A kind of method for avoiding damaging puerpera's eye
Technical field
The present invention relates to health care of body field, more particularly to a kind of method for avoiding damaging puerpera's eye.
Background technology
Electronic reader is used exclusively for showing the number of books, magazine, newspaper and the written material in other printed matter sources
The portable of letter sheet, low energy consumption, high-resolution equipment.Some electronic readers also provide similar blog, website, and news pushes away
The access of these electronic documents such as send.In most cases, the equipment of other major functions is possessed, such as net book and mobile phone,
Also used as electronic reader.Herein, " electronic reader " this term refers to setting specifically for display text
The equipment of meter.Most electronic reader provides the text of black and white adjusting using the technology for being referred to as " electric ink "
This display mode rather than LCD screen display mode, they allow user to store stack room in single equipment, annotated,
Emphasis is drawn on text and bookmark is set.
Current electronic reader, although reducing the influence to the eyes of spectators using the technology of " electric ink ", when
If viewing time is long, the eyes of spectators can still be caused with certain injury.The current eye lacked to electronic reader spectators
Eyeball level of fatigue carries out the technical scheme of effective detection.
For just having given birth to the puerpera of child, earlier month is in lactation and convalescence, and often bed rest time is more.
In order to keep learning new educating knowledge and dismiss the boring time, electronic reader can be all used substantially, and usage time is also
It is longer.Therefore, new technology is also required at present to solve or alleviate the problem.
The content of the invention
In order to solve the above problems, the invention provides a kind of method for avoiding damaging puerpera's eye, face is used
Portion's image capture device, it is embedded in the front framework of electronic reader, the face for the spectators to electronic reader is carried out
View data captures, and to obtain and export high definition face image, and adaptive depth nerve net is carried out to high definition face image
The detection and identification of network, spectators' eye fatigue grade corresponding to confirmation, while accordingly alarmed using voice or image mode.
According to an aspect of the present invention, there is provided a kind of method for avoiding damaging puerpera's eye, methods described bag
Include using electronic reader spectators' fatigue strength analyzing platform with tired to electronic reader spectators based on adaptive deep neural network
Lao Du parse and alarm immediately, and the electronic reader spectators fatigue strength analyzing platform includes:Face image capture device,
It is embedded in the front framework of electronic reader, the face for the spectators to electronic reader carries out view data capture, with
Obtain and export high definition face image.
Preferably, the platform also includes:Electric ink touch screen, it is fixedly mounted on the front framework of electronic reader
It is interior, the related content of the e-book for showing the electronic reader storage, it is additionally operable to receiving the eye strain of spectators
During grade, real-time display word content corresponding with the eye strain grade of spectators.
Preferably, the platform also includes:Boombox, it is embedded in the front framework of electronic reader, is used for
When the eye strain grade of the spectators received transfinites, corresponding fatigue warning voice document is played.
Preferably, the platform also includes:
Noise amplitude analytical equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is set with face image capture
Standby connection, for receiving high definition face image, is analyzed the noise in the high definition face image to obtain various noises
Amplitude, the amplitude based on various noises determines the credit rating of the high definition face image;
Amount of images decision device, is arranged on the surface-mounted integrated circuit of electronic reader, is set with noise amplitude analysis
Standby connection, for when the credit rating is less than or equal to pre-determined lower limit credit rating, selecting and credit rating respective amount
Training image as default training quantity, credit rating is higher, and the quantity of training image is more, and is additionally operable to receiving
Credit rating when being more than pre-determined lower limit credit rating, select the training image of fixed qty as default training quantity;
Color image obtains equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is set with described image quantitative decision
Standby connection, for each type scene, the image of default training quantity is chosen as training image, by the instruction of all types scene
Practice image and be all transformed into YUV color spaces to obtain multiple training color images;
Normalized equipment, is arranged on the surface-mounted integrated circuit of electronic reader, and obtaining equipment with color image is connected,
For receiving the multiple training color image, normalized is performed respectively to be consolidated to the multiple training color image
The multiple standard exercise images being sized;
Feature decision device, is arranged on the surface-mounted integrated circuit of electronic reader, respectively with the normalized equipment
Connected with the noise amplitude analytical equipment, the input quantity type of the model of selection is determined according to credit rating, according to selection
Input quantity type carries out feature extraction to each standard exercise image and meets input quantity type, the standard of selection to obtain
Training characteristics amount corresponding to training image, wherein, credit rating is lower, corresponding to the input quantity type of the model of selection at data
Reason amount is more;
Model training equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is connected with the feature decision device, uses
In receiving each training characteristics amount corresponding to each standard exercise image, by each training characteristics amount be respectively outputted in model with
The training of model parameter is completed, wherein, model includes input layer, hidden layer and output layer, and the output quantity of the output layer of model is
Eye strain grade;
Fatigue strength extraction equipment, is arranged on the surface-mounted integrated circuit of electronic reader, respectively with the feature decision device
Connected with the noise amplitude analytical equipment, for receiving high definition face image, to high definition face image successively YUV color spaces
Conversion, normalized and meet the input quantity type of selection according to the feature extraction of the input quantity type of selection to obtain
, identification feature amount corresponding to the high definition face image, using identification feature amount corresponding to the high definition face image as training after
The input of the input layer of model, to obtain the eye strain grade of spectators.
Preferably, the electric ink touch screen is connected with the fatigue strength extraction equipment, for receiving the eye of spectators
Level of fatigue.
Preferably, the boombox is connected with the fatigue strength extraction equipment, and the eye for receiving spectators is tired
Labor grade.
Preferably, the platform also includes:FLASH storage chips, it is connected with described image quantitative decision equipment, for pre-
Pre-determined lower limit credit rating is first stored, is additionally operable to store the default training quantity of described image quantitative decision equipment output.
The present invention at least possesses the important inventive point in following a few places:
(1) total quality of input picture is assessed using noise amplitude analytical equipment, selected when second-rate
The image recognition apparatus of high intensity, when quality is good, selects the image recognition of regular tenacity to set to improve image recognition quality
It is standby to reduce unnecessary equipment power dissipation, so as to take into account the efficiency of image recognition and effect;
(2) the fatigue strength grade testing mechanism based on adaptive deep neural network has been built, has improved spectators' fatigue strength
The precision of testing result.
Brief description of the drawings
Embodiment of the present invention is described below with reference to accompanying drawing, wherein:
Fig. 1 is the fatigue strength extraction of electronic reader spectators' fatigue strength analyzing platform according to embodiment of the present invention
The interface diagram of equipment.
Fig. 2 is the structure square frame of electronic reader spectators' fatigue strength analyzing platform according to embodiment of the present invention
Figure.
Embodiment
The embodiment of the method for avoiding damaging puerpera's eye of the present invention is carried out in detail below with reference to accompanying drawings
Describe in detail bright.
It is not related to the technical scheme with fatigue strength detection in the prior art.In order to overcome above-mentioned deficiency, the present invention
A kind of method for avoiding damaging puerpera's eye is built, methods described is including the use of electronic reader spectators' fatigue strength solution
Platform is analysed to carry out parsing and alarming immediately to electronic reader spectators fatigue strength based on adaptive deep neural network.The electricity
The specific embodiment of sub- reader spectators fatigue strength analyzing platform is as follows.
Fig. 1 is the fatigue strength extraction of electronic reader spectators' fatigue strength analyzing platform according to embodiment of the present invention
The interface diagram of equipment.Wherein, the fatigue strength extraction equipment is realized using microcontroller as shown in Figure 1.
Fig. 2 is the structure square frame of electronic reader spectators' fatigue strength analyzing platform according to embodiment of the present invention
Figure, the platform include:
Face image capture device, it is embedded in the front framework of electronic reader, for the spectators to electronic reader
Face carry out view data capture, to obtain and export high definition face image.
Then, continue to carry out further the concrete structure of electronic reader spectators' fatigue strength analyzing platform of the present invention
Explanation.
It can also include in the electronic reader spectators fatigue strength analyzing platform:
Electric ink touch screen, it is fixedly mounted in the front framework of electronic reader, for showing the electronic reading
The related content of the e-book of device storage, is additionally operable to when receiving the eye strain grade of spectators, real-time display and spectators'
Word content corresponding to eye strain grade.
It can also include in the electronic reader spectators fatigue strength analyzing platform:
Boombox, it is embedded in the front framework of electronic reader, it is tired for the eye in the spectators received
When labor grade transfinites, corresponding fatigue warning voice document is played.
It can also include in the electronic reader spectators fatigue strength analyzing platform:
Noise amplitude analytical equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is set with face image capture
Standby connection, for receiving high definition face image, is analyzed the noise in the high definition face image to obtain various noises
Amplitude, the amplitude based on various noises determines the credit rating of the high definition face image;
Amount of images decision device, is arranged on the surface-mounted integrated circuit of electronic reader, is set with noise amplitude analysis
Standby connection, for when the credit rating is less than or equal to pre-determined lower limit credit rating, selecting and credit rating respective amount
Training image as default training quantity, credit rating is higher, and the quantity of training image is more, and is additionally operable to receiving
Credit rating when being more than pre-determined lower limit credit rating, select the training image of fixed qty as default training quantity;
Color image obtains equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is set with described image quantitative decision
Standby connection, for each type scene, the image of default training quantity is chosen as training image, by the instruction of all types scene
Practice image and be all transformed into YUV color spaces to obtain multiple training color images;
Normalized equipment, is arranged on the surface-mounted integrated circuit of electronic reader, and obtaining equipment with color image is connected,
For receiving the multiple training color image, normalized is performed respectively to be consolidated to the multiple training color image
The multiple standard exercise images being sized;
Feature decision device, is arranged on the surface-mounted integrated circuit of electronic reader, respectively with the normalized equipment
Connected with the noise amplitude analytical equipment, the input quantity type of the model of selection is determined according to credit rating, according to selection
Input quantity type carries out feature extraction to each standard exercise image and meets input quantity type, the standard of selection to obtain
Training characteristics amount corresponding to training image, wherein, credit rating is lower, corresponding to the input quantity type of the model of selection at data
Reason amount is more;
Model training equipment, is arranged on the surface-mounted integrated circuit of electronic reader, is connected with the feature decision device, uses
In receiving each training characteristics amount corresponding to each standard exercise image, by each training characteristics amount be respectively outputted in model with
The training of model parameter is completed, wherein, model includes input layer, hidden layer and output layer, and the output quantity of the output layer of model is
Eye strain grade;
Fatigue strength extraction equipment, is arranged on the surface-mounted integrated circuit of electronic reader, respectively with the feature decision device
Connected with the noise amplitude analytical equipment, for receiving high definition face image, to high definition face image successively YUV color spaces
Conversion, normalized and meet the input quantity type of selection according to the feature extraction of the input quantity type of selection to obtain
, identification feature amount corresponding to the high definition face image, using identification feature amount corresponding to the high definition face image as training after
The input of the input layer of model, to obtain the eye strain grade of spectators.
In the electronic reader spectators fatigue strength analyzing platform:
The electric ink touch screen is connected with the fatigue strength extraction equipment, eye strain for receiving spectators etc.
Level.
In the electronic reader spectators fatigue strength analyzing platform:
The boombox is connected with the fatigue strength extraction equipment, for receiving the eye strain grade of spectators.
It can also include in the electronic reader spectators fatigue strength analyzing platform:
FLASH storage chips, it is connected with described image quantitative decision equipment, for prestoring pre-determined lower limit quality etc.
Level, it is additionally operable to store the default training quantity of described image quantitative decision equipment output.
In addition, it can also include in the electronic reader spectators fatigue strength analyzing platform:Time division duplex communication interface, with
The fatigue strength extraction equipment connection, for receiving and transmitting wirelessly the eye strain grade of spectators.
Time division duplex is a kind of duplex mode of communication system, is used to separate in mobile communication system to receive and transmit and believes
Road.Positive third generation development, China have submitted 3G (Third Generation) Moblie draft standard (TD- to mobile communication in June, 1997 at present
SCDMA), the characteristic such as its tdd mode and smart antenna new technology is greatly prized by and into one of three main candidate standard.
Fdd mode rules all the land in the first generation and second generation mobile communication system, and tdd mode does not draw attention.But due to new business
Needs and new technology development, and many advantages of tdd mode, tdd mode will be increasingly subject to pay attention to.
The operation principle of time division duplex is as follows:TDD is a kind of duplex mode of communication system, is used in mobile communication system
Received and transmission channel (or uplink downlink) in separation.Received in the GSM of tdd mode and transmission is same
Frequency channels are the different time-gap of carrier wave, are received and transmission channel with the time is ensured to separate;And the mobile communication system of fdd mode
The reception and transmission of system are on two symmetrical frequency channels of separation, are received and transmission channel with frequency range is ensured to separate.
It is different using the GSM feature and communication benefit of different dual-modes.The movement of tdd mode is led to
The same frequency of up-downgoing channel in letter system, thus the reciprocity with up-downgoing channel, this leads to the movement of tdd mode
Letter system brings many advantages.
In tdd mode, in up-link and downlink the transmission of information can be carried out in same carrier frequency, i.e.,
The transmission of information is to be realized on same carrier wave by the time-division in the transmission of information and downlink in up-link.
Using electronic reader spectators' fatigue strength analyzing platform of the present invention, for electronic reader spectators in the prior art
State is difficult to the technical problem detected, by transforming electronic reader framework, adds multiple targetedly IMAQs and sets
Standby and image processing equipment, it is crucial that, employ fatigue of the adaptive deep neural network to electronic reader spectators
Degree is detected in time, is timely reminded or is alerted so as to provide the spectators in fatigue state.
It is understood that although the present invention is disclosed as above with preferred embodiment, but above-described embodiment and it is not used to
Limit the present invention.For any those skilled in the art, without departing from the scope of the technical proposal of the invention,
Many possible changes and modifications are all made to technical solution of the present invention using the technology contents of the disclosure above, or are revised as
With the equivalent embodiment of change.Therefore, every content without departing from technical solution of the present invention, the technical spirit pair according to the present invention
Any simple modifications, equivalents, and modifications made for any of the above embodiments, still fall within the scope of technical solution of the present invention protection
It is interior.