CN110413727A - Method of judging clothing style category through color gamut based on PCCS system - Google Patents

Abstract

The invention relates to a method for judging clothing style categories through color gamut based on the PCCS system, comprising the following steps: establishing clothing perceptual image space, analyzing consumers' perceptual evaluation dimensions of clothing, dividing clothing style categories accordingly, and determining clothing perceptual evaluation representatives Vocabulary; through network collection and research, extract representative clothing color samples, and make clothing sample pictures for perceptual evaluation experiments; use questionnaire survey method combined with semantic difference method to conduct perceptual evaluation experiments; analyze experimental data, and compare clothing perceptual imagery with PCCS tone area and NCD The color space is matched; a discriminant model is established to realize the judgment of the perceptual image of men's shirts through color design variables. On the one hand, the present invention can provide designers with color selection references when the target image is known; on the other hand, it can help designers better grasp the influence of different colors on consumers' perception and apply them flexibly.

Description

Method of judging clothing style category through color gamut based on PCCS system

technical field

The invention relates to the application of kansei engineering in the field of clothing, in particular to a method for judging clothing style categories through color gamut based on the PCCS system.

Background technique

According to Philip Kotler's three stages of consumer behavior, contemporary consumers have reached the third stage of purchasing decisions based on personal preferences - the stage of perceptual consumption, where the "perceptual value" of a product is more important than its "functional value" ". Desmet believes that product appearance is a stimulus that elicits emotional responses from users. The "seven-second law" of the American Fashion Color Research Association shows that people only need seven seconds to determine their likes and dislikes of products, and the role of color in its appearance design accounts for 67%.

At present, the research on color sensibility is most notable by Japanese scholars, such as the NCD color image space developed by Japan Color Research Institute. These research results are mostly used in industrial design, and the pertinence of clothing color is not enough. Domestic scholars have conducted research on the semantics of clothing style and color attributes or color matching, but no system has been formed. There is no systematic theoretical guidance for the color design of many clothing categories, and men's shirts are one of them.

Contents of the invention

The technical problem to be solved by the present invention is to provide a method for judging the clothing style category through the color gamut based on the PCCS system. On the one hand, it can provide a color selection reference for the designer when the target image is known. On the other hand, it can help the designer Better grasp the impact of different colors on consumers' perception and apply them flexibly.

The technical scheme adopted by the present invention to solve its technical problems is: provide a kind of method based on PCCS system to judge clothing style category by color gamut, comprise the following steps:

(1) Establish clothing perceptual image space, analyze consumers' perceptual evaluation dimensions of clothing, divide clothing style categories accordingly, and determine representative vocabulary for clothing perceptual evaluation;

(2) Extract representative clothing color samples through network collection and research, and make clothing sample pictures for perceptual evaluation experiments;

(3) Using the questionnaire survey method combined with the semantic difference method to conduct a perceptual evaluation test;

(4) Analyze the experimental data, and match the clothing perceptual image with the PCCS tone area and NCD color space;

(5) Establish a discrimination model to realize the judgment of the perceptual image of men's shirts through color design variables.

The step (1) is as follows: extensively collect and screen clothing perceptual evaluation vocabulary, use questionnaire survey method, card sorting method and cluster analysis of data statistics to establish perceptual image space of clothing and analyze consumers' perceptual evaluation dimensions of clothing .

The cluster analysis of the card sorting method and data statistics is as follows: prepare blank cards of the same size, and write N words to be classified and number them; invite several people to be tested, and conduct tests one by one; execute the test When testing, ask the subjects to think and put similar cards together according to their personal feelings. The number of groups is controlled at 6-10 groups to make the follow-up analysis results accurate and effective; the grouping frequency is counted, and an N×N similarity matrix is established to carry out systematic Cluster analysis obtained the pedigree chart to determine the number of groups, and the specified number of groups was clustered by K-means to obtain the specific grouping results.

The step (2) is specifically as follows: collect recent show photos and color plans of major clothing brands from professional fashion websites, extract the RGB attribute values of the main colors from the clothing-related pictures, and cluster the color samples to obtain the final experimental samples , and compared to get the PCCS hue and hue attribute values of the sample.

The step (3) uses the semantic difference method to carry out the evaluation experiment, and uses the Likert 5-level evaluation scale to combine the clothing sample image processed in the step (2) and the representative vocabulary of the clothing perceptual evaluation determined by the analysis in the step (1) , the experiment required the subjects to evaluate and score clothing samples of different colors according to their psychological feelings.

The step (4) is specifically: processing the data obtained from the perceptual evaluation experiment, taking the average value of all subjects' scores as the final score of the sample, performing factor analysis, and determining the influence factor of color on consumer perceptual evaluation; at the same time Match the perceptual image of the sample with the color range of PCCS to obtain the representative color gamut of various styles.

Matching the perceptual image of the sample with the color range of the PCCS to obtain representative color gamuts of various styles is specifically: analyzing the perceptual image of monochrome clothing and placing the results in the PCCS color system coordinate map; summarizing K types of clothing The range area of style color in the PCCS color system; the obtained PCCS hue distribution and position of K-type clothing are reflected in the NCD coordinate map, and the fuzzy areas of K-type clothing are uniformly positioned in the NCD space.

Described step (5) is specifically: carry out discriminant analysis to gained data, K/2 that obtains with factor analysis is to clothing image word pair as the dependent variable of discriminant analysis, and independent variable is hue and hue in the PCCS theoretical system, obtains classification function as a discriminant model.

The independent variables include two groups, one group is hue independent variable with 12 dummy variables in total, and the other group is hue independent variable with 11 dummy variables in total.

Beneficial effect

Due to the adoption of the above-mentioned technical solution, the present invention has the following advantages and positive effects compared with the prior art: the present invention classifies clothing styles by investigating and analyzing the perceptual evaluation dimensions of clothing; The color range selected by the clothing is matched with the PCCS color system and the NCD color space; in addition, a discrimination model is established for each style of clothing, and finally it is only necessary to substitute the PCCS hue and hue value of the clothing color to judge the attribution of the corresponding style category of the clothing Happening. Therefore, through the matching relationship between clothing color and perceptual image, on the one hand, it can provide designers with a reference for color selection when the target image is known, and on the other hand, it can help designers better grasp the impact of different colors on consumers' perception And applied to the actual design. At the same time, the biggest feature of the PCCS system selected by the present invention is that it integrates the three-attribute relationship of color into two concepts of hue and tone, which is more scientific, refined, fast and efficient than other color systems.

Description of drawings

Fig. 1 is the systematic clustering pedigree diagram of vocabulary to be classified;

Fig. 2 is the figure of the men's shirt style selected in the perceptual evaluation experiment;

Figure 3 is an example diagram of a male shirt sample in the perceptual evaluation experiment;

Figure 4 is a schematic diagram of the Likert five-level scale used in the perceptual evaluation experiment;

Figure 5A is a line chart of the average score of samples 1-11;

Fig. 5B is a line chart of the mean score of samples No. 12-22;

Fig. 5C is a line chart of the average scores of samples No. 23-33;

Fig. 5D is a line chart of the mean score of samples No. 34-44;

Figure 5E is a line chart of the average score of samples No. 45-55;

Fig. 5F is a line chart of the mean score of samples No. 56-66;

Figure 5G is a line chart of the average score of samples No. 67-77;

Fig. 6A is the corresponding figure of PCCS color coordinates of men's shirts in retro style;

Fig. 6B is the corresponding figure of PCCS color coordinates of the avant-garde style men's shirt;

Fig. 6C is a corresponding figure of PCCS color coordinates of low-key style men's shirts;

Fig. 6D is the corresponding figure of PCCS color coordinates of men's shirts with flamboyant style;

Figure 6E is a corresponding map of PCCS color coordinates of urban style men's shirts;

Fig. 6F is the corresponding map of PCCS color coordinates of rural style men's shirts;

Figure 7 is a summary map of the spatial regions corresponding to the PCCS color system for the six styles of men's shirt colors;

Figure 8 is the color gamut distribution map of the six styles of men's shirts in the NCD space;

Figure 9 is a schematic diagram of the sf12 recommended color for rural style men's shirts;

Figure 10 is a schematic diagram of color recommendation b2 for avant-garde men's shirts.

Detailed ways

Below in conjunction with specific embodiment, further illustrate the present invention. It should be understood that these examples are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the teachings of the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of the present application.

The embodiment of the present invention relates to a method for judging the clothing style category through the color gamut based on the PCCS system. Representative vocabulary for perceptual evaluation; extract representative clothing color samples through network collection and research, and make clothing sample pictures for perceptual evaluation experiments; use questionnaire survey method combined with semantic difference method to conduct perceptual evaluation tests; analyze experimental data, and compare clothing perceptual images with PCCS color The region and NCD color space are matched; a discriminant model is established to realize the judgment of the perceptual image of men's shirts through color design variables.

Take a men's shirt as an example below to further illustrate the present invention.

1. Establish a perceptual image space for men's shirts:

A large number of adjectives about the emotional evaluation of men's shirts were collected by reading literature and magazines, conducting Internet searches, and referring to exhibition copywriting. In order to avoid the lack of key words in the subsequent screening, the collection should cover as wide as possible, and the collection of related adjectives must also follow certain standards: First, the description object is clear. Men’s shirts are only a small branch of many clothing categories, and the collected words must It can properly and appropriately describe this specific object; second, limit the angle of description, and exclude words that only describe styles and fabric characteristics without involving the overall image; third, conform to aesthetics, and the collected words must meet the aesthetic needs of consumers There are mixed opinions; fourth, objective information words that have nothing to do with human perception should be excluded, and words that have nothing to do with clothing but only about the temperament of the wearer should also be excluded.

A total of 168 adjectives were collected and preliminarily screened. The questionnaire survey was used for further screening. The subjects selected from the 168 words that they thought were suitable for describing the visual experience of men’s shirts based on their actual experience. 1/3 of the vocabulary, finally selected 58 qualified adjectives.

Since there are still many adjectives after screening, it is necessary to classify the perceptual image vocabulary on this basis, and finally form the required perceptual image space. Using the card sorting method, the experimental process can be divided into the following four steps: 1) Prepare blank cards of the same size, and write and number 58 words to be classified one by one; 2) Invite 25 subjects to test each user; 3) Execute the experiment, ask the subjects to think and put similar cards together according to their personal feelings, and control the number of groups in 6-10 groups to make the follow-up analysis results accurate and effective; 4) Statistical grouping frequency, establish 58×58 similar cards The sex matrix was analyzed using SPSS 21.0.

First use the system clustering analysis method to obtain the pedigree diagram to determine the number of groups, as shown in Figure 1, it can be seen that the image vocabulary pairs are divided into 6-10 groups is more appropriate, combined with the progress table of the clustering process and the class membership table of the clustering results. After comprehensive consideration, it is finally determined that the image semantics can be divided into 8 groups. Then carry out K-means cluster analysis and set the target number of clusters to 8, and the clustering results are shown in Table 1. The comprehensive analysis concluded that the representative images in the group mean low-key, retro, mature, pastoral, avant-garde, young, publicity, and urban.

Table 1 Grouping table of imagery semantic vocabulary

The representative vocabulary of the obtained 8 groups is paired to form four pairs of antonyms: retro - avant-garde, urban - pastoral, mature - young, low-key - publicity, this The four word pairs reflect consumers' perceptual evaluation of four different dimensions of the style of printed men's shirts: time domain, region, wearer's temperament and personality, and they are used as image semantic representative samples to study the relationship between perceptual imagery and the color of men's shirts.

Second, determine the men's shirt sample:

From professional fashion websites such as WOW-Trend and POP Apparel Fashion Frontline, we collect the show photos and color schemes of major men’s clothing brands in the past five and ten quarters, and take the pictures related to men’s shirts as samples, and extract the samples through image software The RGB value of the color attribute of the men’s shirt in the picture finally obtained 615 colors. In order to reduce the workload, the obtained colors were clustered without affecting the semantic expression measurement and the color difference was allowed, and 77 final samples were obtained. Placed in the PCCS-RGB conversion table to get their hue and tint attribute values. Select the style of men’s shirts as shown in Figure 2, use PhotoshopCS6 to color render them, fill in 77 color samples in turn, and get 77 solid-color men’s shirts, as shown in Figure 3.

3. Using the questionnaire survey method combined with the semantic difference method to conduct a perceptual evaluation test:

In this embodiment, the semantic difference method is used to design the evaluation questionnaire, and the 5-level Likert scale is used to combine the processed sample pictures with the four pairs of semantic vocabulary determined above, and the samples are randomly shuffled in advance. The experiment required the subjects to rate the imagery of 77 solid-color men's shirts with different colors only according to their personal subjective feelings. The scores ranged from -2 to +2, which respectively reflected the tendency and degree of imagery expressed by the subjects. Taking the perceptual image word pair "retro-avant-garde" as an example, the evaluation scale is shown in Figure 4, and each subject conducts the experiment independently. According to "ISO6658 Sensory Analysis - General Guidelines for Methodology", the number of preferred evaluators for sensory evaluation should be more than 20. In this embodiment, 40 questionnaires were distributed in the experiment, and each subject was required to score four groups of image-vocabulary pairs on 77 samples, and scored 77×4=308 times in total.

4. Analyze the experimental data and match the perceptual image of the men's shirt with the PCCS tone area and NCD color space:

Statistical subjects score data to obtain the average of the four groups of vocabulary scores for each sample. Drawing a line graph shows the score distribution of each sample more intuitively. The highest or lowest point of each line indicates that this sample brings the most obvious feeling to people, as shown in Figure 5A-G.

In the formula: i is the serial number of the sample; j is the perceptual image vocabulary; n is the total number of samples.

After correlation analysis, it is found that the correlation coefficients of the four groups of adjective pairs studied in this example are generally high, and factor analysis can be carried out to further reduce the dimension. The results of the factor analysis are shown in Table 2. Finally, three factors were determined to be extracted: fashion factor, personality factor, and temperament factor. These six perceptual image semantics are regarded as six major style types of men's shirts affected by color design variables.

Table 2 Factor loading matrix after rotation

Based on the image scoring results of each sample, the six styles of men's shirts were matched with the PCCS color system and the NCD color image space.

First, analyze the perceptual image of the monochrome men's shirt and place the results in the PCCS color system coordinate diagram, as shown in Figure 6A-F. The range area of the six men's shirt style colors in the PCCS color system can be summarized, as shown in Figure 7. It can be seen that the colors of the six categories of men's shirts overlap more in achromatic, dark gray tones, bright tones and light tones, and other areas have their own characteristics although they overlap. According to the specific performance of the color area map, achromatic black, white and gray are eternal classics, mainly used in men's shirts in retro style, urban style and low-key style. Bright tones, bright tones and light tones are suitable for avant-garde and ostentatious styles, bright, enthusiastic and outstanding; while dark gray tones and light gray tones with low purity are suitable for low-key and retro styles. Each style area presents a fuzzy and semi-closed state, that is, the representative colors of each style have a certain tone tendency, and the styles overlap each other. With the change of fashion, the edge of the area will expand or shrink.

The obtained PCCS tone distribution and position of the six categories of men's shirts are reflected in the NCD coordinate map. Due to the fuzzy characteristics of the region boundaries, comparison and adjustment can be made appropriately, and finally the fuzzy regions of the six categories of men's shirts are uniformly positioned in the NCD space , as shown in Figure 8.

5. Establish a discrimination model to realize the judgment of the perceptual image of men's shirts through color design variables:

In order to obtain a model that can predict the category attributes of men's shirts, a discriminant analysis is performed on the obtained data, and three pairs of men's shirt image words obtained by factor analysis are used as the three dependent variables of the discriminant analysis, and the independent variables are hue and hue in the PCCS theoretical system , the classification function obtained through discriminant analysis, only need to substitute the hue and hue value of the men’s shirt color to judge its category attribution.

Since the dependent variable of the discriminant analysis must be a categorical variable, the three groups of variables are divided into three grades according to the image scoring results: the score of the word pair "retro - avant-garde" divides it into "retro", " Moderate" and "Avant-garde"; the same "low-key-public" can be divided into "avant-garde", "moderate" and "public"; "urban-rural" can be divided into "Urban", "moderate", "rural". Since the two independent variables "hue" and "hue" are categorical variables, they need to be treated as dummy variables. "Hue" includes "achromatic" and 12 hues, a total of 13 categories. Here, "achromatic" is used as a control group, and establish 12 dummy variables; "hue" extracts 12 main hues from the 24 hues of the PCCS hue circle, which are the psychological four primary colors "red, yellow, green, blue" and the complementary colors of the psychological four primary colors "blue-green, blue Purple, red-purple, yellow-orange" and the remaining equidistant divisions of "red-orange, yellow-green, green-blue, purple", with "red-purple" as the reference group, 11 dummy variables were set up.

After the variable processing is completed, SPSS can be used for discriminant analysis. It can be seen that the two independent variables "hue" and "hue" can effectively distinguish the three categories of each dependent variable; the total prediction accuracy reached: 85.7%, 87%, 79.2%, very good discrimination. The classification function coefficients are shown in Table 3, and the classification functions of the three groups of men's shirts can be written separately.

Table 3 Classification function coefficient table of three groups of men's shirts

For the first group of categories "retro - avant-garde", y=-1 means that the category classification function of "retro" is: f ₁₁ =-5.06+9.36v+9.165b+8.994s+10.354dp+9.353Lt ⁺ +10.284sf+10.558d+10.985dk+9.496p ⁺

+8.833Ltg+11.150g+10.932dkg-0.781 red-0.204 red orange-0.361 yellow orange-0.317 yellow+0.181 yellow-green-1.219 green-0.33 green-blue+0.037 blue-green-0.38 blue-0.197 blue-violet-0.426 purple

y=0, that is, the classification function of men's shirt category of "neither retro nor avant-garde" is:

f ₁₂ ＝-10.568+17.837v+19.012b+20.034s+11.877dp+17.882Lt ⁺ +12.295sf+10.649d+8.092dk +17.022p ⁺ +21.002Ltg+7.103g+8.405dkg+4.626red+1.2 +2.163 Yellow Orange+1.905 Yellow-1.086 Yellow Green+7.314 Green+1.978 Green Blue-0.223 Blue Green+2.283 Blue+1.182 Blue Purple+2.558 Purple

The classification function for y=1, i.e. "fashionable" category men's shirts is:

f ₁₃ ＝-14.125+26.356v+26.851b+21.186s+12.962dp+26.240Lt ⁺ +15.868sf+14.122d+9.017dk +18.857p ⁺ +20.209Ltg+10.288g+10.228dkg+3.461 red+0. +0.932 yellow orange+1.897 yellow-0.737 yellow-green+3.687 green-2.543 green blue-0.034 blue-green-4.216 blue-6.214 blue-violet+0.681 purple

For the "Unassuming-Expressive" category:

y=-1 means that the coefficient of the independent variable "hue" in the "low-key" category classification function is very small compared with the "hue" coefficient and constant, and the 12 digits after the decimal point are all zero, so it can be ignored, then

f ₂₁ =-4.904+10.571=5.667 (sample is not achromatic)

f ₂₁ =-4.904 (sample is achromatic)

y=0, that is, the classification function of men's shirt category of "neither low-key nor ostentatious" is:

f ₂₂ ＝-12.098+22.834v+22.701b+22.053s+18.413dp+21.74Lt ⁺ +18.407sf+17.476d+8.548dk+ 14.9p ⁺ +14.740Ltg+7.03g+9.217dkg+4.378 red+4.047 red orange 3.701 yellow orange + 4.643 yellow + 1.027 yellow green + 5.676 green + 3.035 green blue + 5.606 blue green + 1.129 blue - 1.365 blue purple + 0.309 purple

y = 1 i.e. the "flashy" category classification function is:

f ₂₃ ＝-20.297+37.409v+36.269b+26.378s+22.82dp+32.12Lt ⁺ +19.754sf+22.148d+8.612dk+1 7.783p ⁺ +17.441Ltg+9.765g+11.415dkg+8.359 red+3.9 Orange + 2.019 Yellow Orange + 6.556 Yellow - 0.225 Yellow Green + 6.065 Green - 2.306 Green Blue + 6.262 Blue Green - 5.937 Blue - 11.730 Blue Purple - 0.855 Purple

Similarly, write the "urban-rural" group:

y=-1, that is, the category classification function of "urban" is

f ₃₁ =-4.843+10.571=5.728 (sample is not achromatic)

f ₃₁ =-4.843 (sample is achromatic)

y=0, that is, the male shirt category classification function of "neither urban nor pastoral" is:

f ₃₂ ＝-10.896+14.563v+15.182b+17.953s+12.398dp+12.921Lt ⁺ +15.796sf+15.780d+19.711dk +13.442p ⁺ +16.603Ltg+10.755g+6.576dkg+4.9541 red+4. +3.922 Yellow Orange+2.436 Yellow+5.485 Yellow Green+4.78 Green+6.501 Green Blue+5.04 Blue Green+3.516 Blue+4.132 Blue Purple+0.123 Purple

y=1 i.e. the "pastoral" category classification function is:

f ₃₃ ＝-11.947+17.669v+17.29b+16.337s+13.827dp+15.864Lt ⁺ +20.762sf+14.818d+16.441dk +14.396p ⁺ +15.447Ltg+11.661g+7.827dkg+1.024red+3.9 +4.022 Yellow Orange+5.133 Yellow+6.529 Yellow Green+5.701 Green+1.706 Green Blue+7.795 Blue Green+0.208 Blue-0.454 Blue Purple-4.229 Purple

For each dependent variable, substitute the hue and hue data of the color of the men's shirt, and calculate the function values of the three classification functions of the dependent variable, and the category with the largest value is regarded as the category of the men's shirt.

In this embodiment, through research and analysis, men's shirts are divided into six categories according to three perceptual evaluation dimensions: retro, avant-garde, low-key, publicity, urban and pastoral. Through the perceptual evaluation experiment, the color range selected by the six categories of men’s shirts is matched with the PCCS color system and the NCD color space; in addition, the three major groups of the six categories of men’s shirts are paired to establish a discriminant model, and only need to be substituted into the men’s shirts The PCCS hue and hue value of the color can determine the attribution of the men's shirt to the six categories. To sum up, through the matching relationship between the color of men’s shirts and perceptual images, on the one hand, it can provide designers with a reference for color selection when the target image is known, and on the other hand, it can help designers better grasp the impact of different colors on consumption. The impact of the reader's perception and applied to the actual design. See Case 1 and Case 2 for specific application methods.

Case 1: Reference for color selection of rural style men's shirts

Customer H wants to order men’s shirts with a pastoral style. According to the summary map of the spatial regions of the six types of men’s shirt colors corresponding to the PCCS system, the common colors of the pastoral style men’s shirts are mainly concentrated in {sf+Lt ⁺ +b+v}; according to the six types of men’s shirt colors The color gamut distribution of men's shirts in the NCD space also shows that the softness, hardness, and warmth of the men's shirts in rural style are moderate; considering the matching of hue and hue, a qualified color collection can be obtained. Here is a reference color, which is intended to be The sf hue with medium chroma and medium lightness is recommended, the hue selection is 12: green, and the final recommended color is sf12, as shown in Figure 9.

In order to verify whether sf12 conforms to the pastoral style in consumers’ perceptual images, it is substituted into the category function of “urban – pastoral”, and the function values of three categories are obtained: f ₃₁ =-4.843+10.571=5.728; f ₃₂ =- 10.896+15.796+4.78=9.68; f ₃₃ =-11.947+20.762+5.701=14.516. It can be seen that f ₃₃ >f ₃₂ >f ₃₁ , that is, sf12 belongs to the category of "pastoral", and the recommendation is reasonable.

Case 2: Color selection reference for avant-garde style men's shirts

Designers want to design avant-garde men's shirts. When choosing colors, they can refer to the summary map of the spatial regions of the six types of men's shirt colors corresponding to the PCCS system. It is observed that the common colors of avant-garde men's shirts are mainly concentrated in {v+b+ Lt ⁺ +s}, combined with the color gamut distribution of the six styles of men's shirts in the NCD space, the color of the avant-garde style men's shirts transitions from the warm and hard area to the warm and soft area; the recommended b tone, hue selection 2: red, the final recommended color b2, as shown in Figure 10.

In order to verify whether b2 conforms to the pastoral style in consumers’ perceptual images, it is substituted into the “retro-avant-garde” category function, and the function values of the three categories are obtained: f ₁₁ =-5.06+9.165-0.781=3.324; f ₁₂ =-10.568+19.012+4.686=13.13; f ₁₃ =-14.125+26.851+3.461=16.187. It can be seen that f ₁₃ >f ₁₂ >f ₁₁ , that is, b2 belongs to the "avant-garde" category, and the recommendation is reasonable.

Claims (9)

1. one kind judges garment language class method for distinguishing by colour gamut based on PCCS system, which comprises the following steps:

(1) clothes kansei image space is established, analysis consumer divides garment language class to the perceptual evaluative dimension of clothes accordingly Not, determine that the evaluation of clothes perception represents vocabulary；

(2) it collects and investigates by network and extract clothes representativeness color sample, make the style book figure of perceptual evaluation experimental；

(3) perceptual evaluation test is carried out using questionnaire method combination Semantic Differential；

(4) experimental data is analyzed, clothes kansei image is matched with PCCS hue regions and NCD color space；

(5) discrimination model is established, realizes and passes through judgement of the Color Scheme variable to men's shirt kansei image.

2. according to claim 1 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the step (1) specifically: collect extensively and screening clothing perception evaluates vocabulary, with questionnaire method, card-stacking method With the clustering of data statistics, establishes the kansei image space of clothes and analyze consumer to the perceptual evaluative dimension of clothes.

3. according to claim 2 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the clustering of the card-stacking method and data statistics specifically: prepare the identical blank card of size, and write N number of Vocabulary to be sorted is simultaneously numbered；Several people of subject are invited, are tested by the way of testing one by one；It, please quilt when executing test Feel to put together similar card according to individual after examination person's thinking, grouping number is controlled in 6-10 group, so that subsequent analysis knot Fruit accurate and effective；Statistical packet frequency establishes the similarity matrix of N × N, carries out hierarchial-cluster analysis and obtains determining point of family tree Group's number, it is specified that group's number is divided to cluster to obtain specific grouping result by K-means.

4. according to claim 1 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the step (2) specifically: major apparel brand show field photo and color enterprise planning in recent years is collected from professional fashion website, from The RGB attribute value that main col is extracted in clothes picture concerned therein, clusters color sample to obtain final experiment sample, and Control obtains the PCCS tone and form and aspect attribute value of sample.

5. according to claim 1 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the step (3) carries out evaluation experimental using Semantic Differential, and using 5 grades of evaluation charters of Li Kete, step (2) are handled The clothes perception evaluation of good style book figure and step (1) analysis determination represents vocabulary and is combined, requirement of experiment subject Assessment marking is carried out depending on style book figure of the mental impression to different colours.

6. according to claim 1 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the step (4) specifically: handle perceptual evaluation experimental the data obtained, take the average value of all subject's marking As the final score of the sample, factorial analysis is carried out, determines the impact factor that color evaluates consumer's perception；Simultaneously by sample This kansei image is matched with the Color Range of PCCS, obtains the representative colour gamut of all kinds of styles.

7. according to claim 6 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the kansei image by sample is matched with the Color Range of PCCS, obtains the representative colour gamut of all kinds of styles specifically: The kansei image analysis of monochromatic clothes is placed the result in PCCS colour system coordinate diagram；Summarize K kind garment language color Range areas in PCCS colour system；The distribution of PCCS tone and position by obtained K class clothes reflect to NCD coordinate diagram In, the fuzzy region of K class clothes is uniformly located in the space NCD.

8. according to claim 1 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the step (5) specifically: carry out discriminant analysis to the data obtained, the K/2 obtained with factorial analysis is to clothes image word pair As the dependent variable of discriminant analysis, independent variable is tone and form and aspect in PCCS theoretical system, obtains classification function as differentiation Model.

9. according to claim 8 judge garment language class method for distinguishing by colour gamut based on PCCS system, feature exists In the independent variable includes two groups, and one group is tone independent variable, shares 12 dummy variables, and another group is form and aspect independent variable, altogether There are 11 dummy variables.