CN113705681A - Lipstick number identification method based on machine learning - Google Patents
Lipstick number identification method based on machine learning Download PDFInfo
- Publication number
- CN113705681A CN113705681A CN202111000354.2A CN202111000354A CN113705681A CN 113705681 A CN113705681 A CN 113705681A CN 202111000354 A CN202111000354 A CN 202111000354A CN 113705681 A CN113705681 A CN 113705681A
- Authority
- CN
- China
- Prior art keywords
- lipstick
- hsv
- value
- training
- clustering
- 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
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000010801 machine learning Methods 0.000 title claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 36
- 238000003066 decision tree Methods 0.000 claims abstract description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 2
- 238000005286 illumination Methods 0.000 abstract 1
- 238000010606 normalization Methods 0.000 abstract 1
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000001815 facial effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 239000002537 cosmetic Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a lipstick number identification method based on machine learning, which is used for solving the problem of poor stability of a judgment result caused by judging the color of lipstick through human eyes or traditional algorithm identification in the prior art. The method comprises the steps of obtaining a lipstick color number training sample; training an identification model by using chrominance information of a training sample, wherein the model comprises two parallel clustering branches, and an RGB value and an HSV value are respectively input into the two branches to be independently trained until clustering is stable, and the number of clusters and the cluster center are determined; and respectively inputting the RGB and HSV values of the sample to be recognized into the trained recognition model to obtain two results, and inputting the two results into a decision tree to obtain a final recognition result. The method adopts RGB and HSV to respectively judge the color of the lipstick, has higher accuracy compared with single RGB or HSV identification, and simultaneously eliminates the interference on the identification result caused by different illumination environments through image normalization, thereby improving the accuracy of the identification of the color number of the lipstick.
Description
Technical Field
The invention belongs to the technical field of machine learning application, and particularly relates to a red number identification method based on machine learning.
Background
At present, when people select the lipstick, people often judge the color of the lipstick through human eyes, and then decide whether to select the lipstick according to the color of the lipstick.
The method for judging the color of the lipstick through human eyes is easily influenced by various factors, such as the influence of the body condition of the human body and the influence of the light environment in which the lipstick is placed, so that people can obtain different color judgment results for the lipstick with the same color under different scenes, namely the judgment result for the color of the lipstick is unstable.
Furthermore, because the types of the lipsticks are various, the lipsticks of different brands have respective independent color definition standards, the brands have strong identification degrees for the color numbers of the lip color cosmetics, products with different colors have various names and definitions, the lipsticks of different brands are not easy to be put together for color comparison, and the color numbers of the lipsticks cannot be determined only by means of difficulty in identifying the specific colors of the lipsticks with naked eyes. For example, a certain lipstick of different brands is essentially one color, but with different names, it is difficult for people to remember the difference between similar color lipsticks of different brands, thereby causing trouble to people in selecting the color of the lipstick. Therefore, certain difficulties are posed to the analysis of the current prevalence trend of lipstick.
At present, only one RGB value is used as a parameter for identifying a lipstick number, the method adopts two parameters of RGB and HSV simultaneously, double channels are used for identifying simultaneously, results of the RGB and HSV are comprehensively evaluated, and errors are reduced.
Disclosure of Invention
In order to solve the problem of quickly and accurately identifying lipstick numbers of different brands, the invention provides a lipstick number identification method based on machine learning. The specific technical scheme is as follows:
step 1, obtaining a lipstick number training sample, preprocessing the training sample, and obtaining chrominance information of the training sample, wherein the chrominance information comprises RGB (red, green, blue) value information and HSV (hue, saturation and value) information;
step 2, training a recognition model by using chrominance information of a training sample, wherein the recognition model comprises two parallel clustering branches, and the RGB value and the HSV value are respectively input into the two parallel clustering branches for independent training until clustering is stable, so that two sets of independent clustering branches are obtained, and the number and the center of clusters are determined;
and 3, respectively inputting the RGB value and HSV value information of the sample to be recognized into the trained recognition model to obtain two output results, and inputting the two results into a decision tree for decision making to obtain a final recognition result.
Further, the pretreatment comprises:
normalizing the pictures of the training samples;
carrying out noise reduction processing on the pictures of the training samples, and averaging;
and acquiring the RGB value and HSV value of the training sample after the noise reduction treatment.
Further, the training samples are the chroma information of the lipstick obtained on various major lipstick brand official websites, and the samples to be identified are lipstick pictures uploaded by users.
Further, the clustering branches are clustered by using a K-Medians clustering method in machine learning.
Compared with the prior art, the beneficial effect of this disclosure is:
the method and the device realize the identification of the lipstick number of the picture to be detected, realize the intelligent identification of the lipstick number, and the user can identify the lipstick number only by providing the facial image. In addition, the traditional lipstick number identification method is only based on RGB detection standard for detection, and the RGB and HSV double channels are adopted for decision identification, so that the identification precision is improved. The recognition speed is improved through the double-channel training.
Drawings
Fig. 1 is a flowchart of a lipstick number identification method based on machine learning according to the present invention.
Detailed Description
In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be clearly and completely described below with reference to the embodiments and the accompanying drawings, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments.
Step 1: acquiring standard RGB values and HSV values of the lipstick on various official websites of the large lipstick brand;
step 2: the acquired lipstick data is processed.
Normalizing the lipstick picture collected in the above steps, performing noise reduction, and averaging,
acquiring an RGB value and an HSV value of a training sample after noise reduction;
and step 3: training samples are collected by the steps to train the recognition model.
And training an identification model by using the chrominance information of the training sample, wherein the identification model comprises two parallel clustering branches, and the RGB value and the HSV value are respectively input into the two parallel clustering branches to be independently trained until the clustering is stable, so that two sets of independent clustering branches are obtained, and the cluster number and the cluster center are determined. The classification model to be trained comprises at least one class cluster;
in the embodiment, the clustering branches adopt K-Medians in machine learning, and the training processes of the two branches are the same, specifically as follows:
clustering the training samples by using any clustering method to obtain the cluster center of the initial cluster;
calculating the Euclidean distance between the chromaticity information of the training sample and the cluster center of each initial class cluster, wherein the smaller the Euclidean distance is, the higher the similarity between the training sample and the class cluster is;
comparing the Euclidean distance with a preset threshold value, and if the Euclidean distance is smaller than or equal to the preset threshold value, clustering the training sample to a cluster with the minimum Euclidean distance; if the Euclidean distance is larger than a preset threshold value, the similarity between the training sample and the existing cluster center is too low, and the clustering cannot be performed, adding a cluster, wherein the cluster center value of the cluster is equal to the value of the training sample; wherein, the preset threshold belongs to artificial setting and is related to the identification precision; the smaller the preset threshold value is, the higher the identification precision is; the larger the preset threshold value is, the lower the recognition accuracy is.
And when the clustering result is stable, finishing training, storing the number of the trained class clusters and the class cluster center value in a file, finishing the identification model, and directly calling class cluster information from the file for use when the identification model is used.
The clustering method used in the method can be K-Medians in machine learning;
and 4, step 4: acquiring an RGB value and an HSV value of a picture to be recognized, and respectively inputting the RGB value and the HSV value into corresponding recognition models for recognition to obtain an output result of a lipstick number;
acquiring a picture to be identified, wherein in one possible implementation mode, a female facial picture with the heat degree exceeding a preset threshold value or a picture uploaded by a user is acquired through a social media and is used as the picture to be identified;
and respectively inputting the RGB value and HSV value information of the sample to be recognized into the trained recognition model to obtain two output results, and inputting the two results into a decision tree to make a decision to obtain a final recognition result.
Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. While embodiments in accordance with the invention have been described above, these embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments described.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present application and are intended to be covered by the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (4)
1. A lipstick number identification method based on machine learning is characterized by comprising the following steps:
step 1, obtaining a lipstick number training sample, preprocessing the training sample, and obtaining chrominance information of the training sample, wherein the chrominance information comprises RGB (red, green, blue) value information and HSV (hue, saturation and value) information;
step 2, training a recognition model by using chrominance information of a training sample, wherein the recognition model comprises two parallel clustering branches, and the RGB value and the HSV value are respectively input into the two parallel clustering branches for independent training until clustering is stable, so that two sets of independent clustering branches are obtained, and the number and the center of clusters are determined;
and 3, respectively inputting the RGB value and HSV value information of the sample to be recognized into the trained recognition model to obtain two output results, and inputting the two results into a decision tree for decision making to obtain a final recognition result.
2. The machine learning-based lipstick number recognition method according to claim 1, characterized in that said preprocessing comprises:
normalizing the pictures of the training samples;
carrying out noise reduction processing on the pictures of the training samples, and averaging;
and acquiring the RGB value and HSV value of the training sample after the noise reduction treatment.
3. The machine learning-based lipstick number identification method according to claim 1,
the training samples are the chroma information of the lipstick obtained on the official website of each big lipstick brand,
the sample to be identified is a lipstick picture uploaded by a user.
4. The machine learning-based lipstick number identification method according to claim 1,
and the clustering branches are clustered by using a K-Medians clustering method in machine learning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111000354.2A CN113705681A (en) | 2021-08-28 | 2021-08-28 | Lipstick number identification method based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111000354.2A CN113705681A (en) | 2021-08-28 | 2021-08-28 | Lipstick number identification method based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113705681A true CN113705681A (en) | 2021-11-26 |
Family
ID=78656400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111000354.2A Pending CN113705681A (en) | 2021-08-28 | 2021-08-28 | Lipstick number identification method based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113705681A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114818883A (en) * | 2022-04-07 | 2022-07-29 | 中国民用航空飞行学院 | CART decision tree fire image identification method based on optimal combination of color features |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107799116A (en) * | 2016-08-31 | 2018-03-13 | 科大讯飞股份有限公司 | More wheel interacting parallel semantic understanding method and apparatus |
| US20180260974A1 (en) * | 2017-03-09 | 2018-09-13 | Hewlett Packard Enterprise Development Lp | Color recognition through learned color clusters |
| CN110322522A (en) * | 2019-07-11 | 2019-10-11 | 山东领能电子科技有限公司 | A kind of vehicle color identification method based on the interception of target identification region |
| CN112016621A (en) * | 2020-08-28 | 2020-12-01 | 上海第一财经数据科技有限公司 | Training method of classification model, color classification method and electronic equipment |
-
2021
- 2021-08-28 CN CN202111000354.2A patent/CN113705681A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107799116A (en) * | 2016-08-31 | 2018-03-13 | 科大讯飞股份有限公司 | More wheel interacting parallel semantic understanding method and apparatus |
| US20180260974A1 (en) * | 2017-03-09 | 2018-09-13 | Hewlett Packard Enterprise Development Lp | Color recognition through learned color clusters |
| CN110322522A (en) * | 2019-07-11 | 2019-10-11 | 山东领能电子科技有限公司 | A kind of vehicle color identification method based on the interception of target identification region |
| CN112016621A (en) * | 2020-08-28 | 2020-12-01 | 上海第一财经数据科技有限公司 | Training method of classification model, color classification method and electronic equipment |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114818883A (en) * | 2022-04-07 | 2022-07-29 | 中国民用航空飞行学院 | CART decision tree fire image identification method based on optimal combination of color features |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2019100282A1 (en) | Face skin color recognition method, device and intelligent terminal | |
| Elgammal et al. | Skin detection-a short tutorial | |
| Rahmat et al. | Skin color segmentation using multi-color space threshold | |
| Yang et al. | Research on a skin color detection algorithm based on self-adaptive skin color model | |
| CN103634680B (en) | The control method for playing back and device of a kind of intelligent television | |
| CN104636759B (en) | A kind of method and picture filter information recommendation system for obtaining picture and recommending filter information | |
| CN111860538A (en) | Tongue color identification method and device based on image processing | |
| CN104598888B (en) | A kind of recognition methods of face gender | |
| CN110598574A (en) | Intelligent face monitoring and identifying method and system | |
| Lionnie et al. | A comparison of human skin color detection for biometrie identification | |
| CN110599553B (en) | Skin color extraction and detection method based on YCbCr | |
| CN113705681A (en) | Lipstick number identification method based on machine learning | |
| Mohammed et al. | Image segmentation for skin detection | |
| Kumar et al. | Real time detection and tracking of human face using skin color segmentation and region properties | |
| Yusuf et al. | Human face detection using skin color segmentation and watershed algorithm | |
| CN108416782A (en) | View-based access control model identifies and the tobacco leaf ranking method and system of illumination correction | |
| Berbar | Novel colors correction approaches for natural scenes and skin detection techniques | |
| CN110728318A (en) | Hair color identification method based on deep learning | |
| Manjare et al. | Skin detection for face recognition based on HSV color space | |
| CN108875572B (en) | Pedestrian re-identification method based on background suppression | |
| CN110348530B (en) | Method for identifying lipstick number | |
| Berbar | Skin colour correction and faces detection techniques based on HSL and R colour components | |
| Aminian et al. | Face detection using color segmentation and RHT | |
| Nabiyev et al. | Towards a biometric purpose image filter according to skin detection | |
| CN110991223A (en) | Method and system for identifying beautiful pupil based on transfer learning |
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 |