One kind is based on character image identification audio reading method and its device
Technical field
The present invention relates to intelligent arrangement for reading technical fields, and in particular to one kind identifies audio reading side based on character image
Method and its device.
Background technique
As intelligence technologically continues to develop, there are more and more intelligence and read auxiliary tool, wherein talking pen is
It is that the New Generation of Intelligent after learning machine, point reader is read and educated using the high-tech product of optical image recognition technology
Learning tool.At present general talking pen software and hardware architecture include Sensor (infrared photosensitive), MCU, OID algorithm, can reflect it is red
The special coating of outer light prints books (i.e. mating books), the core OID algorithm of talking pen, and OID is also known as photosensitive pen or optics
Identification instrument, principle are exactly by the instrument with lighting apparatus come the digital signal of influence chart on piece, thus to serial book is matched
Certain response occurs in this.There is a kind of thing to be called pen in OID, pen of the pen here namely described in us, but he
It is not our common pens to write again, but the inside is integrated with a kind of optical sensing and identifier of some electronic components
Device.It is to be sensed using contact in OID, i.e. pen touches mating books, then perceives the information of mating books, last root
Certain reflection is made according to the information received.Contain one layer of ins and outs layer on its mating books, contains an institute on ins and outs layer
The information that can be sensed is incorporated into some code code to its picture or text according to the demand of specific mating books, and each
A code is to have digital number index to be identified to it, when pen light pen recognizes a code,
First identify its No. index, then by this No. index reflection to own chip the inside (certain chip be write in advance as
Event driven program corresponding with these pictures, event driven program can also generate according to OID software), then pen meeting
Certain response is made according to the content stored inside chip.This is in fact the same just as computer, when user gives a task,
It can realize this task according to the program being previously stored in light pen.For example, using most voice function in terms of OID
Can, i.e., the content on picture is clicked by pen and made a sound.
But talking pen is confined to the mating books of containing the ins and outs layer, and to common books without recognition capability, mating books in production
Coding it is at high cost, can not be widely used.
Summary of the invention
The purpose of the present invention is to provide one kind based on character image identification audio reading method and its device, to solve
The aid reading tool of existing talking pen etc can not be applied to the problem of common language material.
To achieve the above object, technical solution of the present invention provides a kind of based on character image identification audio reading side
Method, this method comprises:
The character image information on written material is acquired, makes each character image information corresponding at least by autozoom
The figure of one letter symbol;
The character image information is pre-processed, identifies that the character image information is gone forward side by side row text information matches;
The matched text information is matched with the word audio information of audio database, and by the text sound
Frequency information is played in real time.
Further, it is described to the character image information carry out pretreatment include:
Gray proces are carried out to the character image information, and the character image information is carried out using adaptive threshold
Threshold process increases contrast.
Further, the character image information on the acquisition written material, makes each text figure by autozoom
As the figure of the corresponding at least letter symbol of information includes:
When camera is close to written material described in simultaneously face, according to the font size of the written material, by automatic
Zoom makes the figure for only having a letter symbol in each picture.
Further, when the character image information recognition failures or the word audio information matches fail, hair
Warning note out.
Further, the audio database and corresponding text image data library are updated using cloud database.
Further, the text information identified using a display screen real-time display, and show what cloud database was sent
The corresponding literal interpretation of the text information.
Based on the same inventive concept, present invention also provides one kind identifies audio reading device, the dress based on character image
Setting includes: device noumenon, the autozoom camera that the loudspeaker of device noumenon upper end is arranged in, device noumenon end is arranged in
And it is arranged in the intrinsic main control board of described device and lithium battery;It is integrally disposed on the main control board to have micro process
Device, and the figure identification chip, stereo process chip, audio memory and the picture that are electrically connected with the microprocessor store
Device, the character image information of the autozoom camera acquisition written material, passes through the figure identification chip and the figure
Piece memory matched identifies the character image information, and the microprocessor is by the text information identified and the audio storage
The audio database of device is matched, by being sent to the loudspeaker after the stereo process chip.
Further, be additionally provided with power management chip on the main control board, the power management chip with it is described
Microprocessor connection, the corresponding power management chip are provided with charging interface.
Further, display screen and auxiliary camera are additionally provided on described device ontology, the display screen and described auxiliary
Camera is helped to connect with the microprocessor.
Further, described device ontology two sides be respectively arranged with USB data interface, audio key, switching key and
Earphone jack.
Optionally, wireless communication chips are provided on the main control board, for carrying out wireless communication with intelligent terminal.
The wireless communication chips are WiFi communication device, bluetooth communication device, ZigBee communication device or twireless radio-frequency communication device
Part etc..
The present invention has the advantage that
It is provided in an embodiment of the present invention that audio reading method and its device are identified based on character image, pass through and acquires text material
The character image information of material carries out the text information of character image information using image recognition technology and Audio Matching technology
It identifies and passes through loudspeaker and play corresponding audio, it is greatly convenient so as to carry out aid reading to common language material
The reading of user.
Detailed description of the invention
Provided in an embodiment of the present invention kind of Fig. 1 identifies audio reading method flow diagram based on character image.
Provided in an embodiment of the present invention kind of Fig. 2 identifies audio reading device left view structure chart based on character image.
Provided in an embodiment of the present invention kind of Fig. 3 identifies the right view structure chart of audio reading device based on character image.
Provided in an embodiment of the present invention kind of Fig. 4 identifies audio reading device use state structure chart based on character image.
Provided in an embodiment of the present invention kind of Fig. 5 identifies audio reading device system structure diagram based on character image.
Specific embodiment
The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention..
Embodiment 1
Audio reading method, this method packet are identified based on character image as shown in Figure 1, the embodiment of the invention provides one kind
It includes:
Character image information on S101, acquisition written material, makes each character image information pair by autozoom
Answer the figure of an at least letter symbol;
Character image information only one or several letter symbols acquired by autozoom, so that text figure
As the spatial cache that information occupies is small, and the composition of image information is not loaded, the resource expended when reducing Text region.
S102, the character image information is pre-processed, identify the character image information and carries out text information
Matching;
Pass through character image number so that character image information is clear and is conducive to be identified by gray scale, contrast processing
It is matched according to grapholect image pre-stored in library, to achieve the purpose that Text region.Such as pytesser, OCR
In Python using the Tesseract engine from Google, these Text region application programs can be very
The good above-mentioned work of completion.
S103, the matched text information is matched with the word audio information of audio database, and will be described
Word audio information is played in real time.
Wherein, it is described to the character image information carry out pretreatment include:
Gray proces are carried out to the character image information, and the character image information is carried out using adaptive threshold
Threshold process increases contrast.
Wherein, the character image information on the acquisition written material, believes each character image by autozoom
The figure for ceasing a corresponding at least letter symbol includes:
When camera is close to written material described in simultaneously face, according to the font size of the written material, by automatic
Zoom makes the figure for only having a letter symbol in each picture.Due to only one letter symbol of each picture, text
Word identifies low in energy consumption, to be suitable for small-sized mobile device.
Optionally, it when the character image information recognition failures or the word audio information matches fail, issues
Warning note.
Optionally, the audio database and corresponding text image data library are updated using cloud database.
Optionally, the text information identified using a display screen real-time display, and show the institute that cloud database is sent
State the corresponding literal interpretation of text information.
Embodiment 2
Based on the same inventive concept, as shown in Figure 2-5, present invention also provides one kind is read based on character image identification audio
Read apparatus, the device include: device noumenon 10, and the loudspeaker 20 of 10 upper end of device noumenon is arranged in, is arranged in device noumenon 10
The autozoom camera 30 of end and the main control board 40 and lithium battery 50 being arranged in described device ontology 10;It is described
Figure identification chip, mixed integrally disposed on main control board 40 to have a microprocessor, and being electrically connected with the microprocessor
Sound handles chip, audio memory and picture memory, and the autozoom camera 30 acquires the character image of written material
Information passes through character image information described in the figure identification chip and the picture memory match cognization, the micro process
Device matches the text information identified with the audio database of the audio memory, passes through the stereo process chip
After be sent to the loudspeaker 20.
Wherein, the focal length variations section of autozoom camera 30 is between 10mm-30mm, using electronics autozoom,
Make only have one or several letter symbols in character image information under micro-processor control.Figure identification chip can be with
It is the MA2450 chip of Movidius.
Wherein, be additionally provided with power management chip on the main control board 40, the power management chip with it is described micro-
Processor connection, the corresponding power management chip are provided with charging interface 43 and switch key 41.
Wherein, display screen 60 and auxiliary camera 70, the display screen 60 and institute are additionally provided on described device ontology 10
Auxiliary camera 70 is stated to connect with the microprocessor.Auxiliary camera 70 is larger for acquiring distinguishingly remote or volume
Letter symbol.
Wherein, described device ontology two sides are respectively arranged with USB data interface 42, audio key 45, switching key 46, ear
Machine transplanting of rice hole 44 and repeat playing key 47.Switching key 46 is used to show the corresponding explanation information of text by display screen.
Optionally, wireless communication chips are provided on the main control board 40, for carrying out channel radio with intelligent terminal
Letter.The wireless communication chips are WiFi communication device, bluetooth communication device, ZigBee communication device or twireless radio-frequency communication
Device etc..
As shown in Figure 2,4, the light transmission contact 31 of arc is provided in the end of device noumenon, in use, light transmission contact exists
Written material sliding, since in order to adapt to manpower operation, light transmission contact 31 tilts 8-15 degree, corresponding light transmission contact in vertical direction
31 are provided with reflecting optics 32, for enable incident ray vertical enter autozoom camera.
It is provided in an embodiment of the present invention that audio reading method and its device are identified based on character image, pass through and acquires text material
The character image information of material carries out the text information of character image information using image recognition technology and Audio Matching technology
It identifies and passes through loudspeaker and play corresponding audio, it is greatly convenient so as to carry out aid reading to common language material
The reading of user.
Although above having used general explanation and specific embodiment, the present invention is described in detail, at this
On the basis of invention, it can be made some modifications or improvements, this will be apparent to those skilled in the art.Therefore,
These modifications or improvements without departing from theon the basis of the spirit of the present invention are fallen within the scope of the claimed invention.