CN108564113A - A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity - Google Patents
A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity Download PDFInfo
- Publication number
- CN108564113A CN108564113A CN201810259304.8A CN201810259304A CN108564113A CN 108564113 A CN108564113 A CN 108564113A CN 201810259304 A CN201810259304 A CN 201810259304A CN 108564113 A CN108564113 A CN 108564113A
- Authority
- CN
- China
- Prior art keywords
- tongue fur
- complexity
- picture
- neural network
- tongue
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/90—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to alternative medicines, e.g. homeopathy or oriental medicines
-
- 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
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
Abstract
The invention discloses a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity, including：Tongue fur image data is acquired using collecting device；The accurate location of tongue fur is found in tongue fur picture；Grader of the design based on deep neural network is trained tongue fur picture, to extract the feature of tongue fur picture；Use complexity perception algorithm, the complexity of tongue fur data is considered in individual level, to which tongue fur data are divided into easy data set and complex data collection, a complexity discriminator is trained using the two data sets, easy grader and complex classifier is respectively trained with the two data sets again, tongue fur feature is finally input to complexity discriminator to judge the complexity of tongue fur sample；Finally tongue fur feature is input in corresponding softmax graders according to the complexity of tongue fur sample to obtain corresponding constitution classification.The method accurately and rapidly can carry out constitution identification to tongue fur, reduce the workload of doctor, alleviate medical pressure.
Description
Technical field
The present invention relates to the tongue fur constitution knowledges of image recognition, data mining technology and traditional Chinese medical science field in computer realm
Not, and in particular to a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity.
Background technology
In China, tcm constitution theory sees China Shanxi generation earliest《37 biographies the 7th of Shanxi books》One book passes through
Prolonged development and abundant, it is established that perfect tcm constitution theoretical system.On middle Biography of Medical Figures, constitution is described as in people
The congenital and day after tomorrow day in life process is assigned to the form to be formed, and is the general performance of physiological function and psychological condition.《Chinese medicine body
Qualitative classification and criterion》Definition：Tcm constitution refers in human vital process, in native endowment and acquired basis
On be formed by morphosis, physiological function and psychological condition various aspects synthesis, metastable intrinsic speciality；It is that the mankind exist
The human body personal characteristics adaptable with nature, social environment formed in growth, growth course.The classification for studying constitution, must
Extensive comparative analysis must be carried out to complicated phenomenon of constitution, then screen classification, hold the constitution difference rule and body of individual
Matter feature.《The classification of 9 kinds of basic constitution type in TCM and its diagnosis statement foundation》In conjunction with clinical observation and ancient times and modern times
Tcm constitution is divided into gentle matter, deficiency of vital energy matter, deficiency of yang matter, deficiency of Yin matter, phlegm wet matter, damp and hot matter, the stasis of blood by the related understanding of classification of TCM constitution
9 kinds of fundamental types such as blood matter, obstruction of the circulation of vital energy matter, special official report matter, comprehensively reflect somatotypes.Traditional classification of TCM constitution relies primarily on each
The characteristics of constitution, is screened, and is screened one by one, what main criterion was put into effect with reference to 2009 years《Traditional Chinese Medicine Constitution Classification with
Judgement》.For example, deficiency of yang matter is mainly shown as fears cold, hands and feet being not warm usually, hot drink preference food, lassitude, sweet tea is light fat tender, deeptensepulse
Late；Deficiency of Yin matter, which is mainly shown as, defends feverishness in palms and soles, dryness of the mouth and throat, and nose is micro- dry, and cool drink is hard and dry, red tongue with a little fluid, and arteries and veins is thin
Number；It is that skin of face grease is more that phlegm wet matter, which is mainly cashed, hidrosis and glutinous, uncomfortable in chest, abundant expectoration, sticky and greasy in mouth or sweet tea, preference for fat and sweet food sweet tea
Glutinous, tongue is greasy, slippery.
Lingual diagnosis is that traditional Chinese medical science field is important and unique content, is one of the important content of observation and diagnosis.Lingual diagnosis is by Chinese medicine
Clinical workers is used for observing body interior physiology and pathological change at least 3000.Experienced doctor of traditional Chinese medicine is mainly with the naked eye
Come information such as color, texture, the forms of observing tongue to judge the physical condition of patient.Tongue is the seedling of the heart, is waited except spleen, tongue
It is given birth to by stomach Qi.Internal organs are the collateral of heart meridian system tongue sheet by the be connected important visitor of tcm diagnosis disease of passages through which vital energy circulates and tongue, the few the moon of foot
Arteries and veins hold tongue sheet under the arm, the train of thought tongue sheet of the moon of fainting enough, the lunar arteries and veins of foot connects tongue sheet, dissipates sublingual, therefore internal organs lesion, can be in tongue nature and tongue fur
On reflect, lingual diagnosis is mainly examined the form of tongue nature and tongue fur, color and luster, is moisturized, and judges the property of disease, patient's condition with this
Shallow depth, the prosperity and decline of qi and blood, the profit and loss of body fluid and internal organs actual situation etc..
Computer vision is an important research direction in computer realm, and target detection and image recognition are computers
Two vital tasks in vision, with the rapid development of deep learning, target detection also achieves huge with image recognition
Progress.Deep learning is a kind of method carrying out representative learning based on data in machine learning, it is by combining low-level image feature shape
It is indicated at abstract high level.Its essence is the multi-layer perception (MLP)s for including multiple hidden layers, learn to arrive data automatically by reverse conduction
Character representation, to promote the accuracy rate of classification.In machine learning, the performance of learning algorithm depends on its parameter and instruction
Practice data.However, the emphasis of most of algorithm development is all adjustment model parameter, without the number for going to consider to be used to training pattern
According to.It goes to consider a problem from the angle of data, considers the complexity of sample in individual level, understand data by the original of mistake classification
Cause, to promote the accuracy rate of grader.The relationship between tongue fur feature and constitution is excavated using artificial intelligence, is provided for patient
Convenient, fast and accurate constitution identification, alleviates medical pressure, additionally it is possible to develop Traditional Chinese Medicine.
Invention content
It being based on deep neural network and complexity sense in view of the deficiencies of the prior art, it is an object of the present invention to provide a kind of
The tongue fur constitution recognition methods known, the method accurately and rapidly can carry out constitution identification to tongue fur, be carried for the diagnosis of doctor
For assisting foundation, the workload of doctor is reduced, medical pressure is alleviated.
The purpose of the present invention can be achieved through the following technical solutions：
A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity, the method includes following steps
Suddenly：
S1, tongue fur image data is acquired as input using collecting device；
S2, the accurate location that tongue fur is found in the tongue fur picture of input；
S3, grader of the design based on deep neural network are trained tongue fur picture, to extract tongue fur picture
Feature；
S4, complexity perception algorithm, the complexity of consideration tongue fur data in individual level, according to step S3 extractions are used
Tongue fur picture is divided into easy data set and complex data collection by the feature of tongue fur picture, uses the training one jointly of the two data sets
A complexity discriminator, then easy grader and complex classifier is respectively trained with the two data sets, finally by tongue fur picture
Feature be input to complexity discriminator to judge the complexity of tongue fur picture；
S5, the feature of tongue fur picture is input in corresponding grader to obtain phase according to the complexity of tongue fur picture
The constitution classification answered.
Further, in step S1, picture collection is carried out to the tongue fur of patient using the image capture device of non-intrusion type
Work, each tongue fur picture have all corresponded to one kind in nine kinds of somatotypes in Chinese medicine, and to every tongue fur figure of acquisition
Piece all carries out color and goes mean value and normalized.
Further, in step S2, tongue is found using the Faster R-CNN algorithm of target detection based on VGG models first
Then the general location of tongue uses the calibration network based on depth convolutional neural networks, calibrate detection zone, precise positioning tongue fur
Position.
Further, it is trained using deep neural network as model in step S3, and with trained depth
Neural network extracts the feature of tongue fur picture as feature extractor.
Further, the easy grader and complex classifier of training are softmax graders in step S4, training
Complexity discriminator is logistic regression classifier.
Compared with prior art, the present invention having the following advantages that and advantageous effect：
1, the present invention is based on the tongue fur constitution recognition methods that deep neural network and complexity perceive, using depth nerve net
Network carries out target detection and calibration, can quickly locate the position of tongue fur, reduce remaining background parts and done to classification results
It disturbs.
2, the present invention is based on the tongue fur constitution recognition methods that deep neural network and complexity perceive, using depth nerve net
Network extracts the feature of tongue fur picture, can learn and extract different constitution class another characteristics automatically.
3, it the present invention is based on the tongue fur constitution recognition methods that deep neural network and complexity perceive, is perceived using complexity
Algorithm considers the complexity issue of sample, due to tongue fur that illumination, angle, resolution ratio are brought when solving tongue fur picture collection
Picture quality imbalance problem reduces the interference of tongue fur picture from each other in different distributions.
4, the present invention is based on the tongue fur constitution recognition methods that deep neural network and complexity perceive, it is based on computer technology
With artificial intelligence, integrated tongue fur constitution identification framework is constructed, is identified compared to traditional tcm constitution, reduces diagnosis
Time, improve the accuracy rate of constitution identification, the method can be deployed in the equipment of hospital, for patient provide it is convenient soon
Prompt constitution identifies channel, and additional foundation is provided for the diagnosis of doctor, can alleviate the medical pressure of current anxiety, have one
Fixed market value and promotional value.
Description of the drawings
Fig. 1 is the flow for the tongue fur constitution recognition methods that the embodiment of the present invention is perceived based on deep neural network and complexity
Figure.
Fig. 2 is the flow chart of complexity perception algorithm of the embodiment of the present invention.
Specific implementation mode
Present invention will now be described in further detail with reference to the embodiments and the accompanying drawings, but embodiments of the present invention are unlimited
In this.
Embodiment：
A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity is present embodiments provided, it is described
The flow chart of method is as shown in Figure 1, include the following steps：
S1, tongue fur image data is acquired as input using collecting device；
The tongue fur picture is all carried out the mark of constitution by several veteran doctors according to same standard, according to machine
The tongue fur image data of acquisition is divided into training set, verification collection and three parts of test set, every tongue fur figure by the method for device study
Piece all goes mean value and normalized by color.
S2, the accurate location that tongue fur is found in the tongue fur picture of input；
The general location of tongue fur is found using the Faster R-CNN algorithm of target detection based on VGG models first, then
Using the calibration network based on depth convolutional neural networks, detection zone, precise positioning tongue fur position are calibrated.In precise positioning tongue
During tongue position, x offsets, y offset and the scaling s between the tongue fur position and true tongue fur position of detection are judged,
Specific setting is as follows：
xn∈ { -0.17,0,0.17 }, yn∈ { -0.17,0,0.17 }, sn∈{0.83,0.91,1.0,1.10,1.21}
After the processing of step S1, S2, tongue fur region is cut, and scales, dimension 224*224*3.
S3, grader of the design based on deep neural network are trained tongue fur picture, to extract tongue fur picture
Feature；
For the deep neural network used for VGG-16 models, effect is that important tongue fur letter is extracted from tongue fur picture
The feature of breath, detailed design are as shown in table 1：
| Layer name | Parameter | Output size |
| Conv_1 | 3*3,64, padding=same, activation=ReLU | 224*224*64 |
| Conv_2 | 3*3,64, padding=same, activation=ReLU | 224*224*64 |
| MaxPooling_1 | 2*2, strides=2*2 | 112*112*64 |
| Conv_3 | 3*3,128, padding=same, activation=ReLU | 112*112*128 |
| Conv_4 | 3*3,128, padding=same, activation=ReLU | 112*112*128 |
| MaxPooling_2 | 2*2, strides=2*2 | 56*56*128 |
| Conv_5 | 3*3,256, padding=same, activation=ReLU | 56*56*256 |
| Conv_6 | 3*3,256, padding=same, activation=ReLU | 56*56*256 |
| Conv_7 | 3*3,256, padding=same, activation=ReLU | 56*56*256 |
| MaxPooling_3 | 2*2, strides=2*2 | 28*28*256 |
| Conv_8 | 3*3,512, padding=same, activation=ReLU | 28*28*512 |
| Conv_9 | 3*3,512, padding=same, activation=ReLU | 28*28*512 |
| Conv_9 | 3*3,512, padding=same, activation=ReLU | 28*28*512 |
| MaxPooling_4 | 2*2, strides=2*2 | 14*14*512 |
| Conv_10 | 3*3,512, padding=same, activation=ReLU | 14*14*512 |
| Conv_11 | 3*3,512, padding=same, activation=ReLU | 14*14*512 |
| Conv_12 | 3*3,512, padding=same, activation=ReLU | 14*14*512 |
| Ave_Pooling_1 | global pooling | 512 |
| Dense_1 | 1024 | 1024 |
Table 1
Further, step S3 is specifically divided into two stages of training and feature extraction of tongue fur picture, wherein training stage
Detailed process be：
[1] the big of 224*224 will be scaled by tongue fur region detection, calibration and the training set of cutting and verification collection picture
It is small；
[2] it uses training set to train deep neural network model, using the weight of back-propagation algorithm adjusting parameter, utilizes
Mini-batch methods select training data training, carry out successive ignition；
[3] output that softmax graders are last as deep neural network is used；
[4] when training, after often taking turns iteration, performance of the test depth neural network model on verification collection；
[5] more wheel iteration are eventually passed through, the deep neural network model to behave oneself best on verification collection is chosen at and is used as most
Whole deep neural network model.
The detailed process of wherein feature extraction phases is：
[1] the big of 224*224 will be scaled by tongue fur region detection, calibration and the training set of cutting and verification collection picture
It is small；
[2] all tongue fur pictures are input in the trained deep neural network as feature extractor, are taken
Character representation of Dense_1 layers of the output as tongue fur picture, each tongue fur picture are expressed as the vector of one 1024 dimension.
S4, complexity perception algorithm, the complexity of consideration tongue fur data in individual level, thus by tongue fur picture are used
Be divided into easy data set and complex data collection, using the two data sets jointly training one complexity discriminator, then with this two
Easy grader and complex classifier is respectively trained in a data set, and the feature of tongue fur picture is finally input to complexity discriminator
To judge the complexity of tongue fur picture；
The input of step S4 is the tongue fur picture feature for all 1024 dimensions that step S3 is extracted, the complexity sense
The flow chart of algorithm is known as shown in Fig. 2, being as follows：
S41, training set is divided into k disjoint subsets, a subset is taken to train a softmax points every time in order
Class device, and correct situation of whole training sets on this softmax grader is tested, it repeats k times, counts each training sample
Correct number on this k grader is labeled as N (xi), if E_set is to be easy data set, C_set is complex data collection,
θ is set as complexity threshold, the present embodiment sets (k=150, θ=0.6), and design parameter setting can be adjusted according to actual conditions,
It is easy data set and complex data collection division methods is as follows：
S42, easy data set and the easy grader of complex data collection training and complex classifier are utilized respectively；
S43, a complexity discriminator is trained jointly using easy data set and complex data collection；
S44, the feature for the tongue fur picture that step S3 is extracted is input to complexity discriminator to judge answering for tongue fur picture
Miscellaneous degree.
Wherein, the easy grader of step S42 training and complex classifier are softmax graders, step S43 training
Complexity discriminator be logistic regression classifier.
S5, the feature of tongue fur picture is input in corresponding grader to obtain phase according to the complexity of tongue fur picture
The constitution classification answered.
The above, patent preferred embodiment only of the present invention, but the protection domain of patent of the present invention is not limited to
This, any one skilled in the art is in the range disclosed in patent of the present invention, according to the skill of patent of the present invention
Art scheme and its patent of invention design are subject to equivalent substitution or change, belong to the protection domain of patent of the present invention.
Claims (5)
1. a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity, which is characterized in that the method
Include the following steps：
S1, tongue fur image data is acquired as input using collecting device；
S2, the accurate location that tongue fur is found in the tongue fur picture of input；
S3, grader of the design based on deep neural network are trained tongue fur picture, to extract the feature of tongue fur picture；
S4, using complexity perception algorithm, the complexity of tongue fur data is considered in individual level, tongue fur is extracted according to step S3
Tongue fur picture is divided into easy data set and complex data collection by the feature of picture, and using the two data sets, training one is multiple jointly
Miscellaneous degree discriminator, then easy grader and complex classifier is respectively trained with the two data sets, finally by the spy of tongue fur picture
Sign is input to complexity discriminator to judge the complexity of tongue fur picture；
S5, the feature of tongue fur picture is input in corresponding grader to obtain accordingly according to the complexity of tongue fur picture
Constitution classification.
2. a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity according to claim 1,
It is characterized in that：In step S1, picture collection work is carried out to the tongue fur of patient using the image capture device of non-intrusion type, often
One tongue fur picture has all corresponded to one kind in nine kinds of somatotypes in Chinese medicine, and is all carried out to every tongue fur picture of acquisition
Color goes mean value and normalized.
3. a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity according to claim 1,
It is characterized in that：In step S2, the big of tongue fur is found using the Faster R-CNN algorithm of target detection based on VGG models first
Then body position uses the calibration network based on depth convolutional neural networks, calibrate detection zone, precise positioning tongue fur position.
4. a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity according to claim 1,
It is characterized in that：It is trained using deep neural network as model in step S3, and with trained depth nerve net
Network extracts the feature of tongue fur picture as feature extractor.
5. a kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity according to claim 1,
It is characterized in that：The easy grader and complex classifier of training are softmax graders, trained complexity in step S4
Discriminator is logistic regression classifier.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810259304.8A CN108564113A (en) | 2018-03-27 | 2018-03-27 | A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810259304.8A CN108564113A (en) | 2018-03-27 | 2018-03-27 | A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108564113A true CN108564113A (en) | 2018-09-21 |
Family
ID=63533461
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810259304.8A Pending CN108564113A (en) | 2018-03-27 | 2018-03-27 | A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108564113A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109259730A (en) * | 2018-10-09 | 2019-01-25 | 广东数相智能科技有限公司 | A kind of early warning analysis method and storage medium based on lingual diagnosis |
| CN109377479A (en) * | 2018-09-27 | 2019-02-22 | 中国电子科技集团公司第五十四研究所 | Satellite dish object detection method based on remote sensing image |
| CN109448817A (en) * | 2018-09-28 | 2019-03-08 | 小伍健康科技（上海）有限责任公司 | A kind of recipe recommendation method and apparatus based on deep neural network |
| CN109657722A (en) * | 2018-12-20 | 2019-04-19 | 山东农业大学 | Tongue fur image-recognizing method and system based on deep learning algorithm |
| CN109770875A (en) * | 2019-03-27 | 2019-05-21 | 上海铀米机器人科技有限公司 | A kind of human body constitution discrimination method and system based on neural network classifier |
| CN109801269A (en) * | 2018-12-29 | 2019-05-24 | 华南理工大学 | A kind of tongue fur classification of TCM constitution method of extruding competition-based and excitation neural network |
| CN109903836A (en) * | 2019-03-31 | 2019-06-18 | 山西慧虎健康科技有限公司 | A kind of diet intelligent recommendation and matching system and method based on constitution and big data |
| CN109994186A (en) * | 2019-03-31 | 2019-07-09 | 山西慧虎健康科技有限公司 | Tcm constitution intelligent measurement and maintenance system and method based on image big data |
| CN111476260A (en) * | 2019-11-22 | 2020-07-31 | 上海大学 | Putrefaction classification algorithm based on convolutional neural network |
| CN111528833A (en) * | 2020-05-07 | 2020-08-14 | 广州大学 | Method and system for quickly identifying and processing electrocardiosignals |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503479A (en) * | 2016-12-06 | 2017-03-15 | 郭福生 | A kind of Constitution Identification system and discrimination method |
| CN107610087A (en) * | 2017-05-15 | 2018-01-19 | 华南理工大学 | A kind of tongue fur automatic division method based on deep learning |
-
2018
- 2018-03-27 CN CN201810259304.8A patent/CN108564113A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503479A (en) * | 2016-12-06 | 2017-03-15 | 郭福生 | A kind of Constitution Identification system and discrimination method |
| CN107610087A (en) * | 2017-05-15 | 2018-01-19 | 华南理工大学 | A kind of tongue fur automatic division method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| JIAJIONG MA ET AL: "Tongue image constitution recognition based on Complexity Perception method", 《ARXIV:1803.00219V1[CS.CV]》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109377479B (en) * | 2018-09-27 | 2021-10-22 | 中国电子科技集团公司第五十四研究所 | Butterfly satellite antenna target detection method based on remote sensing image |
| CN109377479A (en) * | 2018-09-27 | 2019-02-22 | 中国电子科技集团公司第五十四研究所 | Satellite dish object detection method based on remote sensing image |
| CN109448817A (en) * | 2018-09-28 | 2019-03-08 | 小伍健康科技（上海）有限责任公司 | A kind of recipe recommendation method and apparatus based on deep neural network |
| CN109259730A (en) * | 2018-10-09 | 2019-01-25 | 广东数相智能科技有限公司 | A kind of early warning analysis method and storage medium based on lingual diagnosis |
| CN109657722A (en) * | 2018-12-20 | 2019-04-19 | 山东农业大学 | Tongue fur image-recognizing method and system based on deep learning algorithm |
| CN109801269A (en) * | 2018-12-29 | 2019-05-24 | 华南理工大学 | A kind of tongue fur classification of TCM constitution method of extruding competition-based and excitation neural network |
| CN109770875A (en) * | 2019-03-27 | 2019-05-21 | 上海铀米机器人科技有限公司 | A kind of human body constitution discrimination method and system based on neural network classifier |
| CN109903836A (en) * | 2019-03-31 | 2019-06-18 | 山西慧虎健康科技有限公司 | A kind of diet intelligent recommendation and matching system and method based on constitution and big data |
| CN109994186A (en) * | 2019-03-31 | 2019-07-09 | 山西慧虎健康科技有限公司 | Tcm constitution intelligent measurement and maintenance system and method based on image big data |
| CN111476260A (en) * | 2019-11-22 | 2020-07-31 | 上海大学 | Putrefaction classification algorithm based on convolutional neural network |
| CN111476260B (en) * | 2019-11-22 | 2023-07-21 | 上海大学 | Greasy fur classification algorithm based on convolutional neural network |
| CN111528833A (en) * | 2020-05-07 | 2020-08-14 | 广州大学 | Method and system for quickly identifying and processing electrocardiosignals |
| CN111528833B (en) * | 2020-05-07 | 2023-05-02 | 广州大学 | Rapid identification and processing method and system for electrocardiosignals |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108564113A (en) | A kind of tongue fur constitution recognition methods perceived based on deep neural network and complexity | |
| Zhang et al. | A seven-layer convolutional neural network for chest CT-based COVID-19 diagnosis using stochastic pooling | |
| CN108806792B (en) | Deep learning face diagnosis system | |
| Su et al. | Human identification using finger vein and ECG signals | |
| Tian et al. | Multi-path convolutional neural network in fundus segmentation of blood vessels | |
| CN108257135A (en) | The assistant diagnosis system of medical image features is understood based on deep learning method | |
| CN106683087A (en) | Coated tongue constitution distinguishing method based on depth neural network | |
| CN112132833B (en) | Dermatological image focus segmentation method based on deep convolutional neural network | |
| US9687199B2 (en) | Medical imaging system providing disease prognosis | |
| CN109493954A (en) | A kind of SD-OCT image retinopathy detection system differentiating positioning based on classification | |
| CN107886503A (en) | A kind of alimentary canal anatomical position recognition methods and device | |
| CN109754414A (en) | Image processing method, device, electronic equipment and computer readable storage medium | |
| CN107358605A (en) | For identifying the deep neural network and system of diabetic retinopathy | |
| CN109903836A (en) | A kind of diet intelligent recommendation and matching system and method based on constitution and big data | |
| CN105023023B (en) | A kind of breast sonography characteristics of image self study extracting method for computer-aided diagnosis | |
| CN109300123A (en) | Physical analysis method based on face and tongue Image Acquisition | |
| CN108492300A (en) | Tubular structure enhances the pulmonary vascular tree dividing method combined with energy function | |
| CN106618481A (en) | Traditional Chinese medicine intelligent diagnosis expert system | |
| CN106846380B (en) | A kind of medical image registration method and equipment | |
| Loureiro et al. | Using a skeleton gait energy image for pathological gait classification | |
| CN106780653A (en) | The generation method of collaterals of human and acupuncture points on the human body Visual Graph | |
| CN106169084A (en) | A kind of SVM mammary gland sorting technique based on Gauss kernel parameter selection | |
| CN112957054B (en) | 12-lead electrocardiosignal classification method based on channel attention packet residual error network | |
| CN108877923A (en) | A method of the tongue fur based on deep learning generates prescriptions of traditional Chinese medicine | |
| CN105957092B (en) | Breast molybdenum target characteristics of image self study extracting method for computer-aided 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180921 |