CN112070737A - Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing - Google Patents
Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing Download PDFInfo
- Publication number
- CN112070737A CN112070737A CN202010907702.3A CN202010907702A CN112070737A CN 112070737 A CN112070737 A CN 112070737A CN 202010907702 A CN202010907702 A CN 202010907702A CN 112070737 A CN112070737 A CN 112070737A
- Authority
- CN
- China
- Prior art keywords
- image
- tongue
- tongue picture
- app
- chinese medicine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000012545 processing Methods 0.000 title claims abstract description 56
- 239000003814 drug Substances 0.000 title claims abstract description 41
- 230000004069 differentiation Effects 0.000 title description 2
- 208000011580 syndromic disease Diseases 0.000 title description 2
- 230000008569 process Effects 0.000 claims abstract description 27
- 238000003745 diagnosis Methods 0.000 claims abstract description 22
- 201000010099 disease Diseases 0.000 claims abstract description 18
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 18
- 230000008859 change Effects 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 26
- 238000013527 convolutional neural network Methods 0.000 claims description 16
- 208000024891 symptom Diseases 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 230000003321 amplification Effects 0.000 claims description 3
- 238000004880 explosion Methods 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 238000013467 fragmentation Methods 0.000 claims description 3
- 238000006062 fragmentation reaction Methods 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 230000008707 rearrangement Effects 0.000 claims description 3
- 230000036541 health Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 3
- 239000011248 coating agent Substances 0.000 description 2
- 238000000576 coating method Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000037213 diet Effects 0.000 description 2
- 235000005911 diet Nutrition 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 206010020751 Hypersensitivity Diseases 0.000 description 1
- 208000026350 Inborn Genetic disease Diseases 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 208000035977 Rare disease Diseases 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 208000026935 allergic disease Diseases 0.000 description 1
- 230000007815 allergy Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000036760 body temperature Effects 0.000 description 1
- -1 body temperature Substances 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 208000016361 genetic disease Diseases 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a traditional Chinese medicine tongue observation and dialectic intelligent identification method based on tongue picture image processing, which is realized based on an intelligent terminal, wherein the intelligent terminal is provided with a preset App, and the method comprises the following steps: step S1, the user uses the intelligent terminal to shoot the tongue picture, so that the App can obtain the tongue picture; step S2, the App preprocesses the tongue picture to obtain a tongue picture meeting parameter requirements; step S3, inputting the tongue picture preprocessed in the step S2 into an image recognition buffer area of the App; and step S4, the App identifies the tongue picture according to a preset tongue picture data set, and then obtains a corresponding disease identification result and interpretation information. The invention can provide observation results of continuous image data for doctors of traditional Chinese medicine, provide data such as change conditions and subtle change records of diseases in continuous time periods, and further provide continuous data support for the disease diagnosis process.
Description
Technical Field
The invention relates to an intelligent algorithm for recognizing and distinguishing symptoms of human organ images, in particular to a traditional Chinese medicine tongue observation and distinguishing symptom intelligent recognition method based on tongue picture image processing.
Background
In the prior art, image processing and recognition technology is widely applied to the aspects of face recognition, object recognition, unmanned driving, security protection, dangerous article recognition, medical imaging and the like, and has corresponding technical application in the field of traditional Chinese medicine. The reason is not because the traditional Chinese medicine is traditional medicine, but on the one hand, the traditional Chinese medicine diagnosis depends on the accumulation and inheritance of the experiences asked by the doctors of the traditional Chinese medicine. On the other hand, the application and the conjunction point of the artificial intelligence technology are not seen in the traditional Chinese medicine field.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide an intelligent recognition method based on the combination of image recognition processing technology and traditional Chinese medicine tongue observation and diagnosis method, which can provide the doctor of traditional Chinese medicine with the observation result of continuous image data, provide the change condition of the disease in the continuous time period, and the fine change record, and further provide continuous data support for the disease diagnosis process.
In order to solve the technical problems, the invention adopts the following technical scheme.
A traditional Chinese medicine tongue observation and dialectical intelligent identification method based on tongue picture image processing is realized based on an intelligent terminal, a preset App is installed on the intelligent terminal, and the method comprises the following steps: step S1, the user uses the intelligent terminal to shoot the tongue picture, so that the App can obtain the tongue picture; step S2, the App preprocesses the tongue picture to obtain a tongue picture meeting parameter requirements; step S3, inputting the tongue picture preprocessed in the step S2 into an image recognition buffer area of the App; and step S4, the App identifies the tongue picture according to a preset tongue picture data set, and then obtains a corresponding disease identification result and interpretation information.
Preferably, the step S1 includes the following process: the user starts the App through the intelligent terminal, the App is provided with a function of shooting and acquiring the camera permission of the intelligent terminal, the camera permission comprises a front camera permission and a rear camera shooting permission, and the user switches between the front camera and the rear camera according to requirements.
Preferably, in step S2, the process of preprocessing the tongue picture by the App includes: step S20, image format cropping: comparing image parameters uploaded by a user according to standard sample parameter values by using a sample parameter comparison method, and then cutting the image uploaded by the user; step S21, image processing determination: the method comprises the steps that an image uploaded by a user is cut, then various parameter values are calculated in real time, calculation results are stored as input values of a comparison algorithm, standard sample sheet parameter values are used as comparison input values, the items are compared by utilizing a traversal algorithm, finally, comparison results are output, if the image is qualified, subsequent steps are continuously executed, if the image is unqualified, image processing is carried out, and in the image processing process, a certain item or parameters of the image are adjusted according to the results of the traversal algorithm.
Preferably, in the step S21, when performing image processing judgment, parameterization judgment is performed on five preset situations, if all parameters of the standard proof are simultaneously satisfied, the image recognition buffer area is entered, otherwise, parameter-by-parameter comparison is performed on the standard proof parameters and the user-shot image according to a preset judgment algorithm.
Preferably, in the process of comparing the standard proof parameters with the parameters of the image taken by the user, the types of parameters to be compared include: image fogging parameters, image contrast parameters, lossless magnification, whether stretching is performed, stretching parameters, image sharpness, and color parameters.
Preferably, in step S21, the five preset situations include: image de-fogging: carrying out format change on image characteristics, extracting pixel data of a non-atomized image, and carrying out reconstruction and combination to generate a non-atomized image; image contrast enhancement: carrying out brightness enhancement processing on the excessively dark image, and carrying out darkness adjustment on the excessively bright image even with light explosion until the requirement of identifying the characteristics is met; lossless amplification of the image: amplifying the image by 2-4 times of pixels according to the preset image pixel quality requirement so as to extract the image fine characteristic value; and (3) restoring the stretched image: carrying out compact rearrangement on image pixels, and recombining the image pixels by combining with a conventional contrast; image sharpness and color enhancement: and denoising the compressed blurred and low-pixel image, wherein the denoising process comprises image texture optimization, color saturation adjustment, brightness adjustment and contrast adjustment.
Preferably, the establishing process of the tongue image data set comprises: step S10, collecting tongue picture preprocessing samples, and cutting tongue picture images into corresponding specifications after uploading tongue picture shot by a user through an intelligent terminal camera or a professional camera; step S11, extracting fine features in the tongue picture features through a special layer of the convolutional neural network, and then training item by item to obtain the convolutional neural network; step S12, processing each layer of characteristics of the convolutional neural network to obtain tongue picture partition characteristic values; step S13, showing a characteristic value questionnaire for a user to perform accurate training and characteristic description in image recognition; step S14, the intelligent terminal is used as a trainer, and tongue picture training and questionnaire submission are carried out in time by utilizing fragmentation time; step S15, continuously identifying and training the same tongue picture characteristic value to ensure the image data of the tongue picture data set to be reliable and correct; step S16, counting the answer results; step S17, obtaining tongue image feature values and symptom interpretation data sets through machine training.
Preferably, in step S11, the convolutional neural network includes: a fully-connected layer, which is a hidden layer of the convolutional neural network and contains a weight vector W and an activation function; the convolution layer reserves the spatial characteristics of an input image, can obtain an image with pixel particle dimensionality through convolution operation, continuously stacks 6 different convolution results by generally using one-dimensional convolution and two-dimensional convolution, outputs one convolution Kernel and extracts a tongue picture characteristic value; and the pooling layer is used for compressing the information of the original characteristic layer.
The invention discloses a Chinese medicine tongue observation and dialectic intelligent identification method based on tongue picture processing, which is characterized in that through research and algorithm conception of image processing and identification technology and combination with a Chinese medicine tongue observation and dialectic diagnosis method, innovation is carried out, so that the application of the image processing and identification technology is fused with the Chinese medicine tongue diagnosis, through image acquisition, identification modeling and algorithm deduction of tongue organ blocks and tongue fur of a user, doctors are facilitated to deduce symptoms corresponding to the characteristics, the color and the thickness of the tongue fur and the special characteristics of serious diseases, thereby providing data such as observation results of continuous image data, changes of symptoms in continuous time periods, slight change records and the like for Chinese medicine doctors, providing continuous data support for disease diagnosis, and providing scientific, controllable and effective health management data for users in personal health management science, can provide reliable health management suggestions for users from the traditional Chinese medicine perspective.
Drawings
FIG. 1 is a flow chart of the image processing of the pre-step of the image recognition of tongue diagnosis in TCM;
FIG. 2 is a flow chart of the implementation process of the Chinese medicine tongue diagnosis image recognition trainer;
FIG. 3 is a flow chart of the process of identifying tongue inspection images and interpreting the results of symptoms;
FIG. 4 is a flow chart of the process of identifying tongue images and executing the disease diagnostor in TCM;
FIG. 5 is a block diagram of a system for implementing the method for intelligently identifying tongue observation and dialectics in the present invention.
Detailed Description
The invention is described in more detail below with reference to the figures and examples.
The invention discloses a traditional Chinese medicine tongue observation and dialectic intelligent identification method based on tongue picture image processing, which is realized based on an intelligent terminal as shown in a combined graph from 1 to 5, wherein the intelligent terminal is provided with a preset App, and the method comprises the following steps:
step S1, the user uses the intelligent terminal to shoot the tongue picture, so that the App can obtain the tongue picture;
step S2, the App preprocesses the tongue picture to obtain a tongue picture meeting parameter requirements;
step S3, inputting the tongue picture preprocessed in the step S2 into an image recognition buffer area of the App;
and step S4, the App identifies the tongue picture according to a preset tongue picture data set, and then obtains a corresponding disease identification result and interpretation information.
In the method, through the research and the algorithm conception of the image processing and recognition technology and the combination of the diagnosis method of the tongue observation and dialectical of the traditional Chinese medicine, the innovation is carried out, so that the application of the image processing and recognition technology and the tongue diagnosis of the traditional Chinese medicine are mutually fused, through image acquisition, recognition modeling and algorithm deduction of the tongue organ blocks and the tongue coating of the user, doctors are helped to deduce diseases corresponding to the characteristics of the tongue organ blocks, the color and the thickness of the tongue coating and the special characteristics of serious diseases, therefore, continuous image data observation results, changes of symptoms in a continuous time period, slight change records and other data are provided for a doctor in traditional Chinese medicine, continuous data support is provided for disease diagnosis, scientific, controllable and effective health management data are provided for a user in personal health management science, and reliable health management suggestions can be provided for the user from the perspective of traditional Chinese medicine.
In the embodiment, the algorithm for processing and identifying the image is transplanted to an App pre-installed in an intelligent terminal such as a mobile phone, and a user can obtain and upload the image in real time by taking a self-timer or other shooting modes for the tongue image by means of a front camera or a rear camera of the intelligent terminal. In the App, daily tongue picture monitoring is started firstly, the App can automatically turn on a front camera or a rear camera of the intelligent terminal, a user carries out positioning shooting on an image in a shooting outline drawn by the App, and the App is automatically uploaded to a server side of the App after shooting. To facilitate the user's knowledge and understanding of the App, the App may be named a tongue-watch App. The tongue diagnosis image processing algorithm performs image processing when image data is acquired, and the purpose of image processing is that the image can finally enable the recognition algorithm to recognize diseases and correspond to individual physical signs, so that in the image processing process, serial process judgment and processing of image processing specialization are required.
Further, the step S1 includes the following processes:
the user starts the App through the intelligent terminal, the App is provided with a function of shooting and acquiring the camera permission of the intelligent terminal, the camera permission comprises a front camera permission and a rear camera shooting permission, and the user switches between the front camera and the rear camera according to requirements. Since the captured images vary depending on the camera specifications and pixel arrangement, the App supports such variations and provides buttons to submit the captured images.
In step S2 of this embodiment, the process of the App preprocessing the tongue picture includes:
step S20, image format cropping: comparing image parameters uploaded by a user according to standard sample parameter values by using a sample parameter comparison method, and then cutting the image uploaded by the user; this step is also called image discrimination;
step S21, image processing determination: the method comprises the steps that an image uploaded by a user is cut, then various parameter values are calculated in real time, calculation results are stored as input values of a comparison algorithm, standard sample sheet parameter values are used as comparison input values, the items are compared by utilizing a traversal algorithm, finally, comparison results are output, if the image is qualified, subsequent steps are continuously executed, if the image is unqualified, image processing is carried out, and in the image processing process, a certain item or parameters of the image are adjusted according to the results of the traversal algorithm.
As a preferable mode, in the step S21, when performing image processing judgment, parameterization judgment is performed on five preset situations, if all parameters of the standard proof are met at the same time, the image enters an image recognition buffer area, otherwise, parameter-by-parameter comparison is performed on the standard proof parameters and the user shot image according to a preset judgment algorithm.
Further, in the process of comparing the standard proof parameters with the parameters of the images shot by the user, the types of parameters to be compared include: image fogging parameters, image contrast parameters, lossless magnification, whether stretching is performed, stretching parameters, image sharpness, and color parameters.
The specific algorithm principle of the process is that after the user uploads the image and cuts the image, various parameter values are calculated in real time, the calculation result is stored as an input value of the comparison algorithm, and the standard sample sheet parameter value is a comparison input value, namely an object. The number of combinations of the traversal algorithm which are aligned item by item and are overlapped and aligned for the number of times of alignment C5 is also the factorial number of 5, and the result is 120 overlapped and combined traversal alignments. And finally, outputting a result, entering the next step if the result is qualified, and performing image processing if the result is not qualified, wherein the image processing performs image parameter adjustment on a certain item or a plurality of items according to the result of the traversal algorithm.
In step S21 of this embodiment, the five preset situations include:
image de-fogging: carrying out format change on image characteristics, extracting pixel data of a non-atomized image, and carrying out reconstruction and combination to generate a non-atomized image; in particular, the defogging is lens fogging caused by climate and shooting environment temperature differences, and the image fogging is not easy to recognize and perform defogging processing. If the condition is satisfied, performing a defogged image processing step. When the image is atomized, the format of the characteristic is changed. Extracting pixel data of non-atomized images of the images, and reconstructing and combining the pixel data;
image contrast enhancement: carrying out brightness enhancement processing on the excessively dark image, and carrying out darkness adjustment on the excessively bright image even with light explosion until the requirement of identifying the characteristics is met;
lossless amplification of the image: amplifying the image by 2-4 times of pixels according to the preset image pixel quality requirement so as to extract the image fine characteristic value;
and (3) restoring the stretched image: carrying out compact rearrangement on image pixels, and recombining the image pixels by combining with a conventional contrast; in practical application, the problem of image stretching caused by shaking is inevitable in the self-photographing process, the stretched image is subjected to recovery processing, and the recovery processing process is actually a process of compactly rearranging the image pixels and is combined with a conventional contrast object for recombination;
image sharpness and color enhancement: and denoising the compressed blurred and low-pixel image, wherein the denoising process comprises image texture optimization, color saturation adjustment, brightness adjustment and contrast adjustment.
As the image of the user side may have the above five situations, or simultaneously have the superposition of some concentrated situations, for example, when the defogging operation is performed, the sharpness of the image needs to be enhanced, and the like, for which algorithm processing needs to be performed, first, the App will give a standardized format and perform parameter and quantization on the standardized image, thereby completing the judgment on whether the image needs to be processed.
After four steps of image processing, the tongue image has the basis of recognition. Before formally putting the Tongue image recognition algorithm into the App, a Tongue image acquisition and image recognition trainer based on an intelligent terminal is designed and realized, the trainer is named as a Tongue Image Set (TIS), the TIS is a universal novel human body organ sign image data set different from face recognition and medical images, and functions of Tongue image shooting, uploading, classification, questionnaire corresponding to symptoms, submission and the like are integrated into an image acquisition port of the TIS. The method is used for acquiring tongue picture image data of a patient on the premise of inquiry and consent during diagnosis for a doctor in traditional Chinese medicine, and image acquisition is carried out by taking different diagnosis times and different individuals of different diseases as dimensions. The tongue picture data of about 7000 and more than 10-12 months mobile phones are collected by taking common cases as the first, severe and rare diseases as the opportunity, and a TIS data set and a trainer data basis are established.
The image recognition training aims at adjusting the accuracy and recognition speed of the recognition algorithm, and in this embodiment, the establishing process of the tongue image data set includes:
step S10, collecting tongue picture preprocessing samples, and cutting tongue picture images into corresponding specifications after uploading tongue picture shot by a user through an intelligent terminal camera or a professional camera;
step S11, extracting fine features in the tongue picture features through a special layer of the convolutional neural network, and then training item by item to obtain the convolutional neural network;
step S12, processing each layer of characteristics of the convolutional neural network to obtain tongue picture partition characteristic values;
step S13, showing a characteristic value questionnaire for a user to perform accurate training and characteristic description in image recognition;
step S14, the intelligent terminal is used as a trainer, and tongue picture training and questionnaire submission are carried out in time by utilizing fragmentation time;
step S15, continuously identifying and training the same tongue picture characteristic value to ensure the image data of the tongue picture data set to be reliable and correct;
step S16, counting the answer results;
step S17, obtaining tongue image feature values and symptom interpretation data sets through machine training.
Further, in step S11, the convolutional neural network includes:
a fully-connected layer, which is a hidden layer of the convolutional neural network and contains a weight vector W and an activation function;
the convolution layer reserves the spatial characteristics of an input image, can obtain an image with pixel particle dimensionality through convolution operation, continuously stacks 6 different convolution results by generally using one-dimensional convolution and two-dimensional convolution, outputs one convolution Kernel and extracts a tongue picture characteristic value;
and the pooling layer is used for compressing the information of the original characteristic layer.
The tongue image trainer is used for acquiring a training data set (TIS) through diversified data sample acquisition, and the training algorithm is used for training the recognition capability based on the data set. The method is a precondition of final online real-time tongue picture identification, is used for training to obtain rapid and accurate identification capability, and provides basic preparation for identification and interpretation of corresponding symptoms of tongue picture images. On the basis of the capability, the core recognition capability can be provided, so that after the trained tongue picture image recognition and disease interpretation capability algorithm is integrated into the App, the health management capability of the App based on tongue picture diagnosis can be completely presented.
The invention discloses a traditional Chinese medicine tongue observation and dialectic intelligent recognition method based on tongue picture processing, which realizes tongue picture recognition and symptom result interpretation and has the core principle that a tongue picture is simplified into a training sample, a tongue picture characteristic value is extracted through a convolutional neural network, then a column obtained through training forms an analysis PCA model, the key characteristics of the tongue picture are subjected to dimensionality reduction operation, a Bayesian model is obtained through training, and a Bayesian adjustment factor is calculated. And the real-time tongue picture pretreatment samples are processed in the same way, and the final comparison is to calculate the posterior probability. So as to obtain the characteristic value recognition result and the corresponding explanation of the disease. The real-time tongue image recognition is strongly related to the training in the trainer, which is the function and purpose of the tongue image recognition trainer. The purpose of the tongue picture training data set is to train auxiliary and improve real-time recognition capability through large sample data.
After the tongue picture identification and symptom interpretation are finished, a diagnosis suggestion is provided, on one hand, an identification algorithm can accurately track the tongue picture change of a user, on the other hand, App is based on the targeted data acquisition of basic habits of the user, the data acquisition mainly comprises user physical signs and information, including age, height, weight, pulse, heart rate, electrocardio, blood oxygen, body temperature, medicine taking history, genetic disease history, allergy history, living areas (water quality, climate and diet) and tongue picture dialectical analysis, and on-line doctor diagnosis can clearly and accurately provide diet and medical suggestion for the user through a deep learning and recommending system. The tongue picture image recognition and dialectical analysis is a result of training through a large number of data samples, and can give suggestions to a user by using the experience results of cases according to the current symptoms of the user. The on-line Chinese medicine diagnosis aims to discriminate the accuracy of tongue picture image identification. The two are combined for use, and the algorithm capability of deep learning can be improved.
Compared with the prior art, the method is beneficial to realizing intellectualization of traditional Chinese medicine tongue observation and dialectics on the basis of tongue picture image processing and recognition, can be used as a brand-new health management and a traditional Chinese medicine preventive treatment mode combining professional traditional Chinese medicine tongue picture and image recognition technology, can universally provide convenient and professional traditional Chinese medicine health management tools for users and diagnosis suggestions supported by data, and has actual and profound values and meanings for user health management, so that the method is suitable for popularization and application in the aspect of traditional Chinese medicine diagnosis and treatment and has wide application prospects.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents or improvements made within the technical scope of the present invention should be included in the scope of the present invention.
Claims (8)
1. A traditional Chinese medicine tongue observation and dialectical intelligent identification method based on tongue picture image processing is characterized in that the method is realized based on an intelligent terminal, a preset App is installed on the intelligent terminal, and the method comprises the following steps:
step S1, the user uses the intelligent terminal to shoot the tongue picture, so that the App can obtain the tongue picture;
step S2, the App preprocesses the tongue picture to obtain a tongue picture meeting parameter requirements;
step S3, inputting the tongue picture preprocessed in the step S2 into an image recognition buffer area of the App;
and step S4, the App identifies the tongue picture according to a preset tongue picture data set, and then obtains a corresponding disease identification result and interpretation information.
2. The intelligent recognition method for traditional Chinese medicine tongue observation and dialectics based on tongue picture image processing as claimed in claim 1, wherein said step S1 includes the following procedures:
the user starts the App through the intelligent terminal, the App is provided with a function of shooting and acquiring the camera permission of the intelligent terminal, the camera permission comprises a front camera permission and a rear camera shooting permission, and the user switches between the front camera and the rear camera according to requirements.
3. The method for intelligently recognizing traditional Chinese medicine tongue observation and diagnosis based on tongue picture image processing as claimed in claim 1, wherein in said step S2, said App pre-processes said tongue picture comprising:
step S20, image format cropping: comparing image parameters uploaded by a user according to standard sample parameter values by using a sample parameter comparison method, and then cutting the image uploaded by the user;
step S21, image processing determination: the method comprises the steps that an image uploaded by a user is cut, then various parameter values are calculated in real time, calculation results are stored as input values of a comparison algorithm, standard sample sheet parameter values are used as comparison input values, the items are compared by utilizing a traversal algorithm, finally, comparison results are output, if the image is qualified, subsequent steps are continuously executed, if the image is unqualified, image processing is carried out, and in the image processing process, a certain item or parameters of the image are adjusted according to the results of the traversal algorithm.
4. The method for intelligently recognizing traditional Chinese medicine tongue observation and dialectics based on tongue image processing as claimed in claim 3, wherein in step S21, during image processing judgment, parameterization judgment is performed on five preset situations, if all parameters of the standard sample are met at the same time, the image recognition buffer area is entered, otherwise, parameter-by-parameter comparison is performed on the standard sample parameters and the user shot image according to a preset judgment algorithm.
5. The method of claim 4, wherein the comparing the standard sample sheet parameter with the user's captured image parameter comprises: image fogging parameters, image contrast parameters, lossless magnification, whether stretching is performed, stretching parameters, image sharpness, and color parameters.
6. The method for intelligently recognizing traditional Chinese medicine tongue observation and dialectics based on tongue image processing as claimed in claim 4, wherein in said step S21, five preset situations include:
image de-fogging: carrying out format change on image characteristics, extracting pixel data of a non-atomized image, and carrying out reconstruction and combination to generate a non-atomized image;
image contrast enhancement: carrying out brightness enhancement processing on the excessively dark image, and carrying out darkness adjustment on the excessively bright image even with light explosion until the requirement of identifying the characteristics is met;
lossless amplification of the image: amplifying the image by 2-4 times of pixels according to the preset image pixel quality requirement so as to extract the image fine characteristic value;
and (3) restoring the stretched image: carrying out compact rearrangement on image pixels, and recombining the image pixels by combining with a conventional contrast;
image sharpness and color enhancement: and denoising the compressed blurred and low-pixel image, wherein the denoising process comprises image texture optimization, color saturation adjustment, brightness adjustment and contrast adjustment.
7. The tongue picture image processing-based intelligent recognition method for traditional Chinese medicine tongue observation and dialectics according to claim 1, wherein the tongue picture image data set establishing process comprises:
step S10, collecting tongue picture preprocessing samples, and cutting tongue picture images into corresponding specifications after uploading tongue picture shot by a user through an intelligent terminal camera or a professional camera;
step S11, extracting fine features in the tongue picture features through a special layer of the convolutional neural network, and then training item by item to obtain the convolutional neural network;
step S12, processing each layer of characteristics of the convolutional neural network to obtain tongue picture partition characteristic values;
step S13, showing a characteristic value questionnaire for a user to perform accurate training and characteristic description in image recognition;
step S14, the intelligent terminal is used as a trainer, and tongue picture training and questionnaire submission are carried out in time by utilizing fragmentation time;
step S15, continuously identifying and training the same tongue picture characteristic value to ensure the image data of the tongue picture data set to be reliable and correct;
step S16, counting the answer results;
step S17, obtaining tongue image feature values and symptom interpretation data sets through machine training.
8. The method for intelligently recognizing traditional Chinese medicine tongue observation and dialectics based on tongue image processing as claimed in claim 7, wherein in said step S11, said convolutional neural network comprises:
a fully-connected layer, which is a hidden layer of the convolutional neural network and contains a weight vector W and an activation function;
the convolution layer reserves the spatial characteristics of an input image, can obtain an image with pixel particle dimensionality through convolution operation, continuously stacks 6 different convolution results by generally using one-dimensional convolution and two-dimensional convolution, outputs one convolution Kernel and extracts a tongue picture characteristic value;
and the pooling layer is used for compressing the information of the original characteristic layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010907702.3A CN112070737A (en) | 2020-09-02 | 2020-09-02 | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010907702.3A CN112070737A (en) | 2020-09-02 | 2020-09-02 | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112070737A true CN112070737A (en) | 2020-12-11 |
Family
ID=73666563
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010907702.3A Pending CN112070737A (en) | 2020-09-02 | 2020-09-02 | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112070737A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112971718A (en) * | 2021-02-05 | 2021-06-18 | 北京鹰之眼智能健康科技有限公司 | Syndrome identification method and device, electronic equipment and storage medium |
| CN117611581A (en) * | 2024-01-18 | 2024-02-27 | 之江实验室 | Tongue picture identification method and device based on multi-mode information and electronic equipment |
-
2020
- 2020-09-02 CN CN202010907702.3A patent/CN112070737A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112971718A (en) * | 2021-02-05 | 2021-06-18 | 北京鹰之眼智能健康科技有限公司 | Syndrome identification method and device, electronic equipment and storage medium |
| CN117611581A (en) * | 2024-01-18 | 2024-02-27 | 之江实验室 | Tongue picture identification method and device based on multi-mode information and electronic equipment |
| CN117611581B (en) * | 2024-01-18 | 2024-05-14 | 之江实验室 | Tongue picture identification method and device based on multi-mode information and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9317761B2 (en) | Method and an apparatus for determining vein patterns from a colour image | |
| CN110464366A (en) | A kind of Emotion identification method, system and storage medium | |
| CN109993068B (en) | Non-contact human emotion recognition method based on heart rate and facial features | |
| CN109948476B (en) | Human face skin detection system based on computer vision and implementation method thereof | |
| CN111931758A (en) | Face recognition method and device combining facial veins | |
| KR102162683B1 (en) | Reading aid using atypical skin disease image data | |
| CN112070737A (en) | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing | |
| CN110427881B (en) | Cross-library micro-expression recognition method and device based on face local area feature learning | |
| CN112668486A (en) | Method, device and carrier for identifying facial expressions of pre-activated residual depth separable convolutional network | |
| CN116230234A (en) | Multi-mode feature consistency psychological health abnormality identification method and system | |
| CN114842957B (en) | Senile dementia auxiliary diagnosis system and method based on emotion recognition | |
| CN109241963A (en) | Blutpunkte intelligent identification Method in capsule gastroscope image based on Adaboost machine learning | |
| CN110473176A (en) | Image processing method and device, method for processing fundus images, electronic equipment | |
| CN113128353B (en) | Emotion perception method and system oriented to natural man-machine interaction | |
| Oz et al. | Efficacy of biophysiological measurements at FTFPs for facial expression classification: A validation | |
| CN115205923A (en) | Micro-expression recognition method based on macro-expression state migration and mixed attention constraint | |
| CN110472482A (en) | A kind of method and device of object identification and real time translation | |
| CN115317304A (en) | Intelligent massage method and system based on physiological characteristic detection | |
| CN114067398A (en) | Autistic child communication disorder assisting method based on facial expression recognition | |
| CN109214286B (en) | Face recognition method based on deep neural network multi-layer feature fusion | |
| KR20080109425A (en) | Methods and system for extracting facial features and verifying sasang constitution through image recognition | |
| CN117520826B (en) | Multi-mode emotion recognition method and system based on wearable equipment | |
| CN117671774B (en) | Face emotion intelligent recognition analysis equipment | |
| KR102165487B1 (en) | Skin disease discrimination system based on skin image | |
| Salih et al. | Eye Blinking for Command Generation Based on Deep Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |