CN109101991A - A kind of infant's visual function intelligent evaluation method and device - Google Patents
A kind of infant's visual function intelligent evaluation method and device Download PDFInfo
- Publication number
- CN109101991A CN109101991A CN201810726509.2A CN201810726509A CN109101991A CN 109101991 A CN109101991 A CN 109101991A CN 201810726509 A CN201810726509 A CN 201810726509A CN 109101991 A CN109101991 A CN 109101991A
- Authority
- CN
- China
- Prior art keywords
- visual function
- intelligent evaluation
- assessment
- video data
- evaluation model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
- G06F18/2193—Validation; Performance evaluation; Active pattern learning techniques based on specific statistical tests
-
- 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
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Pathology (AREA)
- Probability & Statistics with Applications (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Databases & Information Systems (AREA)
- Biomedical Technology (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of infant's visual function intelligent evaluation method, comprising: visual function intelligent evaluation model of the building based on space-time double track frame neural network, and use the Temporal Segment Network technique drill visual function intelligent evaluation model；The video data to be evaluated for the mobile device shooting that guardian holds is received, and at least one characteristic value relevant to Assessment of visual function is extracted from video data to be evaluated by visual function intelligent evaluation model；Conversion process is iterated to each characteristic value, obtains damage of visual function probability value corresponding with each characteristic value；Using visual function intelligent evaluation model and damage of visual function probability value as foundation, Assessment of visual function is carried out to video data to be evaluated, obtains Assessment of visual function as a result, and sending Assessment of visual function result to mobile device.As it can be seen that implementing this embodiment, rapidly and accurately the visual function of the object to be detected can be assessed according to the associated video data of object to be detected, detection accuracy is high, advantageously reduces misdiagnosis rate.
Description
Technical field
The present invention relates to field of computer technology, in particular to a kind of infant's visual function intelligent evaluation method and
Device.
Background technique
Vision is the main channel that the mankind obtain information, is the vitals for differentiating shape, in whole other sensory perception works of uniting
There is important role in work, the outness of individual understanding object can be assisted, expand the scope of activities of individual, assist individual
It imitates, study, individual is stimulated to explore the motivation etc. of environment.Detection of the modern medical service for infant's visual function is general using view
Dynamic property nystagmus inspection technique, the main principle according to optokinetic nystagmus have turning for different width black and white strips for one
Drum is placed in infant and rotates at the moment, reaction of the examiner by observation infant to different width stripeds, and records and cause nystagmus
The most thin striped to quiver, finally by conversion it can be concluded that eyesight value.However found in practice, it is examined using optokinetic nystagmus
When looking into method, the rotary drum with grizzly bar that rotary drum used is fixed on a rotatable bracket is detected, it is not only simple to fall behind, but also
It needs manpower to stir the rotary drum during the inspection process, is unable to accurately control the revolving speed of rotation, therefore for infant's visual function
Detection accuracy is low, and error is big.
Summary of the invention
It in view of the above problems, being capable of basis the present invention provides a kind of infant's visual function intelligent evaluation method and device
The associated video data of object to be detected rapidly and accurately assesses the visual function of the object to be detected, and detection accuracy is high,
Advantageously reduce misdiagnosis rate.
To achieve the goals above, the present invention adopts the following technical scheme that:
First aspect present invention discloses a kind of infant's visual function intelligent evaluation method, comprising:
The visual function intelligent evaluation model based on space-time double track frame neural network is constructed, and uses Temporal
Visual function intelligent evaluation model described in Segment Network technique drill；
The video data to be evaluated for the mobile device shooting held from guardian is received, and passes through the visual function intelligence
Assessment models extract at least one characteristic value relevant to Assessment of visual function from the video data to be evaluated；
Conversion process is iterated to each characteristic value, obtains damage of visual function corresponding with each characteristic value
Probability value；
Using the visual function intelligent evaluation model and the damage of visual function probability value as foundation, to the video to be evaluated
Data carry out Assessment of visual function, obtain Assessment of visual function as a result, and sending the Assessment of visual function result to the mobile device.
As an alternative embodiment, the building is based on space-time double track frame mind in first aspect present invention
Visual function intelligent evaluation model through network, and using Temporal SegmentNetwork technique drill visual function intelligence
Assessment models, comprising:
Basic model is constructed by Temporal Segment Network technology；
The basic model is trained by training data, obtains visual function intelligent evaluation model, wherein the instruction
Practice the light stream number of fields that data include training video data, the frame difference data of the training video data and the training video
One or more of according to.
As an alternative embodiment, passing through TemporalSegment described in first aspect present invention
After Network technology constructs basic model, the method also includes:
Obtain original video data relevant to Assessment of visual function；
Data enhancing and Regularization are carried out to the original video data, obtain preprocessed video data；And
Training data is obtained from the preprocessed video data；
It is described that the basic model is trained by training data, obtain visual function intelligent evaluation model, comprising:
The basic model is trained by the training data using input pattern technology is intersected, obtains visual function
Intelligent evaluation model.
As an alternative embodiment, the Assessment of visual function result includes whether in first aspect present invention
Impaired vision rank when having the judgement result of impaired vision and having the impaired vision.
As an alternative embodiment, sending the Assessment of visual function knot described in first aspect present invention
After fruit to the mobile device, the method also includes:
When the judgement result of the impaired vision, which is, the impaired vision, sending includes mentioning for damage of visual function report
Show the communication equipment that information is held to doctor relevant to visual function, wherein the damage of visual function report includes the view function
The information of energy assessment result and the guardian.
Second aspect of the present invention discloses a kind of infant's visual function intelligent evaluation device, comprising:
Module is constructed, for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and is used
The Temporal Segment Network technique drill visual function intelligent evaluation model；
Characteristic extracting module, for receiving the video data to be evaluated for the mobile device shooting held from guardian, and
Relevant to Assessment of visual function at least one is extracted from the video data to be evaluated by the visual function intelligent evaluation model
A characteristic value；
Processing module obtains and each characteristic value pair for being iterated conversion process to each characteristic value
The damage of visual function probability value answered；
Evaluation module is used for using the visual function intelligent evaluation model and the damage of visual function probability value as foundation, right
The video data to be evaluated carries out Assessment of visual function, obtains Assessment of visual function result；
Sending module, for sending the Assessment of visual function result to the mobile device.
As an alternative embodiment, in second aspect of the present invention, the building module includes:
First submodule, for constructing basic model by Temporal Segment Network technology；
Second submodule obtains visual function intelligent evaluation for being trained by training data to the basic model
Model, wherein the training data includes training video data, the frame difference data and the instruction of the training video data
Practice one or more of the light stream field data of video.
Third aspect present invention discloses a kind of server, including memory and processor, and the memory is for storing
Computer program, the processor runs the computer program so that the computer equipment executes the infant and regards function
It can intelligent evaluation method.
Fourth aspect present invention discloses a kind of computer readable storage medium, is stored with used in above-mentioned server
Computer program.
Fifth aspect present invention discloses a kind of visual function intelligent evaluation system, including mobile device, server and communication are set
Standby, the mobile device and the communication equipment are carried out wireless communication with the server:
The mobile device for shooting video data to be evaluated, and sends the video data to be evaluated to server；
The server for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and is adopted
The visual function intelligent evaluation model described in Temporal Segment Network technique drill；And receive the view to be evaluated
Frequency evidence, and by visual function intelligent evaluation model extracted from the video data to be evaluated it is relevant to Assessment of visual function to
A few characteristic value；And conversion process is iterated to each characteristic value, it obtains corresponding with each characteristic value
Damage of visual function probability value；And using visual function intelligent evaluation model and the damage of visual function probability value as foundation, to described
Video data to be evaluated carries out Assessment of visual function, obtains Assessment of visual function as a result, and sending the Assessment of visual function result to institute
State mobile device；And work as and judge that the Assessment of visual function result is deposited when abnormal, sending to the communication equipment includes view
The prompt information of function damage report；Wherein, the damage of visual function report includes the Assessment of visual function result and the prison
Protect the information of people；
The communication equipment, the prompt information sent for receiving the server.
The infant's visual function intelligent evaluation method and device provided according to the present invention, infant's visual function intelligent evaluation
Method includes constructing the visual function intelligent evaluation model based on space-time double track frame neural network first, and use Temporal
Then the Segment Network technique drill visual function intelligent evaluation model receives the mobile device shooting that guardian holds
Video data to be evaluated, and extracted and Assessment of visual function phase from video data to be evaluated by visual function intelligent evaluation model
At least one characteristic value closed；Further, conversion process is iterated to each characteristic value, obtained corresponding with each characteristic value
Damage of visual function probability value；Using visual function intelligent evaluation model and damage of visual function probability value as foundation, to video to be evaluated
Data carry out Assessment of visual function, obtain Assessment of visual function result；Assessment of visual function result is finally sent to mobile device, it is seen then that
Implement infant's visual function intelligent evaluation method and device provided by the present invention, it can be according to the view to be evaluated of object to be detected
Frequency evidence rapidly and accurately assesses the visual function of the object to be detected, obtains the Assessment of visual function of the object to be detected
As a result, detection accuracy is high, misdiagnosis rate is advantageously reduced.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of the scope of the invention.
Fig. 1 is the flow diagram for infant's visual function intelligent evaluation method that the embodiment of the present invention one provides；
Fig. 2 is the flow diagram of infant's visual function intelligent evaluation method provided by Embodiment 2 of the present invention；
Fig. 3 is the structural schematic diagram for infant's visual function intelligent evaluation device that the embodiment of the present invention three provides；
Fig. 4 is the framework schematic diagram of visual function intelligent evaluation system provided by the embodiment of the present invention three.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
For the problems of the prior art, the present invention provides a kind of infant's visual function intelligent evaluation method and devices；
Infant's visual function intelligent evaluation method includes that visual function of the building based on space-time double track frame neural network first is intelligently commented
Estimate model, and use the Temporal Segment Network technique drill visual function intelligent evaluation model, then receives prison
The video data to be evaluated for the mobile device shooting that shield people holds, and pass through visual function intelligent evaluation model from video counts to be evaluated
At least one characteristic value relevant to Assessment of visual function is extracted according to middle；Further, each characteristic value is iterated at transformation
Reason, obtains damage of visual function probability value corresponding with each characteristic value；It is general with visual function intelligent evaluation model and damage of visual function
Rate value is foundation, carries out Assessment of visual function to video data to be evaluated, obtains Assessment of visual function result；Visual function is finally sent to comment
Result is estimated to mobile device.As it can be seen that implementing infant's visual function intelligent evaluation method and device provided by the present invention, Neng Gougen
According to the video data to be evaluated of object to be detected, rapidly and accurately the visual function of the object to be detected is assessed, is somebody's turn to do
The Assessment of visual function of object to be detected advantageously reduces misdiagnosis rate as a result, detection accuracy is high.Also, the technology can use phase
The software or hardware realization of pass, are described below by embodiment.
Embodiment 1
Referring to Fig. 1, Fig. 1 is the process signal of infant's visual function intelligent evaluation method provided in an embodiment of the present invention
Figure.Wherein, as shown in Figure 1, infant's visual function intelligent evaluation method may comprise steps of:
S101, visual function intelligent evaluation model of the building based on space-time double track frame neural network, and use Temporal
Segment Network technique drill visual function intelligent evaluation model.
In the present embodiment, the visual function intelligent evaluation model based on space-time double track frame neural network is first constructed, then again
The visual function intelligent evaluation model is trained using Temporal Segment Network technology, in trained process
In, sparse time sampling technology and supervision strategy are combined to improve the efficiency of model training study.
In the present embodiment, above-mentioned infant is the infant within three one full year of life.
In the present embodiment, Temporal Segment Network is to the new network after the improving of binary-flow network
Model is used for Activity recognition.
In the present embodiment, space-time double track frame neural network is a kind of artificial intelligence nerve based on both time and space
Network.Wherein, space-time double track frame neural network can be time-space domain depth convolutional neural networks.
S102, the video data to be evaluated for receiving the mobile device shooting held from guardian, and pass through visual function intelligence
Energy assessment models extract at least one characteristic value relevant to Assessment of visual function from video data to be evaluated.
In the present embodiment, after receiving video data to be evaluated, ash can also be carried out to the video data to be evaluated
Spend the image preprocessings such as correction, noise filtering.Wherein, carrying out pretreated process to the video data to be evaluated includes that light is mended
It repays, greyscale transformation, histogram equalization, normalization, geometric correction, filtering and sharpening etc., the embodiment of the present invention is not construed as limiting.
In present example, which can be the positive behavior video data of infant, then accordingly
At least one characteristic value extracted from video data to be evaluated by visual function intelligent evaluation model is also behavioural characteristic.Then regard
Function intelligent evaluation model can extract from video data to be evaluated related to Assessment of visual function from the dimension of the time and space
At least one characteristic value, be conducive to promoted Evaluation accuracy.
S103, conversion process is iterated to each characteristic value, it is general obtains damage of visual function corresponding with each characteristic value
Rate value.
S104, using visual function intelligent evaluation model and damage of visual function probability value as foundation, to video data to be evaluated into
Row Assessment of visual function obtains Assessment of visual function as a result, and sending Assessment of visual function result to mobile device.
In the present embodiment, damage of visual function probability value is input to the Softmax classifier of visual function intelligent evaluation model
In, Assessment of visual function result can be obtained.
In infant's visual function intelligent evaluation method described in Fig. 1, building is based on space-time double track frame nerve first
The visual function intelligent evaluation model of network, and intelligently commented using the Temporal Segment Network technique drill visual function
Estimate model, then receive the video data to be evaluated for the mobile device shooting that guardian holds, and passes through visual function intelligent evaluation
Model extracts at least one characteristic value relevant to Assessment of visual function from video data to be evaluated；Further, to each spy
Value indicative is iterated conversion process, obtains damage of visual function probability value corresponding with each characteristic value；With visual function intelligent evaluation
Model and damage of visual function probability value are foundation, carry out Assessment of visual function to video data to be evaluated, obtain Assessment of visual function knot
Fruit；Assessment of visual function result is finally sent to mobile device.As it can be seen that implement method described in Fig. 1, it can be according to be detected right
The video data to be evaluated of elephant, rapidly and accurately assesses the visual function of the object to be detected, obtains the object to be detected
Assessment of visual function as a result, detection accuracy is high, advantageously reduce misdiagnosis rate.
Embodiment 2
Referring to Fig. 2, Fig. 2 is the process signal of infant's visual function intelligent evaluation method provided in an embodiment of the present invention
Figure.Wherein, as shown in Fig. 2, infant's visual function intelligent evaluation method may comprise steps of:
S201, basic model is constructed by Temporal Segment Network technology.
S202, acquisition original video data relevant to Assessment of visual function.
S203, data enhancing and Regularization are carried out to original video data, obtains preprocessed video data.
In the present embodiment, to original video data carry out data enhancing processing include it is horizontal to original video data progress/
Vertical overturning processing, the random cutting process of crop, color change processing, affine/rotation transformation processing, the processing of random size conversion
And one or more of plus noise processing, this embodiment is not limited.
In the present embodiment, by carrying out data enhancing processing to original video data, be conducive to the number of training for promotion data
According to amount, and then being promoted is capability evaluation precision.
Visual function intelligence in the present embodiment, by carrying out Regularization to original video data, after training can be made
There is the problems such as over-fitting when carrying out Assessment of visual function in assessment models, and then appearance situations such as avoid mistaken diagnosis.
In the present embodiment, Regularization is carried out to original video data, does not change the essence of original video data.
S204, training data is obtained from preprocessed video data.
S205, basic model is trained by training data using intersection input pattern technology, obtains visual function intelligence
It can assessment models.
In the present embodiment, during being trained to basic model, the training data of input include training video data,
One or more of frame difference data and the light stream field data of training video of training video data, the present embodiment is not made
It limits.And then improve basic model when being trained study according to training data, moreover it is possible to study into video data frame with
Variation between frame so as to effectively extract characteristic value relevant to Assessment of visual function, and then improves Evaluation accuracy.
S206, the video data to be evaluated for receiving the mobile device shooting held from guardian, and pass through visual function intelligence
Energy assessment models extract at least one characteristic value relevant to Assessment of visual function from video data to be evaluated.
S207, conversion process is iterated to each characteristic value, it is general obtains damage of visual function corresponding with each characteristic value
Rate value.
S208, using visual function intelligent evaluation model and damage of visual function probability value as foundation, to video data to be evaluated into
Row Assessment of visual function obtains Assessment of visual function as a result, and sending Assessment of visual function result to mobile device.
In the present embodiment, when Assessment of visual function result includes whether the judgement result of impaired vision and has impaired vision
Impaired vision rank.
S209, when the judgement result of impaired vision, which is, impaired vision, that holds to doctor relevant to visual function is logical
Believing that equipment is sent includes the prompt information of damage of visual function report, wherein damage of visual function report includes Assessment of visual function result
With the information of guardian.
As it can be seen that implement infant's visual function intelligent evaluation method described in Fig. 2, it can be according to the to be evaluated of object to be detected
Video data is surveyed, rapidly and accurately the visual function of the object to be detected is assessed, obtains the visual function of the object to be detected
Assessment result, detection accuracy is high, advantageously reduces misdiagnosis rate.
Embodiment 3
Referring to Fig. 3, Fig. 3 is the structural representation of infant's visual function intelligent evaluation device provided in an embodiment of the present invention
Figure.As shown in figure 3, infant's visual function intelligent evaluation device includes:
Module 301 is constructed, for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and is adopted
With the Temporal Segment Network technique drill visual function intelligent evaluation model.
Characteristic extracting module 302 for receiving the video data to be evaluated for the mobile device shooting that guardian holds, and is led to
It crosses visual function intelligent evaluation model and extracts at least one characteristic value relevant to Assessment of visual function from video data to be evaluated.
Processing module 303 obtains view corresponding with each characteristic value for being iterated conversion process to each characteristic value
Function damage probability value.
Evaluation module 304 is used for using visual function intelligent evaluation model and damage of visual function probability value as foundation, to be evaluated
Video data carries out Assessment of visual function, obtains Assessment of visual function result.
Sending module 305, for sending Assessment of visual function result to mobile device.
As an alternative embodiment, building module 301 includes:
First submodule 3011, for constructing basic model by Temporal Segment Network technology.
Second submodule 3012 obtains visual function intelligent evaluation for being trained by training data to basic model
Model, training data include the light stream number of fields of training video data, the frame difference data of training video data and training video
One or more of according to.
As it can be seen that implement infant's visual function intelligent evaluation device described in the present embodiment, it can be according to object to be detected
Video data to be evaluated, rapidly and accurately the visual function of the object to be detected is assessed, obtains the object to be detected
Assessment of visual function advantageously reduces misdiagnosis rate as a result, detection accuracy is high.
Referring to Fig. 4, Fig. 4 is a kind of framework signal of visual function intelligent evaluation system provided by the embodiment of the present invention
Figure.As shown in figure 4, the visual function intelligent evaluation system includes:
Mobile device 401 for shooting video data to be evaluated, and sends video data to be evaluated to server 402.
Server 402 for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and uses
Visual function intelligent evaluation model described in Temporal Segment Network technique drill；And receive video counts to be evaluated
According to, and at least one spy relevant to Assessment of visual function is extracted from video data to be evaluated by visual function intelligent evaluation model
Value indicative；And conversion process is iterated to each characteristic value, obtain damage of visual function probability value corresponding with each characteristic value；
And using visual function intelligent evaluation model and damage of visual function probability value as foundation, visual function is carried out to video data to be evaluated and is commented
Estimate, obtains Assessment of visual function as a result, and sending Assessment of visual function result to mobile device 401；And works as and judge that visual function is commented
Estimate result to deposit when abnormal, the prompt information reported including damage of visual function is sent to communication equipment 403；Wherein, visual function damages
Wound report includes the information of Assessment of visual function result and guardian.
Communication equipment 403, for receiving the prompt information of the transmission of server 402.
In the present embodiment, mobile device 401 is the equipment that guardian 405 holds, and communication equipment 403 is related to visual function
The equipment held of doctor 404.
In the present embodiment, mobile device 401, communication equipment 403 are referred to as user equipment (UE, User
Equipment), mobile station, access terminal, subscriber unit, subscriber station, movement station, remote station, remote terminal, mobile device, end
End, wireless telecom equipment, user agent or user apparatus etc. specifically can be website (ST, Station) in WLAN, bee
Cellular telephone, wireless phone, session initiation protocol (SIP, Session Initiation Protocol) phone, Wireless Local Loop
It stands, personal digital assistant (PDA, Personal Digital Assistant), have in road (WLL, Wireless Local Loop)
There is the handheld device of wireless communication function, the other processing equipments for calculating equipment, being connected to radio modem, vehicle-mounted set
Appointing in terminal device in the PLMN network of standby, wearable device, the mobile station in future 5G network and the following evolution etc.
It anticipates one kind, the embodiment of the present invention is not construed as limiting.
In the present embodiment, server 402 can both pass through fortune between mobile device 401, communication equipment 403 respectively
It seeks quotient base station and carries out mobile communication (LTE), and can be communicated by physical communication lines, the embodiment of the present invention does not limit
It is fixed.
In addition, the present invention also provides a kind of server, including memory and processor, memory are calculated for storing
Machine program, processor runs computer program so that computer equipment executes above-mentioned partly or completely infant visual function intelligence
It can appraisal procedure.
In the present embodiment, a kind of computer readable storage medium is also provided, is stored with used in above-mentioned server
Computer program.
In several embodiments provided herein, it should be understood that disclosed device and method can also pass through
Other modes are realized.The apparatus embodiments described above are merely exemplary, for example, flow chart and structure in attached drawing
Figure shows the model support in the cards of the device of multiple embodiments according to the present invention, method and computer program product
Structure, function and operation.In this regard, each box in flowchart or block diagram can represent a module, section or code
A part, a part of module, section or code includes one or more for implementing the specified logical function holds
Row instruction.It should also be noted that function marked in the box can also be to be different from attached drawing in the implementation as replacement
Middle marked sequence occurs.For example, two continuous boxes can actually be basically executed in parallel, they sometimes can also be with
It executes in the opposite order, this depends on the function involved.It is also noted that each of structure chart and/or flow chart
The combination of box and the box in structure chart and/or flow chart can use the dedicated base for executing defined function or movement
It realizes, or can realize using a combination of dedicated hardware and computer instructions in the system of hardware.
In addition, each functional module or unit in each embodiment of the present invention can integrate one independence of formation together
Part, be also possible to modules individualism, an independent part can also be integrated to form with two or more modules.
If function is realized and when sold or used as an independent product in the form of software function module, can store
In a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially in other words to existing
Having the part for the part or the technical solution that technology contributes can be embodied in the form of software products, the computer
Software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be intelligent hand
Machine, personal computer, server or network equipment etc.) execute all or part of the steps of each embodiment method of the present invention.
And storage medium above-mentioned includes: that USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
The various media that can store program code such as reservoir (RAM, Random Access Memory), magnetic or disk.
More than, only a specific embodiment of the invention, but scope of protection of the present invention is not limited thereto, and it is any to be familiar with
Those skilled in the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all cover
Within protection scope of the present invention.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. a kind of infant's visual function intelligent evaluation method characterized by comprising
The visual function intelligent evaluation model based on space-time double track frame neural network is constructed, and uses Temporal Segment
Visual function intelligent evaluation model described in Network technique drill；
The video data to be evaluated for the mobile device shooting held from guardian is received, and passes through the visual function intelligent evaluation
Model extracts at least one characteristic value relevant to Assessment of visual function from the video data to be evaluated；
Conversion process is iterated to each characteristic value, obtains damage of visual function probability corresponding with each characteristic value
Value；
Using the visual function intelligent evaluation model and the damage of visual function probability value as foundation, to the video data to be evaluated
Assessment of visual function is carried out, obtains Assessment of visual function as a result, and sending the Assessment of visual function result to the mobile device.
2. infant's visual function intelligent evaluation method according to claim 1, which is characterized in that the building is based on space-time
The visual function intelligent evaluation model of double track frame neural network, and should using Temporal Segment Network technique drill
Visual function intelligent evaluation model, comprising:
Basic model is constructed by Temporal Segment Network technology；
The basic model is trained by training data, obtains visual function intelligent evaluation model, wherein the trained number
According in the light stream field data including training video data, the frame difference data of the training video data and the training video
One or more.
3. infant's visual function intelligent evaluation method according to claim 2, which is characterized in that pass through described
After Temporal Segment Network technology constructs basic model, the method also includes:
Obtain original video data relevant to Assessment of visual function；
Data enhancing and Regularization are carried out to the original video data, obtain preprocessed video data；And
Training data is obtained from the preprocessed video data；
It is described that the basic model is trained by training data, obtain visual function intelligent evaluation model, comprising:
The basic model is trained by the training data using input pattern technology is intersected, obtains visual function intelligence
Assessment models.
4. infant's visual function intelligent evaluation method according to claim 1, which is characterized in that the Assessment of visual function knot
Impaired vision rank when fruit includes whether the judgement result of impaired vision and has the impaired vision.
5. infant's visual function intelligent evaluation method according to claim 4, which is characterized in that send the view described
After functional assessment result to the mobile device, the method also includes:
When the judgement result of the impaired vision, which is, the impaired vision, the prompt letter including damage of visual function report is sent
The communication equipment that breath is held to doctor relevant to visual function, wherein the damage of visual function report includes that the visual function is commented
Estimate the information of result and the guardian.
6. a kind of infant's visual function intelligent evaluation device characterized by comprising
Module is constructed, for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and is used
The Temporal Segment Network technique drill visual function intelligent evaluation model；
Characteristic extracting module for receiving the video data to be evaluated for the mobile device shooting held from guardian, and passes through
The visual function intelligent evaluation model extracts at least one spy relevant to Assessment of visual function from the video data to be evaluated
Value indicative；
Processing module obtains corresponding with each characteristic value for being iterated conversion process to each characteristic value
Damage of visual function probability value；
Evaluation module is used for using the visual function intelligent evaluation model and the damage of visual function probability value as foundation, to described
Video data to be evaluated carries out Assessment of visual function, obtains Assessment of visual function result；
Sending module, for sending the Assessment of visual function result to the mobile device.
7. infant's visual function intelligent evaluation device according to claim 6, which is characterized in that the building module packet
It includes:
First submodule, for constructing basic model by Temporal Segment Network technology；
Second submodule obtains visual function intelligent evaluation model for being trained by training data to the basic model,
Wherein, the training data includes training video data, the frame difference data of the training video data and the training view
One or more of light stream field data of frequency.
8. a kind of server, which is characterized in that including memory and processor, the memory is for storing computer journey
Sequence, the processor runs the computer program so that the computer equipment executes according to claim 1 to any one of 5
Infant's visual function intelligent evaluation method.
9. a kind of computer readable storage medium, which is characterized in that it is stored with used in server according to any one of claims 8
Computer program.
10. a kind of visual function intelligent evaluation system, which is characterized in that including mobile device, server and communication equipment, the shifting
Dynamic equipment and the communication equipment are carried out wireless communication with the server:
The mobile device for shooting video data to be evaluated, and sends the video data to be evaluated to server；
The server for constructing the visual function intelligent evaluation model based on space-time double track frame neural network, and uses
Visual function intelligent evaluation model described in Temporal Segment Network technique drill；And receive the video to be evaluated
Data, and by visual function intelligent evaluation model extracted from the video data to be evaluated it is relevant to Assessment of visual function at least
One characteristic value；And conversion process is iterated to each characteristic value, obtain view corresponding with each characteristic value
Function damage probability value；And using visual function intelligent evaluation model and the damage of visual function probability value as foundation, to it is described to
It evaluates and tests video data and carries out Assessment of visual function, obtain Assessment of visual function as a result, and sending the Assessment of visual function result to described
Mobile device；And work as and judge that the Assessment of visual function result is deposited when abnormal, sending to the communication equipment includes view function
The prompt information of report can be damaged；Wherein, the damage of visual function report includes the Assessment of visual function result and the monitoring
The information of people；
The communication equipment, the prompt information sent for receiving the server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810726509.2A CN109101991A (en) | 2018-07-04 | 2018-07-04 | A kind of infant's visual function intelligent evaluation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810726509.2A CN109101991A (en) | 2018-07-04 | 2018-07-04 | A kind of infant's visual function intelligent evaluation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109101991A true CN109101991A (en) | 2018-12-28 |
Family
ID=64845731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810726509.2A Pending CN109101991A (en) | 2018-07-04 | 2018-07-04 | A kind of infant's visual function intelligent evaluation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109101991A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113034415A (en) * | 2021-03-23 | 2021-06-25 | 哈尔滨市科佳通用机电股份有限公司 | Method for amplifying small parts of railway locomotive image |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2312931A1 (en) * | 1997-12-04 | 1999-06-10 | Virtual-Eye.Com, Inc. | Automated visual function testing via telemedicine |
| CA2338417A1 (en) * | 1999-05-27 | 2000-12-07 | Virtual-Eye.Com, Inc. | Visual function testing with virtual retinal display |
| CN107358036A (en) * | 2017-06-30 | 2017-11-17 | 北京机器之声科技有限公司 | A kind of child myopia Risk Forecast Method, apparatus and system |
-
2018
- 2018-07-04 CN CN201810726509.2A patent/CN109101991A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2312931A1 (en) * | 1997-12-04 | 1999-06-10 | Virtual-Eye.Com, Inc. | Automated visual function testing via telemedicine |
| CA2338417A1 (en) * | 1999-05-27 | 2000-12-07 | Virtual-Eye.Com, Inc. | Visual function testing with virtual retinal display |
| CN107358036A (en) * | 2017-06-30 | 2017-11-17 | 北京机器之声科技有限公司 | A kind of child myopia Risk Forecast Method, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| WANG, LM 等: "Temporal Segment Networks: Towards Good Practices for Deep Action Recognition", 《COMPUTER VISION - ECCV 2016, PT VIII》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113034415A (en) * | 2021-03-23 | 2021-06-25 | 哈尔滨市科佳通用机电股份有限公司 | Method for amplifying small parts of railway locomotive image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110473147A (en) | A kind of video deblurring method and device | |
| CN101635833B (en) | Method, device and system for video monitoring | |
| CN111523347A (en) | Image detection method and device, computer equipment and storage medium | |
| CN110718235A (en) | Abnormal sound detection method, electronic device and storage medium | |
| Liu et al. | Contrastive self-supervised representation learning for sensing signals from the time-frequency perspective | |
| CN108022171B (en) | Data processing method and equipment | |
| CN112444858B (en) | Edge cloud cooperation-based earthquake real-time monitoring AI (Artificial intelligence) computing method, device and medium | |
| CN106875005A (en) | Adaptive threshold neuronal messages processing method and system | |
| CN111242096B (en) | People number gradient-based people group distinguishing method | |
| Gu et al. | Learning a unified blind image quality metric via on-line and off-line big training instances | |
| CN107426442A (en) | The information gathering of mobile phone full system type pictorial and recognition of face analyses and comparison system and application method | |
| CN114222202B (en) | Environment self-adaptive activity detection method and system based on WiFi CSI | |
| CN109101991A (en) | A kind of infant's visual function intelligent evaluation method and device | |
| CN109840496A (en) | Aquaculture area hierarchical classification extracting method, device, storage medium and electronic equipment | |
| CN106230653A (en) | A kind of data integration method with MAC code as major key | |
| CN109559336A (en) | Object tracking method, device and storage medium | |
| CN109858332A (en) | A kind of human behavior analysis method, device and electronic equipment | |
| CN110222647B (en) | Face in-vivo detection method based on convolutional neural network | |
| CN108460765A (en) | A kind of iris image quality detection method | |
| CN113887519A (en) | Artificial intelligence-based garbage throwing identification method, device, medium and server | |
| CN113989931A (en) | Portable abnormal behavior intelligent analysis system | |
| CN103888704B (en) | A kind of interference signal code method and apparatus | |
| CN113111823A (en) | Abnormal behavior detection method and related device for building construction site | |
| CN111369494A (en) | Winter wheat ear density detection method and device | |
| CN103916617B (en) | A kind of interference signal generation method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181228 |
|
| RJ01 | Rejection of invention patent application after publication |