CN113808118A - Intelligent matching method for colors of upper garment and lower garment of clothes - Google Patents
Intelligent matching method for colors of upper garment and lower garment of clothes Download PDFInfo
- Publication number
- CN113808118A CN113808118A CN202111122781.8A CN202111122781A CN113808118A CN 113808118 A CN113808118 A CN 113808118A CN 202111122781 A CN202111122781 A CN 202111122781A CN 113808118 A CN113808118 A CN 113808118A
- Authority
- CN
- China
- Prior art keywords
- color
- garment
- clothing
- matching
- colors
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
Abstract
The invention relates to the technical field of clothes color matching methods. The provided method for intelligently matching colors of the upper garment and the lower garment of the garment comprises the following steps: a. constructing a clothing color collocation value database comprising a clothing color collocation negative factor database X, a clothing color collocation classic database Y and a clothing color collocation negative database Z based on HSV color collocation rules; b. shooting to obtain pictures of the upper garment and the lower garment; c. respectively calculating texture features and color features of the clothes of the pictures; d. and assigning values for the color matching of the upper garment and the lower garment according to the garment color matching assignment value database X, Y, Z, wherein the assigned values are the appropriate degree of the color matching. By adopting the intelligent matching method for colors of the upper garment and the lower garment of the garment, provided by the invention, the proper color matching of the upper garment and the lower garment can be provided for a user.
Description
Technical Field
The invention relates to the technical field of clothes color matching methods, in particular to an intelligent matching method for colors of upper clothes and lower clothes of clothes.
Background
In the matching of clothes, there are many researches such as style, color matching and the like; in different situations, the color and style need to be studied. The color of the garment is an element which is obvious in impression, and for many users, the colors of the upper garment and the lower garment are difficult to match reasonably and properly, but in the prior art, an intelligent recommendation method for matching the colors of the upper garment and the lower garment of the garment is not provided.
Disclosure of Invention
The invention aims to provide an intelligent matching method for colors of upper clothes and lower clothes of clothes.
The invention adopts the following technical scheme:
the intelligent matching method for colors of the upper garment and the lower garment of the garment comprises the following steps:
a. based on HSV color matching rules, the construction comprises
A negative factor library X for matching clothing colors,
Clothing color matching classic library Y,
Clothing color matching negative library Z
The clothing color matching assigned value database;
b. shooting to obtain pictures of the upper garment and the lower garment;
c. respectively calculating texture features and color features of the clothes of the pictures;
d. and assigning values for the color matching of the upper garment and the lower garment according to the garment color matching assignment value database X, Y, Z, wherein the assigned values are the appropriate degree of the color matching.
By adopting the intelligent matching method for colors of the upper garment and the lower garment of the garment, provided by the invention, the proper color matching of the upper garment and the lower garment can be provided for a user.
Drawings
FIG. 1 is a flow chart of a method for intelligently matching colors of upper garment and lower garment according to the present invention;
fig. 2 is a flowchart of the specific steps of calculating the texture features and color features of the picture according to the intelligent matching method for colors of the upper garment and the lower garment of the garment provided by the invention.
Detailed Description
The invention is further explained with reference to the drawings and the embodiments.
The intelligent matching method for colors of the upper garment and the lower garment of the garment comprises the following steps:
a. based on HSV color matching rules, the construction comprises
A negative factor library X for matching clothing colors,
Clothing color matching classic library Y,
Clothing color matching negative library Z
The clothing color matching assigned value database;
b. shooting to obtain pictures of the upper garment and the lower garment;
c. respectively calculating texture characteristics and color characteristics of the clothes of the pictures;
d. and assigning values for the color matching of the upper garment and the lower garment according to the garment color matching assignment value database X, Y, Z, wherein the assignment value is the appropriate degree of the color matching.
The color characteristics comprise color system, cold and warm, purity, lightness and depth;
the texture features comprise clear color and flower color, wherein the flower color comprises horizontal stripes, vertical stripes, large and small broken flowers, large and small spots, rectangles, circles, edges and corners, diagonal drawing and combinations thereof.
Further, the constructing includes
A negative factor library X for matching clothing colors,
Clothing color matching classic library Y,
Clothing color matching negative library Z
The clothing color matching assigned value database specifically comprises the following steps:
a1. constructing a clothing color matching negative factor library X according to the color characteristics and the texture characteristics based on the HSV color matching rule;
a2. constructing a clothing color matching classic library Y according to clothing matching classic data based on HSV color matching rules;
a3. and constructing a clothing color matching negative library Z according to the clothing matching classical data based on the HSV color matching rule.
Further, the specific steps of calculating the texture features and the color features of the clothing of the picture are as follows:
c1. according to the picture, determining the texture characteristics of the garment by adopting a texture characteristic data regression model, and judging the garment to be clear color or pattern color according to the texture characteristics; wherein the patterns further belong to horizontal stripes, vertical stripes, large and small broken patterns, large and small spots, rectangles, circles, corner shapes, diagonal drawing types and combinations thereof;
c11. if the color is judged to be clear, extracting the average HSV value of the clear color, and determining the color characteristic according to the HSV value;
due to the reasons of photographing light, angles and the like, colors of different positions photographed by the same piece of clear-color clothes are different, so that the average HSV value of the photographed clothes needs to be corrected and calculated through an algorithm;
c12. and if the color is judged to be the flower color, extracting the background color of the flower color through background color extraction data model operation, calculating the average HSV value of the background color, and determining the color characteristic through the HSV value.
Further, the texture feature data regression model construction method comprises the following steps:
firstly, dividing the texture of the clothes into net color and pattern color; secondly, the patterns are subdivided into horizontal stripes, vertical stripes, large and small broken flowers, large and small spots, rectangles, circles, corner shapes, diagonal drawing types and combinations thereof;
according to the classification, a texture feature data regression model is constructed by adopting the following steps:
performing segmentation extraction on the clothing picture, performing minimum region blocking, performing edge extraction on the minimum blocking by using roberts and other algorithms, and performing parameter statistics on the edge picture;
marking the targets in the gallery according to the classification of the clothing texture;
performing regression processing on the statistical parameters and the texture marks by adopting a logistic regression method to obtain a texture feature data regression model;
the use method of the texture feature data regression model comprises the following steps:
use of texture feature data regression model: and (3) dividing and extracting the clothing picture to be processed, partitioning the minimum area, carrying out edge extraction on the minimum partition, carrying out parameter statistics on the edge picture, and then bringing the parameters into a model for regression to obtain a most similar mark value result so as to determine the texture feature class of the clothing picture.
Algorithms for performing edge extraction include roberts algorithms.
Further, the specific method for extracting the average HSV value of the net color comprises the following steps:
histogram statistics is carried out on the RGB values of the target clothing picture, the color characteristic value with the maximum distribution is extracted, and then the color characteristic value is converted into the HSV value.
Further, the background color extraction data model is constructed by the following method:
carrying out edge detection on the picture to obtain an edge graph;
dividing and extracting a target according to the edge graph;
performing histogram statistics on RGB values of the image, and extracting a color value with the maximum distribution as a main background color of a target;
the primary background color RGB values are converted into HSV values.
The algorithm for edge detection includes the sobel algorithm.
In the above intelligent matching method for colors of the upper garment and the lower garment,
the clothing color matching negative factor library X is a color characteristic which is summarized by practice and is not suitable for matching, for example, bright colors can not be matched with bright colors; the color characteristics which are not suitable for matching are combined and gathered together to form a negative factor library X. The color characteristics are called factors and are distinguished from specific colors.
The clothing color matching classic matching library Y is a color matching which is very suitable for matching and summarized by practice, such as pink matching light blue; the classical color matching combinations are gathered together to form a classical matching library Y.
The clothing color matching negative library Z is a very inappropriate color matching after practice summary, for example, deep blue matching deep green; the negative color matching combinations are gathered together to form a matching negative library Z.
The following illustrates a specific use scenario of the intelligent color matching method for clothing top and bottom garments provided by this embodiment:
the user can shoot one or more pieces of top-loading and bottom-loading through intelligent equipment including a mobile phone and import the pictures into the corresponding client program, and the client program assigns values to color matching of the top-loading and the bottom-loading. The client program can also communicate with the server program to perform further data analysis and obtain more matching data.
The specific way of assigning the color matching of the upper garment and the lower garment can be referred to as follows:
1 if the user does not explicitly dress the scene, the assignment rule includes the steps of:
1.1 the color characteristics of the upper garment and the lower garment meet the Z rule of the garment color collocation negative library, the value is 0, and the color characteristics are directly excluded.
1.2 the top and bottom color characteristics conform to the garment color matching classic base Y rule with an initial assignment of 95. The classical matching library is labeled according to the complementation, contrast, proximity, similarity, "colored + colorless" of the HSV color matching rules.
1.3 initial assignment 90 that the top and bottom color characteristics meet the color complementation and contrast rule of the HSV color matching rule, which does not meet the Y, Z rule. And then, matching the colors of the clothes with a negative factor library X, analyzing the color characteristics of the upper garment and the lower garment, and deducting 1 point when each negative factor is met, wherein the upper limit of deduction is 5 points (more than 5 deduction factors appear, and the value is considered to be excluded and is 0). And sorting according to the scores after the negative factors are deducted, and randomly sorting the same scores.
1.4 no compliance with Y, Z rules, top and bottom color features compliance with similar, adjacent, color + no color rules, initial assigned a score of 85. And then, analyzing the color characteristics of the upper garment and the lower garment by contrasting the color matching negative factor library X of the garment, and deducting the upper limit of 5 points when the color characteristics of the upper garment and the lower garment accord with each negative factor and deducing 1 point (more than 5 deduction factors appear, and the value is considered to be excluded and is 0). And sorting according to the scores after the negative factors are deducted, and randomly sorting the same scores.
2 if the user clearly distinguishes the dressing scenario (leisure, dating, commuting, business), the following is done:
2.1 scenes are leisure, appointments. The top-up and bottom-up matches the classic match library (labeled as "complementary" or "comparative") for 95 points; the color characteristics of the top and bottom clothes conform to the complementation and contrast rules of the HSV color matching rules, and the initial value is assigned to 90; the top and bottom color characteristics conform to the similar, adjacent, colored + colorless rule, and are initially assigned a score of 85. And then, analyzing the color characteristics of the upper garment and the lower garment by contrasting the color matching negative factor library X of the garment, and deducting the upper limit of 5 points when the color characteristics of the upper garment and the lower garment accord with each negative factor and deducing 1 point (more than 5 deduction factors appear, and the value is considered to be excluded and is 0). And sorting according to the scores after the negative factors are deducted, and randomly sorting the same scores.
2.2 scenarios are commute, business. The top and bottom color characteristics were in accordance with the classic collocation library (proximity, similar labels) assignments for 95 points; the color characteristics of the top and bottom clothes accord with the rule of 'color and no color', and the initial value is assigned for 90 points; the top and bottom color characteristics meet similar, adjacent rules, and are initially assigned a score of 85. And then, analyzing the color characteristics of the upper garment and the lower garment by contrasting the color matching negative factor library X of the garment, and deducting the upper limit of 5 points when the color characteristics of the upper garment and the lower garment accord with each negative factor and deducing 1 point (more than 5 deduction factors appear, and the value is considered to be excluded and is 0). And sorting according to the scores after the negative factors are deducted, and randomly sorting the same scores.
And the client program feeds back the result of assigning the color matching of the top and bottom clothes to the user so that the user can select the proper color matching of the top and bottom clothes.
Therefore, by adopting the intelligent matching method for colors of the upper garment and the lower garment of the garment, provided by the invention, the proper color matching of the upper garment and the lower garment can be provided for a user.
The above is illustrative of the present invention.
Claims (6)
1. The intelligent matching method for colors of the upper garment and the lower garment of the garment comprises the following steps:
a. based on HSV color matching rules, the construction comprises
A negative factor library X for matching clothing colors,
Clothing color matching classic library Y,
A clothing color matching assignment data base of the clothing color matching negative base Z;
b. shooting to obtain pictures of the upper garment and the lower garment;
c. respectively calculating texture features and color features of the clothes of the pictures;
d. and assigning values for the color matching of the upper garment and the lower garment according to the garment color matching assignment value database X, Y, Z, wherein the assigned values are the appropriate degree of the color matching.
2. The method of claim 1, wherein the colors of the upper garment and the lower garment are matched intelligently,
the construction comprises
A negative factor library X for matching clothing colors,
Clothing color matching classic library Y,
The clothing color matching assignment data base of the clothing color matching negative base Z specifically comprises the following steps:
a1. constructing a clothing color matching negative factor library X according to the color characteristics and the texture characteristics based on the HSV color matching rule;
a2. constructing a clothing color matching classic library Y according to clothing matching classic data based on HSV color matching rules;
a3. and constructing a clothing color matching negative library Z according to the clothing matching classical data based on the HSV color matching rule.
3. The method of claim 1, wherein the colors of the upper garment and the lower garment are matched intelligently,
the specific steps of calculating the texture features and the color features of the clothing of the picture are as follows:
c1. according to the picture, determining the texture characteristics of the garment by adopting a texture characteristic data regression model, and judging the garment to be clear color or pattern color according to the texture characteristics; wherein the patterns further belong to horizontal stripes, vertical stripes, large and small broken patterns, large and small spots, rectangles, circles, corner shapes, diagonal drawing types and combinations thereof;
c11. if the color is judged to be clear, extracting an average HSV value of the clear color, and determining color characteristics according to the HSV value;
due to the reasons of photographing light, angles and the like, colors of different positions photographed by the same piece of clear-color clothes are different, so that the average HSV value of the photographed clothes needs to be corrected and calculated through an algorithm;
c12. and if the color is judged to be the flower color, extracting the background color of the flower color through background color extraction data model operation, calculating the average HSV value of the background color, and determining the color characteristic through the HSV value.
4. The intelligent matching method for colors of clothes tops and bottoms according to claim 3, wherein the texture feature data regression model is constructed by the following steps:
firstly, dividing the texture of the clothes into net color and pattern color; secondly, the patterns are subdivided into horizontal stripes, vertical stripes, large and small broken flowers, large and small spots, rectangles, circles, corner shapes, diagonal drawing types and combinations thereof;
according to the classification, a texture feature data regression model is constructed by adopting the following steps:
performing segmentation extraction on the clothing picture, performing minimum region blocking, performing edge extraction on the minimum blocking by using roberts and other algorithms, and performing parameter statistics on the edge picture;
marking the targets in the gallery according to the classification of the clothing texture;
performing regression processing on the statistical parameters and the texture marks by adopting a logistic regression method to obtain a texture feature data regression model;
the use method of the texture feature data regression model comprises the following steps:
use of texture feature data regression model: and (3) dividing and extracting the clothing picture to be processed, partitioning the minimum area, carrying out edge extraction on the minimum partition, carrying out parameter statistics on the edge picture, and then bringing the parameters into a model for regression to obtain a most similar mark value result so as to determine the texture feature class of the clothing picture.
5. The method of claim 3, wherein the colors of the upper garment and the lower garment are matched intelligently,
the specific method for extracting the average HSV value of the net color comprises the following steps:
histogram statistics is carried out on the RGB values of the target clothing picture, the color characteristic value with the maximum distribution is extracted, and then the color characteristic value is converted into the HSV value.
6. The method of claim 3, wherein the colors of the upper garment and the lower garment are matched intelligently,
the construction method of the background color extraction data model comprises the following steps:
carrying out edge detection on the picture to obtain an edge graph;
dividing and extracting a target according to the edge graph;
performing histogram statistics on RGB values of the image, and extracting a color value with the maximum distribution as a main background color of a target;
the primary background color RGB values are converted into HSV values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111122781.8A CN113808118A (en) | 2021-09-24 | 2021-09-24 | Intelligent matching method for colors of upper garment and lower garment of clothes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111122781.8A CN113808118A (en) | 2021-09-24 | 2021-09-24 | Intelligent matching method for colors of upper garment and lower garment of clothes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113808118A true CN113808118A (en) | 2021-12-17 |
Family
ID=78940398
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111122781.8A Pending CN113808118A (en) | 2021-09-24 | 2021-09-24 | Intelligent matching method for colors of upper garment and lower garment of clothes |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113808118A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184393A (en) * | 2011-06-20 | 2011-09-14 | 苏州两江科技有限公司 | Method for judging automobile type according to license plate recognition |
| CN109978720A (en) * | 2017-12-28 | 2019-07-05 | 深圳市优必选科技有限公司 | Wear methods of marking, device, smart machine and storage medium |
| CN110232253A (en) * | 2019-06-20 | 2019-09-13 | 杭州时趣信息技术有限公司 | Generate computer installation, equipment, storage medium and the method for clothing matching scheme |
| CN111401748A (en) * | 2020-03-17 | 2020-07-10 | 李照宇 | Intelligent dressing matching big data evaluation method based on working scene |
-
2021
- 2021-09-24 CN CN202111122781.8A patent/CN113808118A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184393A (en) * | 2011-06-20 | 2011-09-14 | 苏州两江科技有限公司 | Method for judging automobile type according to license plate recognition |
| CN109978720A (en) * | 2017-12-28 | 2019-07-05 | 深圳市优必选科技有限公司 | Wear methods of marking, device, smart machine and storage medium |
| CN110232253A (en) * | 2019-06-20 | 2019-09-13 | 杭州时趣信息技术有限公司 | Generate computer installation, equipment, storage medium and the method for clothing matching scheme |
| CN111401748A (en) * | 2020-03-17 | 2020-07-10 | 李照宇 | Intelligent dressing matching big data evaluation method based on working scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105046280B (en) | A kind of wardrobe intelligent management apapratus and method | |
| CN108206917B (en) | Image processing method and device, storage medium and electronic device | |
| CN104484450B (en) | Clothing matching based on image recommends method and clothing matching recommendation apparatus | |
| CN110222644A (en) | Forest fire smoke detection method based on image segmentation | |
| CN108121957A (en) | The method for pushing and device of U.S. face material | |
| CN104636759B (en) | A kind of method and picture filter information recommendation system for obtaining picture and recommending filter information | |
| JP3918143B2 (en) | Plant recognition system | |
| CN109215091B (en) | Clothing fashion color automatic extraction method based on graph representation | |
| CN110598017B (en) | Self-learning-based commodity detail page generation method | |
| CN106097354B (en) | A kind of hand images dividing method of combining adaptive Gauss Face Detection and region growing | |
| CN106454064A (en) | Image processing apparatus, and image processing method | |
| CN106803257A (en) | The dividing method of scab in a kind of crop disease leaf image | |
| CN110569859B (en) | Color feature extraction method for clothing image | |
| JP5811416B2 (en) | Image processing apparatus, image processing method, and program | |
| CN109409377A (en) | The detection method and device of text in image | |
| CN108629319A (en) | Image detecting method and system | |
| CN113808118A (en) | Intelligent matching method for colors of upper garment and lower garment of clothes | |
| Dwina et al. | Skin segmentation based on improved thresholding method | |
| CN103984927B (en) | Information processing method and electronic equipment | |
| CN105872516A (en) | Method and device for obtaining parallax parameters of three-dimensional film source | |
| CN112489049A (en) | Mature tomato fruit segmentation method and system based on superpixels and SVM | |
| Ding et al. | Real-time interactive image segmentation using improved superpixels | |
| CN111047669B (en) | Color taking method and device | |
| Funt et al. | Removing outliers in illumination estimation | |
| CN104484324B (en) | A kind of pedestrian retrieval method of multi-model and fuzzy color |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |