CN113537104B - Intelligent prescription picture identification method and system based on Internet hospital - Google Patents
Intelligent prescription picture identification method and system based on Internet hospital Download PDFInfo
- Publication number
- CN113537104B CN113537104B CN202110836008.1A CN202110836008A CN113537104B CN 113537104 B CN113537104 B CN 113537104B CN 202110836008 A CN202110836008 A CN 202110836008A CN 113537104 B CN113537104 B CN 113537104B
- Authority
- CN
- China
- Prior art keywords
- prescription
- picture
- pictures
- training
- sub
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
Abstract
The invention relates to an internet hospital-based prescription picture intelligent identification method and system, which are characterized in that a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels are collected as a training set, a Bit model is trained based on the training set to obtain a prescription picture identification model for identifying prescription pictures, the pictures to be identified sent by a client are input into the prescription picture identification model to realize the automatic identification of the prescription pictures, and when the pictures to be identified are identified as prescription pictures, the prescription pictures are pushed to a corresponding display terminal, so that the internet inquiry flow is simplified, and the inquiry efficiency is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to an intelligent identification method and system for prescription pictures based on an internet hospital.
Background
The internet hospital is a technical platform for providing services such as inquiry, prescription and the like for patients on the internet by integrating promotion and doctor resources, and on the platform, a user can make an inquiry or take a medicine according to the prescription prescribed by a doctor. However, after the user uploads the prescription picture, the doctor or pharmacist is often required to identify whether the prescription picture uploaded by the user is valid, and then perform inquiry or medicine taking according to the identification result, which is time-consuming and low in efficiency.
Disclosure of Invention
The embodiment of the application provides an intelligent prescription picture identification method and system based on an internet hospital, which can automatically identify prescription pictures uploaded by a user and automatically push the prescription pictures to a display end, so that the inquiry efficiency is improved.
In a first aspect of the embodiments of the present application, a method for intelligently identifying a prescription picture based on an internet hospital is provided, where the internet hospital includes a client, a server, and a display;
the intelligent identification method of prescription pictures based on the Internet hospital comprises the following steps:
collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
training a Bit model based on the training set to obtain a prescription picture identification model for identifying a prescription picture;
acquiring a picture to be identified sent by a client;
inputting the picture to be recognized into a prescription picture recognition model, and acquiring a prescription recognition result of the picture to be recognized;
and when the picture to be identified is identified as a prescription picture, pushing the prescription picture to the display end.
In a second aspect of the embodiments of the present application, an internet hospital-based intelligent recognition system for prescription pictures is provided, where the internet hospital includes a client, a server and a display;
the prescription picture intelligent recognition system based on the Internet hospital comprises:
the training set acquisition module is used for collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
the model acquisition module is used for training the Bit model based on the training set and acquiring a prescription picture identification model for identifying a prescription picture;
the image to be identified acquisition module is used for acquiring an image to be identified sent by the client;
the prescription identification module is used for inputting the picture to be identified into a prescription picture identification model and acquiring a prescription identification result of the picture to be identified;
and the pushing module is used for pushing the prescription picture to the display end if the picture to be identified is identified as the prescription picture.
In a third aspect of the embodiments of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps of the internet hospital-based intelligent identification method for prescription pictures.
In a fourth aspect of the embodiments of the present application, there is provided a computer device, including a memory, a processor, and a computer program stored in the memory and executable by the processor, where the processor executes the computer program to implement the steps of the internet hospital-based intelligent identification method for prescription pictures.
In the embodiment of the application, a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels are collected as a training set, a Bit model is trained based on the training set to obtain a prescription picture identification model for identifying prescription pictures, the pictures to be identified sent by a client are input into the prescription picture identification model to realize automatic identification of the prescription pictures, and when the pictures to be identified are identified as prescription pictures, the prescription pictures are pushed to a corresponding display terminal, so that the inquiry flow of the internet is simplified, and the inquiry efficiency is improved.
For a better understanding and practice, the invention is described in detail below with reference to the accompanying drawings.
Drawings
Fig. 1 is a schematic view of an application scenario of an internet hospital-based intelligent identification method for prescription pictures according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method for intelligent identification of prescription pictures for Internet-based hospitals according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an internet hospital-based intelligent identification system for prescription pictures according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
It should be understood that the embodiments described are only some embodiments of the present application, and not all embodiments. All other examples, which can be obtained by a person skilled in the art without making any inventive step based on the embodiments in the present application, belong to the scope of protection of the embodiments in the present application.
The terminology used in the embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the embodiments of the present application. As used in the examples of this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims. In the description of the present application, it is to be understood that the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not necessarily used to describe a particular order or sequence, nor are they to be construed as indicating or implying relative importance. The specific meaning of the above terms in the present application can be understood by those of ordinary skill in the art as appropriate.
In addition, in the description of the present application, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
The intelligent identification method for prescription pictures based on the internet hospital can be applied to the internet hospital shown in fig. 1, wherein the internet hospital comprises a client 101, a server 102 and a display 103. The client 101 and the display terminal 103 are respectively communicated with the server terminal 102 through a network. The client 101 is used for acquiring a picture to be identified input by a user and sending the picture to the server 102 through a network, and the server 102 is used for executing the method for identifying the picture to be identified based on the prescription picture of the internet hospital in the embodiment of the application and pushing the prescription picture to the display terminal 103 when the picture to be identified is identified as the prescription picture; the display terminal 103 is configured to receive and display the prescription picture sent by the server terminal 102.
The client 101 may be various personal computers, notebook computers, smart phones, tablet computers, portable wearable devices, or other devices that can communicate with the server 101 through a network, the server 102 may be an independent server or a server cluster composed of a plurality of servers, and the display 103 may be a device that can display prescription pictures, such as a personal computer, a notebook computer, a smart phone, a tablet computer, and a display.
As shown in fig. 2, the invention provides an intelligent identification method of prescription pictures based on an internet hospital, comprising the following steps:
step S1: collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
the picture with the prescription label can be a prescription picture which is checked and confirmed by a doctor or a pharmacist, and the prescription picture can be a prescription picture. The pictures with prescription labels can be non-prescription pictures on a network, which are taken by a user, and a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels are obtained by adding prescription labels to prescription pictures and adding non-prescription labels to non-prescription pictures.
Step S2: training a Bit model based on the training set to obtain a prescription picture identification model for identifying a prescription picture;
in one embodiment, before the step of training the Bit model based on the training set, the method further comprises:
scaling the training set into a square picture with a preset resolution and converting the square picture into a tensor matrix;
normalizing the tensor matrix.
Specifically, a tensor conversion mode such as a torchvision.transform function can be adopted to convert the picture into a tensor matrix, and the tensor matrix is normalized after conversion, so that the subsequent picture features can be compared conveniently, and the picture identification efficiency is improved.
The preset resolution can be set according to the size of the picture and the identification requirement, and preferably, the side length of the resolution is greater than 320 pixels. In one embodiment, the preset resolution may be set to 512 × 512.
The Bit model is a pre-training model (Big Transfer) developed by Google, and compared with a conventional convolution neural model, the Bit model is pre-trained by using a small amount of general data in advance, and the Bit model can be quickly adapted to a new target task by adjusting the hyper-parameters of the Bit model in a targeted manner, so that the data calculation efficiency and accuracy are improved. In the embodiment of the application, a prescription picture identification model for identifying prescription pictures is constructed based on a Bit model, and the Bit model is trained by utilizing a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels in a training set to obtain the prescription picture identification model for identifying the prescription pictures.
Step S3: acquiring a picture to be identified sent by a client;
the picture to be identified is a picture which is input by a user and needs to be identified as a prescription or not.
In one embodiment, after the step of obtaining the picture to be recognized sent by the client, the method further includes:
zooming the picture to be identified into a square picture with a preset resolution and converting the square picture into a tensor matrix;
normalizing the tensor matrix.
Images can be converted into a tensor matrix by means of tensor conversion modes such as a torchvision.
Step S4: and inputting the picture to be recognized into a prescription picture recognition model, and acquiring a prescription recognition result of the picture to be recognized.
The prescription picture identification model identifies the picture to be identified and outputs a prescription identification result; the prescription identification result comprises that the picture to be identified is identified as a prescription picture or the picture to be identified is identified as a non-prescription picture.
Step S5: and if the picture to be identified is identified as a prescription picture, pushing the prescription picture to the display end.
The display end is an object end where the prescription pictures need to be displayed, and can be a fixed display object preset by a user or determined according to an operation instruction of the user. In one embodiment, the display end is a doctor end or a pharmacist end;
the intelligent identification method of the prescription pictures based on the Internet hospital further comprises the following steps:
responding to a display instruction of a user, and pushing the prescription picture to the doctor end or the pharmacist end;
the display instruction is used for indicating an object to be pushed by the prescription picture, and the user can set the display end as a doctor end or a pharmacist end according to actual requirements, for example, when the user needs to purchase a prescription medicine, the identified prescription picture is pushed to the pharmacist end, so that the pharmacist can conveniently and timely make a corresponding medicine, and the medicine-dispensing process is simplified; when the requirement of the user is inquiry, the doctor can conveniently and quickly know the illness state of the user by pushing the identified prescription picture to the doctor end, and the inquiry efficiency is improved.
In the embodiment of the application, a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels are collected as a training set, a Bit model is trained based on the training set to obtain a prescription picture identification model for identifying prescription pictures, the pictures to be identified sent by a client are input into the prescription picture identification model to realize automatic identification of the prescription pictures, and when the pictures to be identified are identified as prescription pictures, the prescription pictures are pushed to a corresponding display end, so that the inquiry flow of the internet is simplified, and the efficiency is improved.
Because the application relates to private data such as prescription pictures, the number of collected pictures is limited, and it is difficult to improve the generalization ability of the model by collecting a large number of samples, aiming at the problems, the method for training the Bit model based on the training set comprises the following steps:
acquiring the similarity between the pictures with the non-prescription labels and prescription pictures;
the similarity between the picture with the non-prescription label and the prescription picture can be judged manually, and can also be obtained by comparing pixels, calculating a Hamming distance or utilizing a similarity recognition algorithm such as a perceptual Hash algorithm and the like.
Dividing the pictures with the non-prescription labels into a plurality of similar grades according to the corresponding relation between the similarity and the similar grades;
the corresponding relation between the similarity and the similarity level can be set according to actual requirements, wherein the higher the similarity level is, the higher the similarity between the picture and the prescription picture is.
Combining each prescription picture with similar grade with the plurality of pictures with prescription labels to obtain a plurality of sub-training sets;
acquiring the similarity grade of each sub-training set;
determining a label smoothing factor of each sub-training set according to the similarity level of each sub-training set; wherein, the higher the similarity grade is, the larger the label smoothing factor is;
and training the Bit model by utilizing the plurality of sub-training sets and the corresponding label smoothing factors based on preset training parameters to obtain a prescription picture identification model for identifying a prescription picture.
When the existing image recognition algorithm is used for recognition, the label smoothing factor is usually fixed to a certain value, the data set is divided into a plurality of sub-training sets according to the similarity, and each sub-training set uses different label smoothing factors for model training, so that the model learns the sample similarity to a certain degree, the generalization capability of the model is enhanced, and the recognition accuracy is higher.
Specifically, the preset training parameters include a training period and a learning rate; and each sub-training set completes one training in one training period, and each training period is provided with different learning rates. The step of training the Bit model by using the plurality of sub-training sets and the corresponding label smoothing factors based on the preset training parameters specifically comprises:
and according to the similarity grade of each sub-training set, sequentially utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof from low to high to train the Bit model.
Specifically, in the embodiment of the present application, the pictures with the non-prescription labels are divided into: three similarity levels "very similar to prescription" (e.g. a picture obscuring the prescription with no key information recognizable, a prescription with missing key information and blank), "somewhat similar to prescription" (e.g. an outpatient medical record somewhat similar to prescription), and "not at all similar to prescription" (e.g. a person's self-photograph, laboratory sheet) are combined with the prescription picture (positive sample) respectively to yield 3 sub-training sets of positive and negative samples train _0, train _0.1, train _ 0.3. When training is carried out by using a train set of train _0, the epoch number is set to be i, the label smoothing factor is 0, the label is smoothed to be [0,1], training is carried out by using a train set of train _0.1, the epoch number is set to be j, the label smoothing factor is set to be 0.2, the label is smoothed to be [0.1,0.9], training is carried out by using a train set of train _0.3, the epoch number is set to be q, the label smoothing factor is 0.6, and the label is smoothed to be [0.3,0.7], wherein the values of the epoch numbers i, j and q, which are three integers, can be set according to the training requirement, and preferably, the epoch numbers i, j and q are set to values which can enable the iteration number of the model in each train set to be approximately equivalent.
The method comprises the steps of obtaining the similarity grade of each picture and a prescription picture in a training set, combining the pictures with different similarity grades and the pictures with prescription labels to obtain a plurality of sub-training sets, determining the identification difficulty and the corresponding label smoothing factor of each sub-training set according to the similarity grade, and improving the generalization capability and the identification accuracy of a model by adjusting the label smoothing factor of each sub-training set; and training the Bit model by sequentially utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof according to the sequence of the similarity grade from low to high, sequentially training the model by using the sub-training sets with the identification difficulty from small to large, avoiding the model from falling into local optimum, and improving the training efficiency and precision of the prescription picture identification model.
In one embodiment, the step of training the Bit model by using the plurality of sub-training sets and the corresponding label smoothing factors thereof comprises:
and training the Bit model by utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof based on a small-batch random gradient descent algorithm with momentum and preset training parameters to obtain a prescription picture identification model for identifying a prescription picture.
The small-batch random gradient descent algorithm with momentum utilizes gradient to determine the updating direction of parameters, namely in each iteration, for each variable, the corresponding parameter value is updated according to the opposite direction of the gradient of the variable of the target function, and the target function is gradually close to the minimum value by continuously iterating and updating the parameters in the target function. The algorithm calculates the loss function of a small part of training data each time, and updates the parameters according to the average gradient of the part of data; in the embodiment of the application, the Bit model is trained based on the small-batch stochastic gradient descent algorithm with momentum, so that the variance of updated parameters is reduced, and the convergence process is more stable.
As shown in fig. 3, an embodiment of the present application further provides an intelligent prescription picture identification system based on an internet hospital, where the internet hospital includes a client, a server, and a pharmacist;
the prescription picture intelligent recognition system based on the Internet hospital comprises:
the training set acquisition module 1 is used for collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
the model acquisition module 2 is used for training a Bit model based on the training set and acquiring a prescription picture identification model for identifying a prescription picture;
the image to be identified acquisition module 3 is used for acquiring an image to be identified sent by a client;
the prescription identification module 4 is used for inputting the picture to be identified into a prescription picture identification model and acquiring a prescription identification result of the picture to be identified;
and the pushing module 5 is used for pushing the prescription picture to the display end if the picture to be identified is identified as the prescription picture.
It should be noted that, when the internet hospital-based intelligent identification system for prescription pictures is used for executing the internet hospital-based intelligent identification method for prescription pictures, the above-mentioned division of the functional modules is only used for illustration, and in practical application, the above-mentioned function distribution can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules, so as to complete all or part of the above-mentioned functions. In addition, the internet hospital-based intelligent identification system for prescription pictures and the internet hospital-based intelligent identification method for prescription pictures provided by the embodiments belong to the same concept, and specific implementation processes are detailed in the embodiments of the methods and are not repeated herein.
The embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the steps of the internet hospital-based intelligent identification method for prescription pictures.
Embodiments of the present application may take the form of a computer program product embodied on one or more storage media including, but not limited to, disk storage, CD-ROM, optical storage, and the like, in which program code is embodied. Computer readable storage media, which include both non-transitory and non-transitory, removable and non-removable media, may implement any method or technology for storage of information. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of the storage medium of the computer include, but are not limited to: phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, may be used to store information that may be accessed by a computing device.
The embodiment of the application also provides computer equipment, which comprises a memory, a processor and a computer program stored in the memory and executable by the processor, wherein the processor realizes the steps of the intelligent identification method of the prescription pictures based on the internet hospitals when executing the computer program.
According to the intelligent prescription picture identification method based on the Internet hospital, the intelligent pre-inquiry is carried out on the chief complaint information input by the user, the feedback information of the user is obtained, the description of the user is converted into the electronic medical record according with the reading and writing habits of the doctor by using the natural language processing technology, the inquiry time and the inquiry process of the doctor are reduced, and the inquiry efficiency can be effectively improved.
The present invention is not limited to the above-described embodiments, and various modifications and variations of the present invention are intended to be included within the scope of the claims and the equivalent technology of the present invention if they do not depart from the spirit and scope of the present invention.
Claims (9)
1. An intelligent identification method for prescription pictures based on an internet hospital is characterized in that the internet hospital comprises a client, a server and a display end;
the intelligent identification method of prescription pictures based on the Internet hospital comprises the following steps:
collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
training a Bit model based on the training set to obtain a prescription picture identification model for identifying prescription pictures, wherein the training step comprises the following steps: acquiring the similarity between the pictures with the non-prescription labels and prescription pictures; dividing the pictures with the non-prescription labels into a plurality of similar grades according to the corresponding relation between the similarity and the similar grades; the higher the similarity level is, the higher the similarity between the picture and the prescription picture is; combining the prescription picture of each similar grade with the pictures with the prescription labels to obtain a plurality of sub-training sets; acquiring the similarity grade of each sub-training set; determining a label smoothing factor of each sub-training set according to the similarity level of each sub-training set; wherein, the higher the similarity grade is, the larger the label smoothing factor is; training the Bit model by utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof based on preset training parameters to obtain a prescription picture identification model for identifying a prescription picture;
acquiring a picture to be identified sent by a client;
inputting the picture to be recognized into a prescription picture recognition model, and acquiring a prescription recognition result of the picture to be recognized;
and when the picture to be identified is identified as a prescription picture, pushing the prescription picture to the display end.
2. The intelligent recognition method for prescription pictures based on internet hospitals according to claim 1, characterized in that the preset training parameters comprise training period and learning rate; and each sub-training set completes one training in one training period, and each training period is provided with different learning rates.
3. The internet hospital-based intelligent identification method for prescription pictures according to claim 2, wherein the training of the Bit model by using the plurality of sub-training sets and the corresponding label smoothing factors thereof based on the preset training parameters specifically comprises:
and according to the similarity grade of each sub-training set, sequentially utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof from low to high to train the Bit model.
4. The intelligent identification method for prescription pictures based on internet hospitals according to claim 3, characterized in that the step of training the Bit model by using the plurality of sub-training sets and the corresponding label smoothing factors thereof comprises:
and training the Bit model by utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof based on a small-batch random gradient descent algorithm with momentum and preset training parameters to obtain a prescription picture identification model for identifying a prescription picture.
5. The intelligent identification method for prescription pictures based on internet hospitals according to claim 1, characterized in that after the step of obtaining the picture to be identified sent by the client, the method further comprises:
zooming the picture to be identified into a square picture with a preset resolution and converting the square picture into a tensor matrix;
normalizing the tensor matrix.
6. The intelligent recognition method for prescription pictures based on internet hospitals according to claim 1, characterized in that the display end is a doctor end or a pharmacist end;
the intelligent identification method of the prescription pictures based on the Internet hospital further comprises the following steps:
responding to a display instruction of a user, and pushing the prescription picture to the doctor end or the pharmacist end; wherein, the show instruction is used for indicating that the show end is doctor's end or pharmacist's end.
7. An internet hospital-based intelligent prescription picture identification system is characterized in that the internet hospital comprises a client, a server and a display end;
the prescription picture intelligent recognition system based on the Internet hospital comprises:
the training set acquisition module is used for collecting a plurality of pictures with non-prescription labels and a plurality of pictures with prescription labels as a training set;
a model obtaining module, configured to train a Bit model based on the training set, and obtain a prescription picture recognition model for recognizing a prescription picture, where the training step includes: acquiring the similarity between the pictures with the non-prescription labels and prescription pictures; dividing the pictures with the non-prescription labels into a plurality of similar grades according to the corresponding relation between the similarity and the similar grades; the higher the similarity level is, the higher the similarity between the picture and the prescription picture is; combining the prescription picture of each similar grade with the pictures with the prescription labels to obtain a plurality of sub-training sets; acquiring the similarity grade of each sub-training set; determining a label smoothing factor of each sub-training set according to the similarity level of each sub-training set; wherein, the higher the similarity grade is, the larger the label smoothing factor is; training the Bit model by utilizing the plurality of sub-training sets and the corresponding label smoothing factors thereof based on preset training parameters to obtain a prescription picture identification model for identifying a prescription picture;
the image to be identified acquisition module is used for acquiring an image to be identified sent by the client;
the prescription identification module is used for inputting the picture to be identified into a prescription picture identification model and acquiring a prescription identification result of the picture to be identified;
and the pushing module is used for pushing the prescription picture to the display end if the picture to be identified is identified as the prescription picture.
8. A computer device, characterized by: comprising a memory, a processor and a computer program stored in the memory and executable by the processor, the processor implementing the steps of the internet hospital-based intelligent identification method of prescription pictures according to any one of claims 1 to 6 when executing the computer program.
9. A computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program when being executed by a processor realizes the steps of the intelligent identification method of prescription pictures based on internet hospitals according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110836008.1A CN113537104B (en) | 2021-07-23 | 2021-07-23 | Intelligent prescription picture identification method and system based on Internet hospital |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110836008.1A CN113537104B (en) | 2021-07-23 | 2021-07-23 | Intelligent prescription picture identification method and system based on Internet hospital |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113537104A CN113537104A (en) | 2021-10-22 |
| CN113537104B true CN113537104B (en) | 2022-03-08 |
Family
ID=78088799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110836008.1A Active CN113537104B (en) | 2021-07-23 | 2021-07-23 | Intelligent prescription picture identification method and system based on Internet hospital |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113537104B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005211541A (en) * | 2004-01-30 | 2005-08-11 | Yuyama Manufacturing Co Ltd | Tablet storing and dispensing apparatus |
| CN112308649A (en) * | 2020-05-29 | 2021-02-02 | 北京沃东天骏信息技术有限公司 | Method and device for pushing information |
| CN112541544A (en) * | 2020-12-09 | 2021-03-23 | 福州大学 | Garbage classification method based on deep learning |
| CN112883731A (en) * | 2021-04-29 | 2021-06-01 | 腾讯科技（深圳）有限公司 | Content classification method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9916538B2 (en) * | 2012-09-15 | 2018-03-13 | Z Advanced Computing, Inc. | Method and system for feature detection |
| CN107451592A (en) * | 2017-06-30 | 2017-12-08 | 广东数相智能科技有限公司 | A kind of ethical goods checking method and device |
-
2021
- 2021-07-23 CN CN202110836008.1A patent/CN113537104B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005211541A (en) * | 2004-01-30 | 2005-08-11 | Yuyama Manufacturing Co Ltd | Tablet storing and dispensing apparatus |
| CN112308649A (en) * | 2020-05-29 | 2021-02-02 | 北京沃东天骏信息技术有限公司 | Method and device for pushing information |
| CN112541544A (en) * | 2020-12-09 | 2021-03-23 | 福州大学 | Garbage classification method based on deep learning |
| CN112883731A (en) * | 2021-04-29 | 2021-06-01 | 腾讯科技（深圳）有限公司 | Content classification method and device |
Non-Patent Citations (3)
| Title |
|---|
| "Sparse Label Smoothing Regularization";JEAN-PAUL AINAM.et al;《IEEE》;20190313;全文 * |
| "Big Transfer (BiT):General Visual Representation Learning";Alexander Kolesnikov.et al;《arXiv:1912.11370v3》;20200505;全文 * |
| "正则化方法：标签平滑";会飞的猪;《https://blog.csdn.net/qwer7512090/article/details/00700261》;20190910;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113537104A (en) | 2021-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7331171B2 (en) | Methods and apparatus for training image recognition models, methods and apparatus for recognizing images, electronic devices, storage media, and computer programs | |
| CN111191791B (en) | Picture classification method, device and equipment based on machine learning model | |
| CN108280477B (en) | Method and apparatus for clustering images | |
| CN111062871B (en) | Image processing method and device, computer equipment and readable storage medium | |
| US9087271B2 (en) | Learning semantic image similarity | |
| EP3968179A1 (en) | Place recognition method and apparatus, model training method and apparatus for place recognition, and electronic device | |
| US20150278350A1 (en) | Recommendation System With Dual Collaborative Filter Usage Matrix | |
| WO2018059016A1 (en) | Feature processing method and feature processing system for machine learning | |
| CN111324774B (en) | Video duplicate removal method and device | |
| US20190095788A1 (en) | Supervised explicit semantic analysis | |
| WO2022105117A1 (en) | Method and device for image quality assessment, computer device, and storage medium | |
| US11501110B2 (en) | Descriptor learning method for the detection and location of objects in a video | |
| CN112395390B (en) | Training corpus generation method of intention recognition model and related equipment thereof | |
| CN113821668A (en) | Data classification identification method, device, equipment and readable storage medium | |
| CN113705733A (en) | Medical bill image processing method and device, electronic device and storage medium | |
| CN114550051A (en) | Vehicle loss detection method and device, computer equipment and storage medium | |
| CN114519397A (en) | Entity link model training method, device and equipment based on comparative learning | |
| CN113537104B (en) | Intelligent prescription picture identification method and system based on Internet hospital | |
| CN116863116A (en) | Image recognition method, device, equipment and medium based on artificial intelligence | |
| CN116756281A (en) | Knowledge question-answering method, device, equipment and medium | |
| CN116662488A (en) | Service document retrieval method, device, equipment and storage medium | |
| CN114241411B (en) | Counting model processing method and device based on target detection and computer equipment | |
| CN116958852A (en) | Video and text matching method and device, electronic equipment and storage medium | |
| CN115661472A (en) | Image duplicate checking method and device, computer equipment and storage medium | |
| CN111221880B (en) | Feature combination method, device, medium, and electronic apparatus |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: Room 206-208, Building 4, No. 2 Tiantai 1st Road, Huangpu District, Guangzhou City, Guangdong Province, 510700 Patentee after: Guangzhou Fangzhou Information Technology Co.,Ltd. Address before: 510700 room 1005, No. 118, banhe Road, Huangpu District, Guangzhou, Guangdong Patentee before: Guangzhou Fangzhou Information Technology Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |