CN103034874A - Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science - Google Patents
Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science Download PDFInfo
- Publication number
- CN103034874A CN103034874A CN2011102985837A CN201110298583A CN103034874A CN 103034874 A CN103034874 A CN 103034874A CN 2011102985837 A CN2011102985837 A CN 2011102985837A CN 201110298583 A CN201110298583 A CN 201110298583A CN 103034874 A CN103034874 A CN 103034874A
- Authority
- CN
- China
- Prior art keywords
- sample
- gloss
- face
- training sample
- test
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a face gloss analytical method based on inspection diagnosis of traditional Chinese medical science. The method sequentially comprises the following steps of automatically collecting face images by using a digital camera controlled by a computer in a stable light environment and automatically storing the images in the computer; obtaining the face images from an area at the position of the cheek and adjusting resolution to be same; using a hue, saturation and value color pattern to conduct feature extraction on test samples and training samples of the face images by using an improved two-dimensional principal component analysis method on a standard face image gloss image and a test sample image; and using a nearest neighbor method to conduct qualitative and quantitative analysis on the test samples by calculating a cosine distance of characteristics of the test samples and the training samples after dimensionality reduction. The face gloss analytical method can directly conduct quantitative analysis and qualitative description of the gloss of the face images and can assist diagnosis of traditional Chinese medical.
Description
Technical field
The invention belongs to tcm diagnosis digitizing and computing machine field of medical image processing, be specifically related to a kind of facial gloss analytical approach based on tcm inspection, the inventive method can directly be carried out quantitative test and the qualitative description of gloss to face-image, auxiliary tcm diagnosis.
Background technology
Facial observation is a kind of method of tcm diagnosis.Traditional Chinese medicine is always paid attention to facial observation, mainly comprises observation of complexion and hopes gloss two parts.Miraculous Pivot cloud " its vim and vigour all go up in face and walk the sky key for the twelve regular channels, 365 networks ", heart governing blood and vessels, its China is at face, and brothers' three yang channels are all up in women's head-ornaments, so the colour inspection of face can be diagnosed the prosperity and decline of internal organs vital essence and the profit and loss of passages through which vital energy circulates qi and blood." the Plain Questions arteries and veins is wanted precise and tiny opinion " said: " husband's eye expression person, the China of gas are also ".Think qi and blood the up women's head-ornaments of elite and outer reach skin, can show different color and lusters." observation abide by through look take moist for this " said: " bright moist person, gas also, the blue or green helvolus person that deceives in vain, look also.Have gas not suffer from colourless, coloured can not be without gas ".The variation of observing facial gloss can be diagnosed the prosperity and decline of internal organs vital essence, to judging state of an illness weight, inferring that prognosis is most important.But traditional facial gloss examination mainly is to rely on clinician's subjective description, is described as glossy, few gloss and tarnish, lacks the Data support that objectifies, and has very strong subjectivity and ambiguity, and this certainly will affect the overall development of tcm diagnosis.The modernization that traditional Chinese medical science face is examined, the research that objectifies have important theory value and clinical meaning to further developing that standardization, evaluation of clinical curative effect and the traditional Chinese medical science face of Chinese medical discrimination are examined.
Along with the continuous progress of science and technology, the technology such as computer technology, especially pattern-recognition, computer vision, data mining progressively are incorporated into the research process that objectifies, standardizes of the traditional Chinese medical science, have obtained interim achievement.The gloss that face is examined quantitatively is the importance that face is examined, find by prior art documents, at present face examine aspect the gloss analysis also without any method and technology report.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of facial gloss analytical approach of examining towards traditional Chinese medical science face, is intended to the patient's face-image that collects is carried out the automatic analysis of gloss.The present invention is from the angle of pattern-recognition, by sample image being carried out dimensionality reduction, calculate the cosine distance between test sample book and the training sample, and differentiates classification and the quantitative value of test sample book gloss by the arest neighbors method, auxiliary tcm diagnosis.
In order to solve the problems of the technologies described above, design of the present invention is: be to have used qualitatively the gloss samples pictures as training sample, by training sample being used a kind of improved two-dimensional principal component analysis obtain projection matrix, with this projection matrix gloss picture training sample and test sample book are carried out feature extraction, calculate the cosine distance of all training sample features of test sample book feature, use at last the arest neighbors method that test sample book is carried out the gloss analysis.
According to the foregoing invention design, the technical solution used in the present invention comprises the steps:
(1) face-image is gathered automatically by the digital camera of computer control under stable photoenvironment, guarantees to gather the stable of environment;
(2) samples pictures is taken from a zone at face-image cheek place, and all adjusts to identical resolution;
(3) use the HSV color mode, if samples pictures is the RGB pattern, will be converted to the HSV color mode;
(4) select a part of picture as training sample, and manually gloss is classified;
(5) the improved two-dimensional principal component analysis method of employing is carried out feature extraction to test sample book and the training sample of face-image;
(6) the cosine distance between calculating test sample book feature and each the training sample feature;
(7) use the method for arest neighbors that test sample book is carried out facial gloss analysis.
The improved two-dimensional principal component analysis method of above-mentioned employing is as follows to the method that test sample book and the training sample of face-image carries out feature extraction:
1. represent sample with matrix I: if I
A, bThe element at the capable b row of representing matrix I a place is listed as the picture that big or small matrix I represents the HSV color mode that a resolution is n * m, I with the capable 3n of m so
3i-2, j, I
3i-1, j, I
3i, jRespectively storage be the capable j row of this image i place pixel H, S, V value (i=1,2,3 ..., m, j=1,2,3 ..., n);
2. the covariance matrix that comes Two-dimensional to analyze according to training sample with a kind of Innovative method:
Wherein k represents the classification number of sample, n
iBe the number of training of i class, N is the training sample sum,
Be the sample average of i class training sample, I
iBe the training sample population mean;
3. ask the Eigenvalues Decomposition of covariance matrix Cov, with the eigenwert that obtains by descending sort: λ
1, λ
2K λ
3n, proper vector is pressed the order ordering of character pair value, so obtain eigenvectors matrix: V=[v
1v
2K v
3n]
T
4. choose the front capable formation of d projection matrix: the V of V
Project=[v
1v
2K v
d]
T
5. according to projection matrix V
ProjectSample is carried out feature extraction obtain sample characteristics F=V
ProjectI
TSo the feature set that obtains training sample is: F
Train={ F
1, F
2K F
N, test sample book is characterized as F
Test
Cosine distance between above-mentioned calculating test sample book feature and each the training sample feature is namely asked F
TestWith F
TrainIn the included angle cosine value of all samples, and the result taken absolute value.
The method of above-mentioned arest neighbors is carried out facial gloss analysis to test sample book, and namely selection and Ftest cosine are the classification of test sample book apart from the classification of the training sample feature of minimum.
The present invention has following apparent outstanding substantive distinguishing features and remarkable advantage:
Method of the present invention can obtain higher matching accuracy rate: owing to the gloss picture sample based on known class, take full advantage of existing sample, two-dimensional principal component analysis after the improvement is a kind of learning method that supervision is arranged, can finely utilize training sample, extract the gloss characteristics that has more the property distinguished.The method has been applicable to the quantitative and qualitative analysis to the automatic gloss of people's face test sample book picture in the qualitative gloss picture of the some situation, can obtain higher test accuracy rate, objectifying of examining of traditional Chinese medical science face there is certain impetus with modernization, has more actual use value.
Description of drawings
Fig. 1 is the process flow diagram of the facial gloss analytical approach based on tcm inspection of the present invention.
Fig. 2 is the computing machine automatic interpretation figure as a result of the facial gloss analytical approach based on tcm inspection of the present invention.
Fig. 3 is of the present invention based on samples pictures interception position exemplary plot in the facial gloss analytical approach of tcm inspection.
Embodiment
Below in conjunction with the drawings and specific embodiments, further set forth the present invention.These embodiment are interpreted as only being used for explanation the present invention and are not used in restriction protection scope of the present invention.After the content of having read the present invention's record, those skilled in the art can make various changes or modifications the present invention, and these equivalences change and modification falls into claim limited range of the present invention equally.
In following examples of the present invention the face-image sample is divided into " glossy ", " few gloss " and " tarnish " three classes, 30 training samples of every class.
As shown in Figure 1, in the facial gloss analytical approach based on tcm inspection that the preferred embodiment of the present invention provides, whole facial gloss analytic process comprises the steps:
(1) under unified photoenvironment, gathers needed face-image.
(2) image at all face-image cutting cheek face places is adjusted into 100 * 100 (such as Fig. 3) as sample image with the sample resolution.
(3) carry out the classification of gloss classification to training sample, be divided into " glossy ", " few gloss " and " tarnish " three class pictures, every class is chosen 30 as training sample.
(4) all pictures are adjusted to the HSV color mode.
(5) the improved two-dimensional principal component analysis method of employing is carried out feature extraction to test sample book and the training sample of face-image, and method is as follows:
1. each sample uses one 100 * 300 matrix I to represent: suppose I
A, bThe element at the capable b row of representing matrix I a place, so I
I, 3j-2, I
I, 3j-1, I
I, 3jRespectively storage be the capable j row of this image i place pixel H, S, V value (i=1,2,3 ..., 100, j=1,2,3 ..., 100);
2. calculate the covariance matrix that a kind of Innovative method comes Two-dimensional to analyze according to training sample:
Wherein
Be the sample average of i class training sample, I
iBe the training sample population mean;
3. ask the Eigenvalues Decomposition of covariance matrix Cov, with the eigenwert that obtains by descending sort: λ
1, λ
2K λ
300, proper vector is pressed the order ordering of character pair value, so obtain eigenvectors matrix: V=[v
1v
2K v
300]
T
4. choose front 10 row of V and consist of projection matrix: V
Project=[v
1v
2K v
10]
T
5. according to projection matrix V
ProjectSample is carried out feature extraction obtain sample characteristics F=V
ProjectI
TSo the feature set that obtains training sample is: F
Train={ F
1, F
2K F
30, test sample book is characterized as F
Test
Cosine distance between above-mentioned calculating test sample book feature and each the training sample feature is namely asked F
TestWith F
TrainIn the absolute value of included angle cosine value of all samples, and deduct this absolute value with 1.
The method of above-mentioned arest neighbors is carried out facial gloss quantitatively and qualitative analysis to test sample book, namely selects and F
TestCosine is the classification of test sample book apart from the classification of the training sample feature of minimum.
As shown in Figure 2, cosine distance between certain test sample book and certain the training sample feature is 0.933, and this distance is the minimum value of all training samples, and class another edition of a book of this training sample is " few gloss ", and this test sample book also belongs to " few gloss " class so; Class another edition of a book of this training sample is " glossy ", and this test sample book also belongs to " glossy " class so; The classification of this training sample is " tarnish ", and this test sample book also belongs to " tarnish " class so.
Test sample book belongs to " glossy ", " few gloss " or " tarnish ", can assist tcm diagnosis.
Claims (3)
1. facial gloss analytical approach based on tcm inspection, it is characterized in that, the method in turn includes the following steps: the digital camera by computer control under stable photoenvironment gathers face-image automatically, the image automatic storage is in computing machine, the cheek zone is split from face-image, use the picture of HSV color mode as test sample book and training sample, adopt improved two-dimensional principal component analysis method that test sample book and the training sample of face-image carried out feature extraction, calculate the cosine distance between test sample book feature and each the training sample feature, and with the method for arest neighbors test sample book is carried out facial gloss analysis, wherein:
The improved two-dimensional principal component analysis method of aforementioned employing is carried out feature extraction to test sample book and the training sample of face-image, and method is as follows:
1. represent sample with matrix I: if I
A, bThe element at the capable b row of representing matrix I a place is listed as the picture that big or small matrix I represents the HSV color mode that a resolution is n * m, I with the capable 3n of m so
I, 3j-2, I
I, 3j-1, I
I, 3jRespectively storage be the capable j row of this image i place pixel H, S, V value (i=1,2,3 ..., m, j=1,2,3 ..., n);
2. calculate the covariance matrix that a kind of Innovative method comes Two-dimensional to analyze according to training sample:
Wherein k represents the classification number of sample, n
iBe the number of training of i class, N is the training sample sum,
Be the sample average of i class training sample, I
iBe the training sample population mean;
3. ask the Eigenvalues Decomposition of covariance matrix Cov, with the eigenwert that obtains by descending sort: λ
1, λ
2K λ
3n, proper vector is pressed the order ordering of character pair value, so obtain eigenvectors matrix: V=[v
1v
2K v
3n]
T
4. choose the front capable formation of d projection matrix: the V of V
Project=[v
1v
2K v
d]
T
5. according to projection matrix V
ProjectSample is carried out feature extraction obtain sample characteristics F=V
ProjectI
TSo the feature set that obtains training sample is: F
Train={ F
1, F
2K F
N, test sample book is characterized as F
Test
2. the facial gloss analytical approach based on tcm inspection according to claim 1 is characterized in that, the cosine distance between described calculating test sample book feature and each the training sample feature is namely asked F
TestWith F
TrainIn the absolute value of included angle cosine value of all samples, and to deducting this absolute value with 1.
3. the facial gloss analytical approach based on tcm inspection according to claim 1 and 2 is characterized in that, the method for described arest neighbors is carried out facial gloss analysis to test sample book, selects and F
TestCosine is the classification of test sample book apart from the classification of the training sample feature of minimum.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102985837A CN103034874A (en) | 2011-09-29 | 2011-09-29 | Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102985837A CN103034874A (en) | 2011-09-29 | 2011-09-29 | Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103034874A true CN103034874A (en) | 2013-04-10 |
Family
ID=48021752
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011102985837A Pending CN103034874A (en) | 2011-09-29 | 2011-09-29 | Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103034874A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103750822A (en) * | 2014-01-06 | 2014-04-30 | 上海金灯台信息科技有限公司 | Laser three-dimensional image acquisition device for inspection of traditional Chinese medicine |
| CN103767684A (en) * | 2014-01-06 | 2014-05-07 | 上海金灯台信息科技有限公司 | Three-dimensional image collection device for inspection diagnosis in traditional Chinese medicine |
| CN104586365A (en) * | 2015-01-26 | 2015-05-06 | 北京工业大学 | Method for quantifying complexion psychological perception and judging adaptability |
| CN105956382A (en) * | 2016-04-26 | 2016-09-21 | 北京工商大学 | Traditional Chinese medicine constitution optimized classification method based on improved CART decision-making tree and fuzzy naive Bayes combined model |
| CN106407645A (en) * | 2016-08-08 | 2017-02-15 | 北京工商大学 | Principal component analysis method-based facial skin state evaluation method |
| CN109711306A (en) * | 2018-12-19 | 2019-05-03 | 新绎健康科技有限公司 | A kind of method and apparatus obtaining facial characteristics based on depth convolutional neural networks |
| CN109978873A (en) * | 2019-03-31 | 2019-07-05 | 山西慧虎健康科技有限公司 | A kind of intelligent physical examination system and method based on Chinese medicine image big data |
| CN110414578A (en) * | 2019-07-16 | 2019-11-05 | 上海电机学院 | A kind of transfer learning method based on the multiple batches of training of dynamic and colour gamut conversion |
| CN110459304A (en) * | 2019-07-19 | 2019-11-15 | 汕头大学 | A kind of health status diagnostic system based on face-image |
| CN112200075A (en) * | 2020-10-09 | 2021-01-08 | 西安西图之光智能科技有限公司 | Face anti-counterfeiting method based on anomaly detection |
| CN115931738A (en) * | 2023-01-09 | 2023-04-07 | 云南烟叶复烤有限责任公司 | Method and system for evaluating quality stability of finished tobacco flakes |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100329569A1 (en) * | 2009-06-29 | 2010-12-30 | Fujitsu Semiconductor Limited | Image processing program, image processing apparatus, and image processing method |
-
2011
- 2011-09-29 CN CN2011102985837A patent/CN103034874A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100329569A1 (en) * | 2009-06-29 | 2010-12-30 | Fujitsu Semiconductor Limited | Image processing program, image processing apparatus, and image processing method |
Non-Patent Citations (2)
| Title |
|---|
| 曾接贤等: "基于KPCA和FMSD的人脸识别", 《计算机工程》, vol. 37, no. 17, 5 September 2011 (2011-09-05) * |
| 李福凤等: "基于计算机技术的中医面色诊信息采集与识别研究", 《世界科学技术-中医药现代化》, no. 06, 15 December 2008 (2008-12-15), pages 28 - 36 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103750822A (en) * | 2014-01-06 | 2014-04-30 | 上海金灯台信息科技有限公司 | Laser three-dimensional image acquisition device for inspection of traditional Chinese medicine |
| CN103767684A (en) * | 2014-01-06 | 2014-05-07 | 上海金灯台信息科技有限公司 | Three-dimensional image collection device for inspection diagnosis in traditional Chinese medicine |
| CN104586365A (en) * | 2015-01-26 | 2015-05-06 | 北京工业大学 | Method for quantifying complexion psychological perception and judging adaptability |
| CN105956382A (en) * | 2016-04-26 | 2016-09-21 | 北京工商大学 | Traditional Chinese medicine constitution optimized classification method based on improved CART decision-making tree and fuzzy naive Bayes combined model |
| CN105956382B (en) * | 2016-04-26 | 2018-06-19 | 北京工商大学 | A kind of tcm constitution Optimum Classification method |
| CN106407645A (en) * | 2016-08-08 | 2017-02-15 | 北京工商大学 | Principal component analysis method-based facial skin state evaluation method |
| CN109711306A (en) * | 2018-12-19 | 2019-05-03 | 新绎健康科技有限公司 | A kind of method and apparatus obtaining facial characteristics based on depth convolutional neural networks |
| CN109978873A (en) * | 2019-03-31 | 2019-07-05 | 山西慧虎健康科技有限公司 | A kind of intelligent physical examination system and method based on Chinese medicine image big data |
| CN110414578A (en) * | 2019-07-16 | 2019-11-05 | 上海电机学院 | A kind of transfer learning method based on the multiple batches of training of dynamic and colour gamut conversion |
| CN110459304A (en) * | 2019-07-19 | 2019-11-15 | 汕头大学 | A kind of health status diagnostic system based on face-image |
| CN112200075A (en) * | 2020-10-09 | 2021-01-08 | 西安西图之光智能科技有限公司 | Face anti-counterfeiting method based on anomaly detection |
| CN112200075B (en) * | 2020-10-09 | 2024-06-04 | 西安西图之光智能科技有限公司 | Human face anti-counterfeiting method based on anomaly detection |
| CN115931738A (en) * | 2023-01-09 | 2023-04-07 | 云南烟叶复烤有限责任公司 | Method and system for evaluating quality stability of finished tobacco flakes |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103034874A (en) | Face gloss analytical method based on inspection diagnosis of traditional Chinese medical science | |
| Yang et al. | Clinical skin lesion diagnosis using representations inspired by dermatologist criteria | |
| Zhang et al. | Detecting diabetes mellitus and nonproliferative diabetic retinopathy using tongue color, texture, and geometry features | |
| Pang et al. | Tongue image analysis for appendicitis diagnosis | |
| Bevilacqua et al. | A novel approach to evaluate blood parameters using computer vision techniques | |
| Binder et al. | Epiluminescence microscopy-based classification of pigmented skin lesions using computerized image analysis and an artificial neural network | |
| Burroni et al. | Melanoma computer-aided diagnosis: reliability and feasibility study | |
| Dimauro et al. | An intelligent non-invasive system for automated diagnosis of anemia exploiting a novel dataset | |
| Dimauro et al. | Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva | |
| CN106971147A (en) | A kind of traditional Chinese medical science facial diagnosis system and facial diagnosis method split based on human face region | |
| Patil et al. | An association between fingerprint patterns with blood group and lifestyle based diseases: a review | |
| Appiahene et al. | Detection of iron deficiency anemia by medical images: a comparative study of machine learning algorithms | |
| CN107505268A (en) | Blood sugar detecting method and system | |
| CN106340000A (en) | Bone age assessment method | |
| Jordan et al. | A review of feature-based retinal image analysis | |
| CN110495888B (en) | Standard color card based on tongue and face images of traditional Chinese medicine and application thereof | |
| TW201100058A (en) | System and method of using digital image multi-region analysis to establish and analyze skin parameters | |
| Wang et al. | Facial image medical analysis system using quantitative chromatic feature | |
| Tirado et al. | Bruise dating using deep learning | |
| CN110036407A (en) | System and method for correcting digital image color based on human sclera and pupil | |
| Flament et al. | Developing an artificial intelligence (AI)‐based descriptor of facial appearance that fits with the assessments of makeup experts | |
| Sacha et al. | Development and qualification of a machine learning algorithm for automated hair counting | |
| Meng et al. | Deep Learning-based Trichoscopic Image Analysis and Quantitative Model for Predicting Basic and Specific Classification in Male Androgenetic Alopecia | |
| Collenne et al. | Fusion between an Algorithm Based on the Characterization of Melanocytic Lesions' Asymmetry with an Ensemble of Convolutional Neural Networks for Melanoma Detection | |
| Vasanthakumari et al. | Classification of skin-cancer lesions based on fluorescence lifetime imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130410 |