tactile roughness perception prediction model. Claim 9 In claim 1, the tactile roughness perception prediction model is a method for generating a fiber tactile roughness perception prediction model including a Gaussian process regression model. Claim 10 A method for evaluating the perception of tactile roughness of a fiber to be analyzed, comprising: a step of collecting physical characteristic information for each of a plurality of fiber samples; a step of collecting perception grade information for the tactile roughness of each of the plurality of fiber samples; a step of selecting a first valid parameter from a plurality of parameters constituting the physical characteristic information; a step of training a plurality of regression analysis models using the perception grade information and the first valid parameter, and selecting a tactile roughness perception prediction model by evaluating the performance of each of the trained plurality of regression analysis models; a step of collecting a second valid parameter for a fiber to be analyzed; and a step of applying the second valid parameter to the tactile roughness perception prediction model to calculate a tactile roughness perception value of the fiber to be analyzed. Claim 11 A tactile roughness perception analysis device for a fiber generated by a method comprising: an analysis module that, when valid parameters related to the physical characteristics of a fiber to be analyzed are input, applies said valid parameters to a tactile roughness perception prediction model to generate a tactile roughness perception value of said fiber to be analyzed, wherein the tactile roughness perception prediction model comprises the steps of: collecting physical characteristic information for each of a plurality of fiber samples; collecting perception grade information regarding the tactile roughness of each of the plurality of fiber samples; selecting valid parameters from a plurality of parameters constituting the physical characteristic information; and training a plurality of regression analysis models using the perception grade information and the valid parameters, and evaluating the performance of each of the trained plurality of regression analysis models to select a tactile roughness perception prediction model.

Images