CN110334649A - A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device - Google Patents
A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device Download PDFInfo
- Publication number
- CN110334649A CN110334649A CN201910599178.5A CN201910599178A CN110334649A CN 110334649 A CN110334649 A CN 110334649A CN 201910599178 A CN201910599178 A CN 201910599178A CN 110334649 A CN110334649 A CN 110334649A
- Authority
- CN
- China
- Prior art keywords
- internal organs
- region
- facial
- image
- key point
- 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
- 230000001815 facial effect Effects 0.000 title claims abstract description 119
- 238000003745 diagnosis Methods 0.000 title claims abstract description 43
- 239000003814 drug Substances 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 34
- 210000001835 viscera Anatomy 0.000 claims abstract description 146
- 239000003086 colorant Substances 0.000 claims abstract description 28
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 9
- 230000009466 transformation Effects 0.000 claims abstract description 6
- 238000004590 computer program Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 8
- 210000004709 eyebrow Anatomy 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 6
- 238000013145 classification model Methods 0.000 claims description 5
- 210000001747 pupil Anatomy 0.000 claims description 5
- 210000001061 forehead Anatomy 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 2
- 239000000049 pigment Substances 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 13
- 210000004185 liver Anatomy 0.000 description 10
- 210000004072 lung Anatomy 0.000 description 10
- 210000000952 spleen Anatomy 0.000 description 10
- 210000003734 kidney Anatomy 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 7
- 230000006872 improvement Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000003703 image analysis method Methods 0.000 description 3
- 210000000056 organ Anatomy 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 208000035850 clinical syndrome Diseases 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 229940079593 drug Drugs 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000003902 lesion Effects 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000002673 Dioscorea communis Nutrition 0.000 description 1
- 241000544230 Dioscorea communis Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 208000035753 Periorbital contusion Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000036285 pathological change Effects 0.000 description 1
- 231100000915 pathological change Toxicity 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Theoretical Computer Science (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Epidemiology (AREA)
- Data Mining & Analysis (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis and examines survey method and device.Comprising: obtain facial image from facial video image by HAAR feature and ADABOOST algorithm;68 key points of every frame facial image are extracted, using the library DLIB to use HOUGH transformation algorithm, the positioning of Delaunay Triangulation criterion and be partitioned into its eye the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, the hot the five internal organs region of facial Plain Questions thorn;Multicolored the five internal organs feature in every face-image and its five internal organs region is extracted, and is converted into the five internal organs surface characteristics collection;The five internal organs surface characteristics collection is sent into preset SVM multi-categorizer, to identify the five internal organs status information, provides digitlization objective basis for tcm diagnosis.Surveying square law device using of the invention examining can recognize that the five internal organs status information of user, accuracy with higher and analysis efficiency.
Description
Technical field
The present invention relates to artificial intelligence technology and traditional Chinese medicine field more particularly to a kind of artificial vision's intelligence Chinese medicine facial diagnosis five
Dirty situation examines survey method and device.
Background technique
Traditional Chinese medicine is to be studied health in human life activity based on traditional Chinese medical theory and practical experience and converted with disease
Rule and its Comprehensive Science of prevention, diagnosis, treatment, rehabilitation and health care.Traditional Chinese medicine, will using yin-yang and five elements as theoretical basis
Human body regards the entity of gas, shape, mind as, intracorporal to analyze the interpretation of the cause, onset and process of an illness and people by prestige, the method for hearing, asking, cutting four methods of diagnosis conjunction ginseng
The vital organs of the human body, to suit the remedy to the case.
Facial diagnosis is that doctor closes the method joined with the four methods of diagnosis come whole to face and face face are observed, to judge
The method of the lesion situation of Whole Body and part.Since the pathological change of human five internal organs' six internal organs can show the correlation of face eventually
Region, thus facial diagnosis can most see clearly the interpretation of the cause, onset and process of an illness and grasp the state of an illness.Traditional facial diagnosis relies primarily on doctor's range estimation and judges, subjective
Property it is stronger, accuracy and consistency are insufficient, are unfavorable for the succession, development and application of tcm diagnosis technology.
With the continuous development of the technologies such as image recognition, machine learning and computer vision, image processing techniques point is utilized
Analysis Chinese medicine facial diagnosis information may advantageously facilitate the accuracy of Chinese medicine facial diagnosis.However, since the analysis of existing Chinese medicine facial diagnosis is usually benefit
Patient's face image is split with individual face-image of user, it is qualitative fixed to obtain facial color, gloss and lip
Amount analysis as a result, and existing Chinese medicine facial diagnosis analysis method can not analyze five dirty situations of user, therefore there are accuracy it is low and
The low problem of analysis efficiency.
Summary of the invention
In view of the above-mentioned problems, five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis of the invention examines survey method and device, it can
The five internal organs status information of user is analyzed, and its accuracy is high, analysis efficiency is high.
In order to solve the above technical problems, five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis of the invention examines survey method, comprising:
Possessed by HAAR feature and ADABOOST algorithm from input computer and obtains face figure in the video image of face
Picture;
68 key points of every frame facial image are extracted, using the library DLIB to convert algorithm, Delaunay tri- using HOUGH
Angle subdivision criterion positions and is partitioned into its eye the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, facial hot five dirty districts of Plain Questions thorn
Domain;
Multicolored the five internal organs feature in every frame facial image and its each the five internal organs region is extracted, and it is special to be converted into corresponding the five internal organs
Levy sample set;The five colors are used to indicate the body surface sickly complexion that traditional Chinese medicine defines, and five color characteristic of the five internal organs is used to indicate often
The pigment ratio of body surface sickly complexion in image is opened, the five internal organs feature is used to indicate the five internal organs state grade of every image;
Each the five internal organs feature samples collection is sent into preset SVM multi-categorizer, to identify the five internal organs status information.
As an improvement of the above scheme, the eye the five internal organs region of every frame facial image is determined as follows:
Using eyebrow key point, upper palpebra inferior key point and the interior external eyes in 68 key points of every frame facial image
Angle key point determines ocular image and its canthus region, palpebral region, and converts positioning using HOUGH and divide the eye
Eye benevolence region, white of the eye region and pupil region in portion's area image, to form eye the five internal organs region.
As an improvement of the above scheme, the facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region is determined as follows:
Using 68 key points of every frame facial image as triangulation point set to carry out Delaunay Triangulation,
Obtain triangle gridding;
According in 68 key points of every frame facial image nasion key point, bridge of the nose key point, nose key point,
Brows key point and lower lip key point position and be partitioned into from corresponding triangle gridding respectively nasion region, bridge of the nose region,
Nose region, place between the eyebrows region and chin portion region, to form facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region.
As an improvement of the above scheme, determine that the facial Plain Questions pierces hot the five internal organs region as follows:
It is crucial according to the brows key point in 68 key points of every frame facial image, left nose point key point, nose
Point, right nose key point and lower lip key point position from corresponding triangle gridding respectively and are partitioned into forehead region, left face
Buccal region domain, nasal region, right cheek region and chin portion region ask thorn hot the five internal organs region to form face.
As an improvement of the above scheme, multicolored the five internal organs feature in each region determines as follows:
The RGB average value of each pixel is calculated according to the R component of each each pixel in region, G component, B component;
According to red ratio, the cyan ratio, yellow in each region described in the RGB mean value calculation of each pixel
Ratio, white proportion and black ratio, to determine that the maximum color of ratio is five color characteristics in each region.
As an improvement of the above scheme, each the five internal organs feature samples collection is converted to as follows:
Interlocked the corresponding relationship between theoretical transformation table according to five color characteristic of each region the five internal organs and preset sickly complexion, is determined
Its five internal organs feature samples collection;The five internal organs feature include eye the five internal organs feature, facial the five internal organs feature, facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs feature,
Facial Plain Questions pierces hot the five internal organs feature.
As an improvement of the above scheme, the binocular camera that the facial video image is made of two high-resolution cameras
It collects.
As an improvement of the above scheme, the five internal organs status information is identified as follows:
According to each the five internal organs surface characteristics collection building training set and test set;
The SVM multi-categorizer is constructed by the training set;
Construct train classification models;
Model prediction is carried out to the test set by the train classification models of building, obtains the five internal organs status information.
The present invention also provides a kind of five dirty situations of artificial vision's intelligence Chinese medicine facial diagnosis to examine survey device, artificial vision's intelligence
Five dirty situation of Chinese medicine facial diagnosis examines survey device and contains computer program, and the computer program is suitable for being performed any of the above-described to realize
Five dirty situation of Visual intelligent Chinese medicine facial diagnosis described in kind examines survey method.
The present invention also provides a kind of five dirty situations of artificial vision's intelligence Chinese medicine facial diagnosis to examine survey device, including processor and institute
The memory of processor connection is stated, the memory is stored with computer program, and the computer program is suitable for being performed with reality
Five dirty situation of Visual intelligent Chinese medicine facial diagnosis described in any of the above-described kind existing examines survey method.
Compared with prior art, a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis of the invention examines survey method and dress
It sets, has the advantages that
(1) eye the five internal organs feature, face five are obtained by the way that theory of traditional Chinese medical science and artificial vision's intellectual technology to be combined
Dirty feature, facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs feature, facial Plain Questions pierce hot the five internal organs feature, then using SVM multi-categorizer to the five internal organs feature into
Row classification judgement, realizes the five internal organs status information quantization analysis and identification, can effectively improve the standard of the analysis of the five internal organs status information
Exactness and analysis efficiency.
(2) the hot theory of traditional Chinese medical science of five wheel five colors theories, the Ling Shu Miraculous Pivot or Divine Axis five colors, Plain Questions thorn is referred and synthesized to set about from Chinese medicine facial diagnosis to the five internal organs shape
State carries out quantitative analysis, provides objective basis for auxiliary clinical syndrome differentiation diagnosis;
(3) mesh is examined into five wheel five colors theories and facial diagnosis is added, the content of abundant facial diagnosis more objectively reflects five to more acurrate
Dirty situation.
Detailed description of the invention
Fig. 1 is the process that a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis provided in an embodiment of the present invention examines survey method
Schematic diagram.
Fig. 2 is the structural model of SVM multi-categorizer in the embodiment of the present invention.
Fig. 3 is that a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis provided in an embodiment of the present invention examines the structure for surveying device
Schematic diagram.
Specific embodiment
In the following description, numerous specific details are set forth in order to facilitate a full understanding of the present invention.But the present invention can be with
It is different from the other modes of this description much to implement, those skilled in the art can be without violating the connotation of the present invention
Similar popularization is done, therefore the present invention is not limited by the specific embodiments disclosed below.
Clear, complete description is carried out to technical solution of the present invention with attached drawing combined with specific embodiments below.
It referring to Figure 1, is that a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis provided in an embodiment of the present invention examines survey side
The flow diagram of method.
As shown in Figure 1, a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method, include the following steps:
S1, the facial video image for obtaining input computer;
In step sl, the facial video image is acquired by the binocular camera that two high-resolution cameras form
It arrives.
S2, facial image is obtained from the facial video image using HAAR feature and ADABOOST algorithm.
S3, it is extracted using 68 key points of the library DLIB to every frame facial image, to position and divide respectively
The eye the five internal organs region of every frame facial image, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, facial Plain Questions pierce hot the five internal organs region out;
In step s3, since current Chinese medicine thinks, the lesion of internal organ can reflect body surface, so-called " to have Zhu Neibi shape
It is all outer ".Therefore, Chinese medicine passes through the examination to internal organ body surface biological characteristic, it will be appreciated that the state change of the five internal organs is examined as Chinese medicine
The basis of disconnected disease and foundation.And body surface biological characteristic showing in body surface organ, have not in different theories of traditional Chinese medical science
The blood network category heart, black eye benevolence category liver, the upper palpebra inferior category spleen, white of the eye category lung, pupil that size canthus is used in same description, such as " five wheel theories "
Hole belongs to kidney positioning and divides the five internal organs body surface biological characteristic presentation;" the The Yellow Emperor's Canon of Internal Medicine Ling Shu Miraculous Pivot or Divine Axis five colors " are special by the five internal organs body surface biology
The region that sign is presented on face is the nasion category heart, bridge of the nose category liver, nose category spleen, place between the eyebrows category lung, chin subordinate's kidney;" The Yellow Emperor's Canon of Internal Medicine
Plain Questions thorn is warm " the five internal organs body surface biological characteristic reflecting regional of face is set to the forehead category heart, left cheek category liver, nose category spleen, the right side
Cheek category lung, chin subordinate's kidney.Thus the present invention extracts 68 key points of every frame facial image using the library DLIB, in conjunction with above-mentioned Chinese medicine
Observation is theoretical and the HOUGH transformation algorithm of artificial vision's intellectual technology, Delaunay Triangulation criterion position and divide its eye
Portion the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, facial Plain Questions pierce hot the five internal organs region.
Wherein, in step s3, the eye the five internal organs region of every frame facial image is determined as follows:
S31, the eyebrow key point using in 68 key points of every frame facial image, upper palpebra inferior key point and interior
Tail of the eye key point determines ocular and its canthus region, palpebral region, and converts positioning using HOUGH and divide the eye
Eye benevolence region, white of the eye region and pupil region in portion's area image, with the eye the five internal organs region of formation " five wheel theories ".
Further, the facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs region is determined as follows:
S32, using 68 key points of every frame facial image as triangulation point set to carry out Delaunay triangle
Subdivision obtains triangle gridding;
S33, the nasion key point according in 68 key points of every frame facial image, bridge of the nose key point, nose close
Key point, brows key point and lower lip key point position from corresponding triangle gridding respectively and are partitioned into nasion region, bridge of the nose
Region, nose region, place between the eyebrows region and chin portion region, to form facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs region.
Further, the five internal organs region of the facial Plain Questions thorn heat is determined as follows:
S34, the brows key point according in 68 key points of every frame facial image, left nose point key point, nose
Key point, right nose key point and lower lip key point position and be partitioned into from corresponding triangle gridding respectively forehead region,
Hot the five internal organs region is pierced to form Plain Questions in left cheek region, nasal region, right cheek region and chin portion region.
In step s3, it is extracted using 68 key points of the library DLIB to every frame facial image, in conjunction with Chinese medicine
The HOUGH of observation theory and artificial vision's intellectual technology converts algorithm, Delaunay Triangulation criterion, to position and divide respectively
Hot the five internal organs region is pierced in the eye the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, facial Plain Questions for cutting out every frame facial image,
The accuracy of image segmentation can be effectively improved, and then promotes the accuracy of the five internal organs body surface living things feature recognition.However, Chinese medicine is logical
Observation is crossed by point reflection energy and blood of human body prosperity and decline, viscera function and the feelings for experiencing disease and evil of the normal color sickly complexion of body surface biological characteristic for the person of examining
Condition, the sickly complexion are green, red, yellow, white, the black five colors.Equally, visual diagnostics think the five internal organs body surface biological characteristic table in the form of multicolored
On present each the five internal organs region of body surface organ, including overall face.Therefore, present invention combination visual diagnostics extract facial and each
Multicolored the five internal organs feature in the five internal organs region, and theoretical transformation is interlocked into facial and each the five internal organs region the five internal organs surface characteristics by sickly complexion
Collection.Wherein, the five internal organs surface characteristics be illness present, be general, deepening, is serious, very serious 5 grades.Step specifically includes:
S4, according to every frame facial image septum reset, eye the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, face
Plain Questions pierces the color-ratio in hot the five internal organs region, determines five color characteristics in face and each the five internal organs region, and staggeredly manage by sickly complexion
By by face and each the five internal organs region multicolored Feature Conversion at the five internal organs surface characteristics collection;Five color characteristic is used to indicate each five
The color in dirty district domain is one of red, cyan, yellow, white or black;
Wherein, in step s 4, five color characteristics determine as follows:
S41, each pixel is calculated according to the R component of the face and each each pixel in the five internal organs region, G component, B component
RGB average value;
S42, the face according to the RGB mean value calculation of each pixel and the red ratio in each the five internal organs region
Example, cyan ratio, yellow ratio, white proportion and black ratio, to determine the maximum color of ratio as five color characteristics of image.
For example, setting face-image is concentrated comprising N facial images, then calculated for each pixel in every facial image
Its RGB average value, since the R component of each pixel, G component, B component have 256 (0~255) grade gray values, and then being capable of root
The color that the pixel is accurately determined according to the RGB average value of each pixel may then pass through under type such as and calculate every face figure
Red ratio, cyan ratio, yellow ratio, white proportion and black ratio as in: red pixel quantity/sum of all pixels, cyan
Pixel quantity/sum of all pixels, yellow pixel quantity/sum of all pixels, white pixel quantity/sum of all pixels, black picture element quantity/as
Plain sum, the last set maximum color of ratio value are five color characteristics of facial image.For example, when determining in a facial image
When cyan ratio maximum, then cyan is five color characteristic of face of this facial image.
Illustrate the forming process of five color characteristic of eye, the facial Ling Shu Miraculous Pivot or Divine Axis five colors five colors by taking ocular image set as an example below
Feature is similar with the facial hot formation of five color characteristics of Plain Questions thorn, and this will not be repeated here.
For example, each five color characteristic of eye includes five color characteristics in five eye the five internal organs regions, i.e. the canthus area of heart
Five color characteristic of domain, five color characteristic of eye benevolence region of liver, five color characteristic of palpebral region of spleen, the white of the eye region five colors of lungs are special
Five color characteristic of pupil region of sign, kidney, accurately determines the pixel according to the RGB average value of each pixel in each the five internal organs region
Color, then calculate in each the five internal organs region red ratio, cyan ratio, yellow ratio, white proportion and black ratio, into
And setting ratio is worth five color characteristics that maximum color is corresponding region image, it is possible thereby to determine the five colors in eye the five internal organs region
Feature:
In formula i be indicate i-th facial image, and
xThe i hearts、xI livers、xI spleens、xI lungs、xI kidneysIt isIt is corresponded to respectively in five color characteristics of eye maximum in this characteristic value of the five internal organs region
Color-ratio value.
It should be understood that the facial multicolored feature set of the Ling Shu Miraculous Pivot or Divine Axis five colors may be set to:
The facial multicolored feature set of Plain Questions thorn heat may be set to:
For this purpose, it is theoretical to be interlocked according to Chinese medicine sickly complexion, by special into the five internal organs body surface according to five color characteristics are converted as shown in table 1 below
Sign:
1 sickly complexion of table is interlocked table
It is also to say, it is assumed that facial and each the five internal organs region is presented below as correspondence:
Cyan is presented in liver area, red, spleen region presentation yellow is presented in heart area, white is presented in lungs region,
Black is presented in kidney region, and the five internal organs surface characteristics are illness presentation;
Black is presented in liver area, cyan is presented in heart area, red, lungs region presentation yellow is presented in spleen region,
White is presented in kidney region, and the five internal organs surface characteristics are that illness is general;
Red is presented in liver area, yellow is presented in heart area, white is presented in spleen region, black is presented in lungs region,
Cyan is presented in kidney region, and the five internal organs surface characteristics are illness intensification;
Yellow is presented in liver area, white is presented in heart area, black is presented in spleen region, cyan is presented in lungs region,
Red is presented in kidney region, and the five internal organs surface characteristics are disease serious;
White is presented in liver area, black is presented in heart area, cyan is presented in spleen region, the presentation of lungs region is red,
Yellow is presented in kidney region, and the five internal organs surface characteristics are that illness is very serious;
S5, each the five internal organs surface characteristics collection is sent into preset SVM multi-categorizer, so that the SVM multi-categorizer is known
Not Chu the five internal organs status information, realize face-image analysis.
N number of overall face the five internal organs surface characteristics, N number of eye the five internal organs region the five internal organs surface characteristics, N number of face spirit are collected respectively
Pivot five colors the five internal organs region the five internal organs surface characteristics, N number of facial Plain Questions pierce hot the five internal organs region the five internal organs surface characteristics as sample data, and
The sample data is normalized to obtain the five internal organs surface characteristics sample set;
Wherein,
a1,nN number of overall face the five internal organs surface characteristics after indicating normalization, as the first subsample in sample set P;
a2,nN number of eye the five internal organs region the five internal organs surface characteristics after indicating normalization, as the second increment in sample set P
This;
a3,nN number of facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region the five internal organs surface characteristics after indicating normalization, as in sample set P
Third subsample;
a4,nN number of facial Plain Questions after indicating normalization pierces hot the five internal organs region the five internal organs surface characteristics, as in sample set P
4th subsample;
Wherein, n=1,2,3 ..., N;
In step s 5, the building of SVM classifier is specific as follows:
S511, dimensionality reduction is carried out to each the five internal organs surface characteristics sample set by principal component analysis, so that the sample in sample set P
Notebook data reduces half, obtains sample set P1;
S512, training set D of the sample data as SVM classifier that 70% is chosen from the sample set P1 after dimensionality reduction,
Remaining 30% sample data as test set T;
S513, setting training set D={ (x1,l1), (x2,l2) ..., (xm,lm), wherein D ∈ P1 li∈{-1,1}xi
It is sample data, liIt is sample labeling;
S514, assume that the training set can be by a hyperplane ωTX+b=0 linear partition, wherein ω is to determine hyperplane
Normal vector, b is the position of origin Yu hyperplane distance, then problem be converted into optimize hyperplane problem:
Wherein, ξiIt is 0/1 loss function, c is punishment parameter;
S55, for Nonlinear separability situation, select RBF Radial basis kernel functionThen
The optimization problem of SVM classifier is eventually converted into the select permeability of parameter (C, δ);Wherein, δ >=0 is the width of Radial basis kernel function
Parameter is spent, for controlling the radial effect range of Radial basis kernel function.
Specifically, in step s 5, train classification models are constructed as follows:
S521, with 2-10≤c≤27With 2-10≤δ≤23For all c and δ that range, step pitch are in 0.1 building value range
The parameter pair of composition;
S522, initial value of the parameter to (C, δ) as the SVM classifier parameter based on RBF Radial basis kernel function is successively taken,
Every group of parameter is obtained using cross validation K-CV method, and classification accuracy is verified to training set D under (C, δ);
S523, parameter corresponding to highest classification accuracy is chosen to (Co, δo) it is used as optimal parameter, obtain required SVM
The model parameter of disaggregated model.
Further, model prediction is carried out as follows in step s 5:
Model parameter (Co, the δ obtained using above-mentioned stepso) svm classifier prediction is carried out to test set T, obtain the person of being examined
The five internal organs status information.
In step s 5, as shown in Fig. 2, SVM multi-categorizer is provided with 10 bis- classifiers of SVM, first layer 1 classification altogether
The sample of classification non-1 is positioned positive sample by device, and the sample of classification non-5 positions negative sample, 2 classifiers of the second layer, along non-1 classification
The sample of non-2 classification is set to positive sample, non-5 sample is set to negative sample, second classifier is obtained, in this way, available
10 binary classifiers.
The embodiment of the present invention also provides a kind of face-image analytical equipment based on Chinese medicine facial diagnosis, which contains computer
Program, the computer program are suitable for being performed to realize any one of the above face-image analysis method.
The embodiment of the present invention also provides another face-image analytical equipment based on Chinese medicine facial diagnosis, as shown in figure 3, it is wrapped
The memory 2 for including processor 1, connecting with the processor 1, the memory 2 are stored with computer program, the computer journey
Sequence is suitable for being performed to realize any one of the above face-image analysis method.
Compared with prior art, the face-image analysis method and device of the invention based on Chinese medicine facial diagnosis has following
The utility model has the advantages that
(1) eye the five internal organs feature, face five are obtained by the way that theory of traditional Chinese medical science and artificial vision's intellectual technology to be combined
Dirty feature, facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs feature, facial Plain Questions pierce hot the five internal organs feature, then using SVM multi-categorizer to the five internal organs feature into
Row classification judgement, realizes the five internal organs status information quantization analysis and identification, can effectively improve the standard of the analysis of the five internal organs status information
Exactness and analysis efficiency.
(2) the hot theory of traditional Chinese medical science of five wheel five colors theories, the Ling Shu Miraculous Pivot or Divine Axis five colors, Plain Questions thorn is referred and synthesized to set about from Chinese medicine facial diagnosis to the five internal organs shape
State carries out quantitative analysis, provides objective basis for auxiliary clinical syndrome differentiation diagnosis;
(3) mesh is examined into five wheel five colors theories and facial diagnosis is added, the content of abundant facial diagnosis more objectively reflects five to more acurrate
Dirty situation.
The above described is only a preferred embodiment of the present invention, limitation in any form not is done to the present invention, therefore
All contents without departing from technical solution of the present invention, it is made to the above embodiment according to the technical essence of the invention any simply to repair
Change, equivalent variations and modification, all of which are still within the scope of the technical scheme of the invention.
Claims (10)
1. a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method characterized by comprising
Possessed by HAAR feature and ADABOOST algorithm from input computer and obtains facial image in the video image of face;
68 key points of every frame facial image are extracted, using the library DLIB to cut open using HOUGH transformation algorithm, Delaunay triangle
The positioning of point criterion and be partitioned into its eye the five internal organs region, facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region, facial Plain Questions pierce hot the five internal organs region;
Five color characteristics in every frame facial image septum reset and its each the five internal organs region are extracted, and are converted into corresponding the five internal organs body surface
Feature set;The five colors are used to indicate the body surface sickly complexion that traditional Chinese medicine defines, and five color characteristic is used to indicate every image
The pigment ratio of middle body surface sickly complexion, the five internal organs surface characteristics are used to indicate the five internal organs state grade of every image;
Each the five internal organs surface characteristics collection is sent into preset SVM multi-categorizer, to identify the five internal organs status information.
2. examining survey method as described in claim 1, which is characterized in that determine every frame facial image as follows
Eye the five internal organs region:
It is closed using eyebrow key point, upper palpebra inferior key point and the interior tail of the eye in 68 key points of every frame facial image
Key point determines ocular image and its canthus region, palpebral region, and converts positioning using HOUGH and divide the eye area
Eye benevolence region, white of the eye region and pupil region in area image, to form eye the five internal organs region.
3. examining survey method as described in claim 1, which is characterized in that determine the facial Ling Shu Miraculous Pivot or Divine Axis five colors five as follows
Dirty district domain:
Using 68 key points of every frame facial image as triangulation point set to carry out Delaunay Triangulation, obtain
Triangle gridding;
According to nasion key point, the bridge of the nose key point, nose key point, brows in 68 key points of every frame facial image
Key point and lower lip key point position from corresponding triangle gridding respectively and are partitioned into nasion region, bridge of the nose region, nose
Region, place between the eyebrows region and chin portion region, to form facial Ling Shu Miraculous Pivot or Divine Axis five colors the five internal organs region.
4. examining survey method as claimed in claim 3, which is characterized in that determine the facial Plain Questions thorn heat five as follows
Dirty district domain:
According to brows key point, the left nose point key point, nose key point, the right side in 68 key points of every frame facial image
Nose key point and lower lip key point position from corresponding triangle gridding respectively and are partitioned into forehead region, left cheek area
Domain, nasal region, right cheek region and chin portion region ask thorn hot the five internal organs region to form face.
5. examining survey method as described in claim 1, which is characterized in that five color characteristics in each region are as follows
It determines:
The RGB average value of each pixel is calculated according to the R component of each each pixel in region, G component, B component;
According to the red ratio in each region described in the RGB mean value calculation of each pixel, cyan ratio, yellow ratio
Example, white proportion and black ratio, to determine that the maximum color of ratio is five color characteristics in each region.
6. examining survey method as claimed in claim 5, which is characterized in that each the five internal organs surface characteristics collection is converted as follows
It obtains:
Interlocked the corresponding relationship between theoretical transformation table according to five color characteristic of each region the five internal organs and preset sickly complexion, determine its five
Dirty surface characteristics collection;The five internal organs feature includes eye the five internal organs feature, facial the five internal organs feature, facial Ling Shu Miraculous Pivot or Divine Axis the five internal organs feature, face
Plain Questions pierces hot the five internal organs feature.
7. examining survey method as described in claim 1, which is characterized in that the facial video image is imaged by two high-resolution
The binocular camera of machine composition collects.
8. as described in claim 1 examines survey method, which is characterized in that identify the five internal organs status information as follows:
According to each the five internal organs surface characteristics collection building training set and test set;
The SVM multi-categorizer is constructed by the training set;
Construct train classification models;
Model prediction is carried out to the test set by the train classification models of building, obtains the five internal organs status information.
9. a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey device, which is characterized in that in artificial vision's intelligence
Doctor's five dirty situation of facial diagnosis examines survey device and is stored with computer program, and the computer program is suitable for being performed to realize as right is wanted
Five dirty situations described in asking any one of 1~8 examine survey method.
10. a kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey device, which is characterized in that including processor, with it is described
The memory of processor connection, the memory are stored with computer program, and the computer program is suitable for being performed to realize
As five dirty situation according to any one of claims 1 to 8 examines survey method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910599178.5A CN110334649A (en) | 2019-07-04 | 2019-07-04 | A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910599178.5A CN110334649A (en) | 2019-07-04 | 2019-07-04 | A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110334649A true CN110334649A (en) | 2019-10-15 |
Family
ID=68144153
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910599178.5A Pending CN110334649A (en) | 2019-07-04 | 2019-07-04 | A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110334649A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111191574A (en) * | 2019-12-26 | 2020-05-22 | 新绎健康科技有限公司 | Method and device for acquiring viscera partition temperature of facial examination |
CN111227795A (en) * | 2020-01-13 | 2020-06-05 | 天津众阳科技有限公司 | Method for extracting color change information of human face reflection area |
CN113140309A (en) * | 2021-04-14 | 2021-07-20 | 五邑大学 | Traditional Chinese medicine complexion diagnosis method and device |
CN114445895A (en) * | 2022-01-28 | 2022-05-06 | 上海掌门科技有限公司 | Method, device, medium and program product for determining face color |
CN116030042A (en) * | 2023-02-24 | 2023-04-28 | 智慧眼科技股份有限公司 | Diagnostic device, method, equipment and storage medium for doctor's diagnosis |
CN117809841A (en) * | 2024-02-23 | 2024-04-02 | 四川大学华西医院 | Skin special patient management method and system based on large model technology |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102270308A (en) * | 2011-07-21 | 2011-12-07 | 武汉大学 | Facial feature location method based on five sense organs related AAM (Active Appearance Model) |
CN102426652A (en) * | 2011-10-10 | 2012-04-25 | 北京工业大学 | Traditional Chinese medicine face color identifying and retrieving method based on image analysis |
CN106971147A (en) * | 2017-03-06 | 2017-07-21 | 武汉嫦娥医学抗衰机器人股份有限公司 | A kind of traditional Chinese medical science facial diagnosis system and facial diagnosis method split based on human face region |
CN107153805A (en) * | 2016-03-02 | 2017-09-12 | 北京美到家科技有限公司 | Customize makeups servicing unit and method |
CN109300123A (en) * | 2018-09-18 | 2019-02-01 | 天津大学 | Physical analysis method based on face and tongue Image Acquisition |
-
2019
- 2019-07-04 CN CN201910599178.5A patent/CN110334649A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102270308A (en) * | 2011-07-21 | 2011-12-07 | 武汉大学 | Facial feature location method based on five sense organs related AAM (Active Appearance Model) |
CN102426652A (en) * | 2011-10-10 | 2012-04-25 | 北京工业大学 | Traditional Chinese medicine face color identifying and retrieving method based on image analysis |
CN107153805A (en) * | 2016-03-02 | 2017-09-12 | 北京美到家科技有限公司 | Customize makeups servicing unit and method |
CN106971147A (en) * | 2017-03-06 | 2017-07-21 | 武汉嫦娥医学抗衰机器人股份有限公司 | A kind of traditional Chinese medical science facial diagnosis system and facial diagnosis method split based on human face region |
CN109300123A (en) * | 2018-09-18 | 2019-02-01 | 天津大学 | Physical analysis method based on face and tongue Image Acquisition |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111191574A (en) * | 2019-12-26 | 2020-05-22 | 新绎健康科技有限公司 | Method and device for acquiring viscera partition temperature of facial examination |
CN111227795A (en) * | 2020-01-13 | 2020-06-05 | 天津众阳科技有限公司 | Method for extracting color change information of human face reflection area |
CN111227795B (en) * | 2020-01-13 | 2022-07-22 | 天津众阳科技有限公司 | Method for extracting color change information of human face reflection area |
CN113140309A (en) * | 2021-04-14 | 2021-07-20 | 五邑大学 | Traditional Chinese medicine complexion diagnosis method and device |
CN114445895A (en) * | 2022-01-28 | 2022-05-06 | 上海掌门科技有限公司 | Method, device, medium and program product for determining face color |
CN116030042A (en) * | 2023-02-24 | 2023-04-28 | 智慧眼科技股份有限公司 | Diagnostic device, method, equipment and storage medium for doctor's diagnosis |
CN117809841A (en) * | 2024-02-23 | 2024-04-02 | 四川大学华西医院 | Skin special patient management method and system based on large model technology |
CN117809841B (en) * | 2024-02-23 | 2024-04-30 | 四川大学华西医院 | Skin special patient management method and system based on large model technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110334649A (en) | A kind of five dirty situation of artificial vision's intelligence Chinese medicine facial diagnosis examines survey method and device | |
Bartlett et al. | Classifying facial action | |
CN103793690B (en) | A kind of human-body biological biopsy method detected based on subcutaneous haematic flow and application | |
CN110084259A (en) | A kind of facial paralysis hierarchical synthesis assessment system of combination face texture and Optical-flow Feature | |
CN110335266B (en) | Intelligent traditional Chinese medicine visual inspection image processing method and device | |
CN110338763A (en) | A kind of intelligence Chinese medicine examines the image processing method and device of survey | |
CN106845328B (en) | A kind of Intelligent human-face recognition methods and system based on dual camera | |
Matuszewski et al. | Hi4D-ADSIP 3-D dynamic facial articulation database | |
Huan et al. | Deep convolutional neural networks for classifying body constitution based on face image | |
CN109508755B (en) | Psychological assessment method based on image cognition | |
CN106778047A (en) | A kind of traditional Chinese medical science facial diagnosis integrated system based on various dimensions medical image | |
Chetty et al. | A multilevel fusion approach for audiovisual emotion recognition | |
CN106725341A (en) | A kind of enhanced lingual diagnosis system | |
CN113946217B (en) | Intelligent auxiliary evaluation system for enteroscope operation skills | |
Matuszewski et al. | High-resolution comprehensive 3-D dynamic database for facial articulation analysis | |
CN117935339A (en) | Micro-expression recognition method based on multi-modal fusion | |
US20100150405A1 (en) | System and method for diagnosis of human behavior based on external body markers | |
Liu et al. | Quantitative assessment of facial paralysis using infrared thermal imaging | |
Liu et al. | Baseline-independent stress classification based on facial StO2 | |
CN109711306B (en) | Method and equipment for obtaining facial features based on deep convolutional neural network | |
Li et al. | Computer-aided disease diagnosis system in TCM based on facial image analysis | |
CN110164547A (en) | A kind of Chinese medicine facial diagnosis system based on human face region and tongue fur | |
CN115713800A (en) | Image classification method and device | |
KR100915922B1 (en) | Methods and System for Extracting Facial Features and Verifying Sasang Constitution through Image Recognition | |
KR20160127296A (en) | Face image analysis method for sasang constituition diagnosis |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191015 |
|
RJ01 | Rejection of invention patent application after publication |