CN108229428A - A kind of character recognition method, device, server and medium - Google Patents
A kind of character recognition method, device, server and medium Download PDFInfo
- Publication number
- CN108229428A CN108229428A CN201810088803.5A CN201810088803A CN108229428A CN 108229428 A CN108229428 A CN 108229428A CN 201810088803 A CN201810088803 A CN 201810088803A CN 108229428 A CN108229428 A CN 108229428A
- Authority
- CN
- China
- Prior art keywords
- identified
- text information
- graph text
- extraction model
- feature point
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
- G06V30/333—Preprocessing; Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
- G06V30/36—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Discrimination (AREA)
Abstract
The embodiment of the invention discloses a kind of character recognition method, device, server and medium, the method includes：Obtain graph text information to be identified；The characteristic point of the graph text information to be identified is extracted from the graph text information to be identified by the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance；The characteristic point with the character features point in the literal pool pre-established is matched, determines word corresponding with the graph text information to be identified.Method provided in an embodiment of the present invention goes out the characteristic point of more accurately graph text information to be identified by personalized feature point extraction model extraction, the word of handwriting input in electronic memo is accurately identified according to the user of current operation, and then user is facilitated to perform more multioperation, improve user experience.
Description
Technical field
The present embodiments relate to information technology field more particularly to a kind of character recognition method, device, server and Jie
Matter.
Background technology
With the development of smart electronics product, traditional hard copy data strides forward towards electronization gradually, due to electronic bits of data
It easily preserves and carries, traditional papery data is replaced gradually by electronic bits of data, is gradually evolved to using electronic device
The record of carry out data records memorandum to provide user and the functions such as even draws.For example, attend class or carry out meeting note
During record, without the use of traditional notebook, as long as carrying the electronic memo for having word input function, you can use it to record
The content that need to be recorded.
The common electronic memo for having hand-written transmission function at present is mostly by by the word of user's handwriting input
It is preserved in the form of picture.But can only again be checked for user using the word content that picture preserves, it can not provide to the user
It is more multi-functional, such as pass through keyword search, again edition function.
Invention content
In view of the above-mentioned problems, an embodiment of the present invention provides a kind of character recognition method, device, server and medium, with
It realizes the word of identification handwriting input, and then user is facilitated to perform more multioperation, improve user experience.
In a first aspect, an embodiment of the present invention provides a kind of character recognition method, including：
Obtain graph text information to be identified；
By the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance from described
The characteristic point of the graph text information to be identified is extracted in graph text information to be identified；
The characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described to be identified
The corresponding word of graph text information.
Second aspect, an embodiment of the present invention provides a kind of character recognition device, including：
Data obtaining module, for obtaining graph text information to be identified；
Feature point extraction module, for passing through the spy corresponding with the graph text information owning user to be identified of training in advance
Sign point extraction model extracts the characteristic point of the graph text information to be identified from the graph text information to be identified；
A word determining module, for the character features point in the characteristic point and the literal pool that pre-establishes to be carried out
Match, determine word corresponding with the graph text information to be identified.
The third aspect, an embodiment of the present invention provides a kind of server, the server includes：
One or more processors；
Storage device, for storing one or more programs；
When one or more of programs are performed by one or more of processors so that one or more of processing
Device realizes the character recognition method provided such as any embodiment of the present invention.
Fourth aspect, the embodiment of the present invention additionally provide a kind of computer readable storage medium, are stored thereon with computer
Program realizes the character recognition method provided such as any embodiment of the present invention when the program is executed by processor.
The embodiment of the present invention passes through training in advance and the graph text information to be identified by obtaining graph text information to be identified
The corresponding feature point extraction model of owning user extracts the graph text information to be identified from the graph text information to be identified
Characteristic point goes out the characteristic point of more accurately graph text information to be identified by personalized feature point extraction model extraction, by described in
Characteristic point is matched with the character features point in the literal pool pre-established, is determined corresponding with the graph text information to be identified
Word can accurately identify the word of handwriting input in electronic memo according to the user of current operation, and then user is facilitated to hold
Row more multioperation, improves user experience.
Description of the drawings
Fig. 1 is the flow chart of the character recognition method in the embodiment of the present invention one；
Fig. 2 is the flow chart of the character recognition method in the embodiment of the present invention two；
Fig. 3 is the flow chart of the character recognition method in the embodiment of the present invention three；
Fig. 4 is the structure diagram of the character recognition device in the embodiment of the present invention four；
Fig. 5 is the structure diagram of the server in the embodiment of the present invention five.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limitation of the invention.It also should be noted that in order to just
Part related to the present invention rather than entire infrastructure are illustrated only in description, attached drawing.
Embodiment one
Fig. 1 is the flow chart of the character recognition method in the embodiment of the present invention one, and the present embodiment is applicable to electronic notebook
The situation of the word of this identification handwriting input.This method can be performed by character recognition device, which can adopt
It is realized with the mode of software and/or hardware, for example, the character recognition device is configured in server.As shown in Figure 1, the party
Method specifically includes：
S110, graph text information to be identified is obtained.
In the present embodiment, graph text information to be identified is the text information for needing to identify.Optionally, text information to be identified
Form can be picture format.For example, when user needs to input word on an electronic device by handwriting mode, Ke Yitong
It crosses input equipment (such as felt pen, writing pencil or finger) and word is slidably inputed on electronic device touchscreen.Pass through inspection at this time
The touch location for calculating input equipment is surveyed, according to touch location formation and the corresponding word of user's input, and is stored as figure
Piece form.It should be noted that word input position and text importing position can be same position, or different positions
It puts.When word input position and text importing position are in same position, go out the text of input in word input position simultaneous display
Word, at this time user possess with similarly being experienced using papery notebook, meet the writing style of user.When word input position and
Text importing position is not in same position, after word input position obtains text information input by user, in text importing
Position shows word input by user successively.
Optionally, the material of touch screen is unlimited.For example, touch screen can be resistive touch screen or capacitive touch screen.
Resistive touch screen is made of one layer of deformable resistance film and one layer of fixed resistance film, and centre is separated by by air
From.Its operation principle is：When with pen or finger contact resistance formula touch tablet, upper resistors compressive deformation simultaneously connects with lower resistors
It touches, lower resistors film is with regard to that can induce the position of pen or finger.Capacitive touch screen is to sense to carry out using the electric current of human body
Work.Capacitive touch screen is one piece of four-layer compound glass screen, and the inner surface and interlayer of glass screen are respectively coated with a conductive film, most
Outer layer is a thin layer silica protective layer for glass, and conductive interlayer membrane coat draws four electrodes, internal layer as working face on four angles
Conductive film is shielded layer.When finger touches on the metal layer, due to human body electric field, user and touch screen surface are formed with one
Coupled capacitor, for high-frequency current, capacitance is direct conductor, and then finger siphons away the electric current of a very little from contact point.
This electric current point is flowed out from the electrode on the quadrangle of touch screen, and flows through the electric current of this four electrodes and finger arrives quadrangle
Apart from directly proportional, by the accurate calculating to this four current ratios, the position of touch point is obtained.
When detecting that input equipment presses touch screen, word is formed in text importing position, is shown input by user
Text information finally includes the picture of text information input by user.
In the present embodiment, the formation to graph text information to be identified and acquisition modes are not limited.For example, it can detect
After the completion of being fully entered to user, formed and include the pictures of all text informations, can also prefixed time interval, timing formed
Include the picture of the inputting word information before current time, full line or fixed line number can also be completed detecting user
During input, the picture for including corresponding row text information is formed.
Optionally, the acquisition graph text information to be identified includes：
Obtain the picture of graph text information to be identified；
The text profile information of the graph text information to be identified is extracted from the pictorial information.
Specifically, when having detected word input, the picture of graph text information to be identified is obtained first, then from acquisition
The text profile of graph text information to be identified is extracted in picture.In the present embodiment, to the word of extraction graph text information to be identified
The mode of profile is not limited.For example, word paragraph segmentation and identification can be carried out to picture, fall into a trap from whole profiles of image
It calculates and extracts the boundary rectangle for meeting text profile and obtain the position of each text profile, finally obtained according to the position of character contour
To the text profile information of paragraph segmentation.
In another embodiment of the invention, it can also be grasped in real time according to input when having detected word input
Make detected touch point position, directly according to the text profile information of touch point position acquisition graph text information to be identified.It obtains in real time
Text profile information can operate the segmentation of image and identification to avoid in subsequent step, make the text profile information of acquisition more
Accurately.
S120, by the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance from
The characteristic point of the graph text information to be identified is extracted in the graph text information to be identified.
In the present embodiment, graph text information owning user to be identified is the user for inputting graph text information to be identified.Due to every
Personal ways of writing is different, therefore carries out based on fixed character recognition method that Text region is more difficult and accuracy rate is relatively low.Though
So everyone ways of writing in different moments has nuance, but everyone whole ways of writing is similar.Therefore,
In the present embodiment, different users is corresponded to different feature point extraction models, when having detected word input, is used
Feature point extraction model corresponding with the user of current input word carries out the extraction of characteristic point.Personalized sign point extraction mould
Type can make the extraction of character features point more accurate, so as to improve the accuracy rate of Text region.Optionally, acquisition is waited to know
The text profile information of other graph text information is input in feature point extraction model, it is right that feature point extraction model passes through as input
The analysis of the text profile information of input and feature point extraction extract characteristic point and the output of graph text information to be identified, at this time
The characteristic point of the graph text information to be identified of feature point extraction model output is obtained, to be carried out in subsequent step using this feature point
The identification of word.
Optionally, to determining that the mode of graph text information owning user to be identified is not limited.For example, it can be marked by user
Know and determine graph text information owning user to be identified；When detecting user's operation, the mode of operation of active user can also be analyzed,
And match current user operation mode with each user's history operating habit, the high user of matching degree is determined as currently holding
The user of row operation, i.e., the owning user of graph text information to be identified.
S130, the characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described
The corresponding word of graph text information to be identified.
After the characteristic point for getting graph text information to be identified, it is carried out with the character features point that is stored in literal pool
Match, the corresponding word of graph text information to be identified is determined according to matching result.
Optionally, matching degree threshold value can be pre-set.After the characteristic point for obtaining graph text information to be identified, by figure to be identified
The characteristic point of literary information characteristic point corresponding with word in literal pool is matched, and is calculated in graph text information to be identified and literal pool
The quantity of the word same characteristic features point of storage and the ratio of characteristic point total quantity, and as graph text information to be identified and currently
The Feature Points Matching degree of word.When the matching degree of graph text information to be identified characteristic point corresponding with word a certain in literal pool is higher than
During preset matching degree threshold value, as the corresponding word of graph text information to be identified.If the characteristic point of graph text information to be identified with
The highest word of matching degree then is determined as waiting to know by the matching degree of the corresponding characteristic point of multiple words higher than preset matching degree threshold value
The corresponding word of other graph text information, using other words as the corresponding candidate character of graph text information to be identified, to know in word
Candidate character is provided when not wrong to change for user.Wherein, preset matching degree threshold value can be 80%.
The embodiment of the present invention passes through training in advance and the graph text information to be identified by obtaining graph text information to be identified
The corresponding feature point extraction model of owning user extracts the graph text information to be identified from the graph text information to be identified
Characteristic point goes out the characteristic point of more accurately graph text information to be identified by personalized feature point extraction model extraction, by described in
Characteristic point is matched with the character features point in the literal pool pre-established, is determined corresponding with the graph text information to be identified
Word can accurately identify the word of handwriting input in electronic memo according to the user of current operation, and then user is facilitated to hold
Row more multioperation, improves user experience.
Embodiment two
Fig. 2 is the flow chart of the character recognition method in the embodiment of the present invention two, and the present embodiment is using above-described embodiment as base
Plinth is further optimized.As shown in Fig. 2, the method includes：
S210, graph text information to be identified is obtained.
S220, identification information corresponding with the graph text information owning user to be identified is obtained.
In the present embodiment, in order to more accurately identify text information input by user, graph text information to be identified is being identified
Before, it is thus necessary to determine that the user of current input word.Optionally, current use can be determined by obtaining the log-on message of user
The identification information (such as user name) at family；When detecting user's operation, the mode of operation of active user can also be analyzed, and ought
Preceding user's operation mode is matched with each user's history operating habit, and the high user of matching degree is determined as currently to perform operation
User, i.e., the owning user of graph text information to be identified obtains the identification information of the user as belonging to graph text information to be identified
The corresponding identification information of user.
S230, feature point extraction mould trained in advance corresponding with the word to be identified is determined according to the identification information
Type.
Optionally, the correspondence of feature point extraction model and user identity information is pre-established, determines current input text
After the user identity information of word, the corresponding feature point extraction model of active user is determined according to the identification information of user.
For example, when the identification information for obtaining active user is " ID：During yonghu ", feature corresponding with " yonghu " is selected
Point extraction model, as the feature point extraction model for the characteristic point for extracting graph text information to be identified.
S240, the characteristic point for going out the graph text information to be identified by the feature point extraction model extraction.
In the present embodiment, after determining feature point extraction model corresponding with word owning user to be identified, using determining
Feature point extraction model to graph text information to be identified carry out characteristic point extraction.
S250, the characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described
The corresponding word of graph text information to be identified.
The technical solution of the present embodiment is embodied through training in advance and the graph text information owning user to be identified
Corresponding feature point extraction model extracts the characteristic point of the graph text information to be identified from the graph text information to be identified
Process by obtaining the identification information of graph text information owning user to be identified, determines and figure to be identified according to user identity information
The corresponding feature point extraction model of literary information owning user, carries out graph text information to be identified personalized feature point extraction, makes
The feature point extraction of graph text information to be identified is more accurate, so as to improve the accuracy of graph text information Text region to be identified.
Embodiment three
Fig. 3 is the flow chart of the character recognition method in the embodiment of the present invention three, and the present embodiment is using above-described embodiment as base
Plinth is further optimized.As shown in figure 3, the method includes：
S310, original character and grapholect corresponding with the original character are obtained, by the original character and described
Word training sample of the grapholect as feature point extraction model to be trained.
In the present embodiment, it before the Text region for carrying out graph text information to be identified, needs to feature point extraction model
Carry out personalized training.
Optionally, feature point extraction model to be trained corresponding with user identifier is initially set up, user is prompted to carry out special
The input of sign point extraction model word training sample, to obtain the word training sample of feature point extraction model.Such as obtain user
The original character repeatedly inputted and the grapholect of user's selection, as the word training sample of feature point extraction model.
S320, the feature point extraction model to be trained is trained using the word training sample, is instructed
The feature point extraction model perfected.
In the present embodiment, the feature point extraction model of foundation is trained using the word training sample of acquisition.Example
Such as, the corresponding multiple original characters of grapholect user selected are as input, using grapholect as output, training characteristics
Extracting parameter in point extraction model, the final extracting parameter confirmed in feature point extraction model, obtains trained characteristic point
Extraction model.
It optionally, can also be to trained feature point extraction model after trained feature point extraction model is obtained
It is corrected.For example, using the original character in word training pattern as input, extraction is joined according to the character features of output point
Number is corrected, and improves the accuracy of feature point extraction model.
S330, graph text information to be identified is obtained.
S340, by the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance from
The characteristic point of the graph text information to be identified is extracted in the graph text information to be identified.
S350, the characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described
The corresponding word of graph text information to be identified.
The technical solution of the present embodiment on the basis of said program, is increased and is instructed according to original character and grapholect
Practice the operation of feature point extraction model, original character and grapholect training characteristics point extraction model repeatedly inputted by user,
The feature point extraction that can make feature point extraction model is more accurate, and then improves the accuracy rate of Text region.
On the basis of said program, after trained feature point extraction model is obtained, further include：
Go out the characteristic point of each original character, and establish literal pool using trained feature point extraction model extraction, it is described
Characteristic point comprising each original character and the correspondence of the grapholect in literal pool.
Optionally, Text region is in addition to the characteristic point for needing to extract text information to be identified, it is also necessary to will extract
Characteristic point be compared with the character features point in literal pool, to determine the corresponding word of text information to be identified.Therefore, it needs
Establish the literal pool for including grapholect and feature point correspondence.
Optionally, can be after feature point extraction model be trained, according to feature point extraction model, corresponding user marks
Know, establish literal pool corresponding with user identifier.As each user establishes personalized literal pool, wherein being stored with the user
The characteristic point of original character and the correspondence of grapholect.It can also be established total after feature point extraction model is trained
Literal pool, wherein including the characteristic point of the original character of grapholect different user input corresponding with the word.
Correspondingly, if each user is corresponding with personalized literal pool, need by the characteristic point with pre-establishing
Literal pool in character features point matched, before determining word corresponding with the graph text information to be identified, according to
Family mark determines literal pool corresponding with active user, and of character features point is carried out from the corresponding literal pool of active user
Match；If all users correspond to same literal pool, without determining literal pool corresponding with active user according to user identifier, directly
The matching of character features point is carried out in literal pool.
Example IV
Fig. 4 is the structure diagram of the character recognition device in the embodiment of the present invention four.The character recognition device can be adopted
It is realized with the mode of software and/or hardware, such as the character recognition device can be configured in server, as shown in figure 4, described
Device includes：
Data obtaining module 410, for obtaining graph text information to be identified；
Feature point extraction module 420, for passing through the corresponding with the graph text information owning user to be identified of training in advance
Feature point extraction model the characteristic point of the graph text information to be identified is extracted from the graph text information to be identified；
Word determining module 430, for the character features point in the characteristic point and the literal pool pre-established to be carried out
Matching determines word corresponding with the graph text information to be identified.
On the basis of said program, the feature point extraction module 420 is specifically used for：
Obtain identification information corresponding with the graph text information owning user to be identified；
Feature point extraction model trained in advance corresponding with the word to be identified is determined according to the identification information；
Go out the characteristic point of the graph text information to be identified by the feature point extraction model extraction.
On the basis of said program, described device further includes：
Sample acquisition module, in the spy corresponding with the graph text information owning user to be identified by training in advance
Before sign point extraction model extracts the characteristic point of the graph text information to be identified from the graph text information to be identified, obtain former
Beginning word and grapholect corresponding with the original character, using the original character and the grapholect as to be trained
The word training sample of feature point extraction model；
Model training module, for being carried out using the word training sample to the feature point extraction model to be trained
Training, obtains trained feature point extraction model.
On the basis of said program, described device further includes：
Literal pool establishes module, for after trained feature point extraction model is obtained, using trained feature
Point extraction model extracts the characteristic point of each original character, and establish literal pool, comprising each original character in the literal pool
Characteristic point and the correspondence of the grapholect.
On the basis of said program, described information acquisition module 410 is specifically used for：
Obtain the picture of graph text information to be identified；
The text profile information of the graph text information to be identified is extracted from the pictorial information.
The character recognition device that the embodiment of the present invention is provided, which can perform the word that any embodiment of the present invention is provided, to be known
Other method has the corresponding function module of execution method and advantageous effect.
Embodiment five
Fig. 5 is the structure diagram of the server in the embodiment of the present invention five.Fig. 5 shows to be used for realizing the present invention
The block diagram of the exemplary servers 512 of embodiment.The server 512 that Fig. 5 is shown is only an example, should not be to the present invention
The function and use scope of embodiment bring any restrictions.
As shown in figure 5, server 512 is showed in the form of universal computing device.The component of server 512 can include but
It is not limited to：One or more processing unit 516, system storage 528, connection different system component is (including system storage
728 and processing unit 516) bus 518.
Bus 518 represents one or more in a few class bus structures, including memory bus or Memory Controller,
Peripheral bus, graphics acceleration port, processing unit 516 or total using the local of the arbitrary bus structures in a variety of bus structures
Line.For example, these architectures include but not limited to industry standard architecture (ISA) bus, microchannel architecture
(MAC) bus, enhanced isa bus, Video Electronics Standards Association (VESA) local bus and peripheral component interconnection (PCI) are total
Line.
Server 512 typically comprises a variety of computer system readable media.These media can be it is any being capable of bedding and clothing
The usable medium that business device 512 accesses, including volatile and non-volatile medium, moveable and immovable medium.
System storage 528 can include the computer system readable media of form of volatile memory, such as deposit at random
Access to memory (RAM) 530 and/or cache memory 532.Server 512 may further include it is other it is removable/can not
Mobile, volatile/non-volatile computer system storage medium.Only as an example, storage device 534 can be used for read-write not
Movably, non-volatile magnetic media (Fig. 5 do not show, commonly referred to as " hard disk drive ").Although being not shown in Fig. 5, can with
It provides for moving the disc driver of non-volatile magnetic disk (such as " floppy disk ") read-write and to removable non-volatile
The CD drive of CD (such as CD-ROM, DVD-ROM or other optical mediums) read-write.In these cases, each driving
Device can be connected by one or more data media interfaces with bus 518.Memory 528 can include at least one program
Product, the program product have one group of (for example, at least one) program module, these program modules are configured to perform the present invention
The function of each embodiment.
Program/utility 540 with one group of (at least one) program module 542, can be stored in such as memory
In 528, such program module 542 includes but not limited to operating system, one or more application program, other program modules
And program data, the realization of network environment may be included in each or certain combination in these examples.Program module 542
Usually perform the function and/or method in embodiment described in the invention.
Server 512 can also be with one or more external equipments 514 (such as keyboard, sensing equipment, display 524 etc.)
Communication can also enable a user to the equipment interacted with the server 512 communication and/or with causing the clothes with one or more
Any equipment (such as network interface card, modem etc.) that business device 512 can communicate with one or more of the other computing device
Communication.This communication can be carried out by input/output (I/O) interface 722.Also, server 512 can also be fitted by network
Orchestration 720 and one or more network (such as LAN (LAN), wide area network (WAN) and/or public network, such as because of spy
Net) communication.As shown in the figure, network adapter 520 is communicated by bus 518 with other modules of server 512.It should be understood that
Although not shown in the drawings, can combine server 512 uses other hardware and/or software module, including but not limited to：Micro- generation
Code, device driver, redundant processing unit, external disk drive array, RAID system, tape drive and data backup are deposited
Storage system etc..
Processing unit 516 is stored in program in system storage 528 by operation, so as to perform various functions using with
And data processing, such as realize the character recognition method that the embodiment of the present invention is provided, this method includes：
Obtain graph text information to be identified；
By the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance from described
The characteristic point of the graph text information to be identified is extracted in graph text information to be identified；
The characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described to be identified
The corresponding word of graph text information.
Certainly, it will be understood by those skilled in the art that processing unit can also realize that any embodiment of the present invention is provided
Character recognition method technical solution.
Embodiment six
The embodiment of the present invention six additionally provides a kind of computer readable storage medium, is stored thereon with computer program, should
The character recognition method provided such as the embodiment of the present invention is realized when program is executed by processor, this method includes：
Obtain graph text information to be identified；
By the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance from described
The characteristic point of the graph text information to be identified is extracted in graph text information to be identified；
The characteristic point is matched with the character features point in the literal pool pre-established, determine with it is described to be identified
The corresponding word of graph text information.
Certainly, a kind of computer readable storage medium that the embodiment of the present invention is provided, the computer program stored thereon
The method operation being not limited to the described above, can also be performed the phase in the character recognition method that any embodiment of the present invention is provided
Close operation.
The arbitrary of one or more computer-readable media may be used in the computer storage media of the embodiment of the present invention
Combination.Computer-readable medium can be computer-readable signal media or computer readable storage medium.It is computer-readable
Storage medium for example may be-but not limited to-the system of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, device or
Device or arbitrary above combination.The more specific example (non exhaustive list) of computer readable storage medium includes：Tool
There are one or the electrical connections of multiple conducting wires, portable computer diskette, hard disk, random access memory (RAM), read-only memory
(ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-
ROM), light storage device, magnetic memory device or above-mentioned any appropriate combination.In this document, computer-readable storage
Medium can be any tangible medium for including or storing program, which can be commanded execution system, device or device
Using or it is in connection.
Computer-readable signal media can include in a base band or as a carrier wave part propagation data-signal,
Wherein carry computer-readable program code.Diversified forms may be used in the data-signal of this propagation, including but it is unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By instruction execution system, device either device use or program in connection.
The program code included on computer-readable medium can be transmitted with any appropriate medium, including --- but it is unlimited
In wireless, electric wire, optical cable, RF etc. or above-mentioned any appropriate combination.
It can write to perform the computer that operates of the present invention with one or more programming language or combinations
Program code, described program design language include object oriented program language-such as Java, Smalltalk, C++,
Further include conventional procedural programming language-such as " C " language or similar programming language.Program code can be with
It fully performs, partly perform on the user computer on the user computer, the software package independent as one performs, portion
Divide and partly perform or perform on a remote computer or server completely on the remote computer on the user computer.
Be related in the situation of remote computer, remote computer can pass through the network of any kind --- including LAN (LAN) or
Wide area network (WAN)-be connected to subscriber computer or, it may be connected to outer computer (such as is carried using Internet service
Pass through Internet connection for quotient).
Note that it above are only presently preferred embodiments of the present invention and institute's application technology principle.It will be appreciated by those skilled in the art that
The present invention is not limited to specific embodiment described here, can carry out for a person skilled in the art various apparent variations,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above example to the present invention
It is described in further detail, but the present invention is not limited only to above example, without departing from the inventive concept, also
It can include other more equivalent embodiments, and the scope of the present invention is determined by scope of the appended claims.
Claims (10)
1. a kind of character recognition method, which is characterized in that including：
Obtain graph text information to be identified；
It waits to know from described by the feature point extraction model corresponding with the graph text information owning user to be identified of training in advance
The characteristic point of the graph text information to be identified is extracted in other graph text information；
The characteristic point with the character features point in the literal pool pre-established is matched, is determined and the picture and text to be identified
The corresponding word of information.
2. according to the method described in claim 1, it is characterized in that, described believed by training in advance and the picture and text to be identified
The corresponding feature point extraction model of breath owning user extracts the graph text information to be identified from the graph text information to be identified
Characteristic point include：
Obtain identification information corresponding with the graph text information owning user to be identified；
Feature point extraction model trained in advance corresponding with the word to be identified is determined according to the identification information；
Go out the characteristic point of the graph text information to be identified by the feature point extraction model extraction.
3. according to the method described in claim 1, it is characterized in that, passing through training in advance and the graph text information to be identified
The corresponding feature point extraction model of owning user extracts the graph text information to be identified from the graph text information to be identified
Before characteristic point, further include：
Original character and grapholect corresponding with the original character are obtained, the original character and the grapholect are made
Word training sample for feature point extraction model to be trained；
The feature point extraction model to be trained is trained using the word training sample, obtains trained feature
Point extraction model.
4. according to the method described in claim 3, it is characterized in that, after trained feature point extraction model is obtained, go back
Including：
Go out the characteristic point of each original character, and establish literal pool using trained feature point extraction model extraction, the word
Characteristic point comprising each original character and the correspondence of the grapholect in library.
5. according to the method described in claim 1, it is characterized in that, the acquisition graph text information to be identified includes：
Obtain the picture of graph text information to be identified；
The text profile information of the graph text information to be identified is extracted from the pictorial information.
6. a kind of character recognition device, which is characterized in that including：
Data obtaining module, for obtaining graph text information to be identified；
Feature point extraction module, for passing through the characteristic point corresponding with the graph text information owning user to be identified of training in advance
Extraction model extracts the characteristic point of the graph text information to be identified from the graph text information to be identified；
Word determining module, for the characteristic point to be matched with the character features point in the literal pool pre-established, really
Fixed word corresponding with the graph text information to be identified.
7. device according to claim 6, which is characterized in that the feature point extraction module is specifically used for：
Obtain identification information corresponding with the graph text information owning user to be identified；
Feature point extraction model trained in advance corresponding with the word to be identified is determined according to the identification information；
Go out the characteristic point of the graph text information to be identified by the feature point extraction model extraction.
8. device according to claim 6, which is characterized in that further include：
Sample acquisition module, in the characteristic point corresponding with the graph text information owning user to be identified by training in advance
Before extraction model extracts the characteristic point of the graph text information to be identified from the graph text information to be identified, original text is obtained
Word and grapholect corresponding with the original character, using the original character and the grapholect as feature to be trained
The word training sample of point extraction model；
Model training module, for being instructed using the word training sample to the feature point extraction model to be trained
Practice, obtain trained feature point extraction model.
9. a kind of server, which is characterized in that the server includes：
One or more processors；
Storage device, for storing one or more programs；
When one or more of programs are performed by one or more of processors so that one or more of processors are real
The now character recognition method as described in any in claim 1-5.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
The character recognition method as described in any in claim 1-5 is realized during execution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810088803.5A CN108229428A (en) | 2018-01-30 | 2018-01-30 | A kind of character recognition method, device, server and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810088803.5A CN108229428A (en) | 2018-01-30 | 2018-01-30 | A kind of character recognition method, device, server and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108229428A true CN108229428A (en) | 2018-06-29 |
Family
ID=62669466
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810088803.5A Pending CN108229428A (en) | 2018-01-30 | 2018-01-30 | A kind of character recognition method, device, server and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108229428A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111079490A (en) * | 2019-05-29 | 2020-04-28 | 广东小天才科技有限公司 | Recognition method for written words and electronic equipment |
| CN111160352A (en) * | 2019-12-27 | 2020-05-15 | 创新奇智(北京)科技有限公司 | Workpiece metal surface character recognition method and system based on image segmentation |
| CN113408373A (en) * | 2021-06-02 | 2021-09-17 | 中金金融认证中心有限公司 | Handwriting recognition method, system, client and server |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463101A (en) * | 2014-11-06 | 2015-03-25 | 科大讯飞股份有限公司 | Answer recognition method and system for textual test question |
| CN105550643A (en) * | 2015-12-08 | 2016-05-04 | 小米科技有限责任公司 | Medical term recognition method and device |
| CN107563382A (en) * | 2017-09-21 | 2018-01-09 | 曾传德 | The text recognition method of feature based capturing technology |
-
2018
- 2018-01-30 CN CN201810088803.5A patent/CN108229428A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463101A (en) * | 2014-11-06 | 2015-03-25 | 科大讯飞股份有限公司 | Answer recognition method and system for textual test question |
| CN105550643A (en) * | 2015-12-08 | 2016-05-04 | 小米科技有限责任公司 | Medical term recognition method and device |
| CN107563382A (en) * | 2017-09-21 | 2018-01-09 | 曾传德 | The text recognition method of feature based capturing technology |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111079490A (en) * | 2019-05-29 | 2020-04-28 | 广东小天才科技有限公司 | Recognition method for written words and electronic equipment |
| CN111160352A (en) * | 2019-12-27 | 2020-05-15 | 创新奇智(北京)科技有限公司 | Workpiece metal surface character recognition method and system based on image segmentation |
| CN111160352B (en) * | 2019-12-27 | 2023-04-07 | 创新奇智(北京)科技有限公司 | Workpiece metal surface character recognition method and system based on image segmentation |
| CN113408373A (en) * | 2021-06-02 | 2021-09-17 | 中金金融认证中心有限公司 | Handwriting recognition method, system, client and server |
| CN113408373B (en) * | 2021-06-02 | 2024-06-07 | 中金金融认证中心有限公司 | Handwriting recognition method, handwriting recognition system, client and server |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10043308B2 (en) | Image processing method and apparatus for three-dimensional reconstruction | |
| US9086798B2 (en) | Associating information on a whiteboard with a user | |
| TWI724669B (en) | Lesion detection method and device, equipment and storage medium | |
| EP3183640B1 (en) | Device and method of providing handwritten content in the same | |
| US12033411B2 (en) | Stroke based control of handwriting input | |
| CN109189879B (en) | Electronic book display method and device | |
| US20160203194A1 (en) | User terminal for displaying image and image display method thereof | |
| JPH07182492A (en) | Operating method of system | |
| Besançon et al. | Hybrid touch/tangible spatial 3D data selection | |
| CN108229428A (en) | A kind of character recognition method, device, server and medium | |
| CN109542278B (en) | Touch data processing method and device and touch equipment | |
| CN104951234B (en) | A kind of data processing method and system based on touch screen terminal | |
| CN110673747B (en) | Input candidate character display method and related device | |
| KR102440198B1 (en) | VIDEO SEARCH METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM | |
| CN109190615A (en) | Nearly word form identification decision method, apparatus, computer equipment and storage medium | |
| CN108196675A (en) | For the exchange method, device and touch control terminal of touch control terminal | |
| US20240143163A1 (en) | Digital ink processing system, method, and program | |
| CN106502536A (en) | A kind of page info choosing method and device | |
| US9058104B2 (en) | Gestures for special characters | |
| CN112840622B (en) | Pushing method and related product | |
| CN109766159A (en) | It fills in a form method for determining position, computer equipment and storage medium | |
| WO2022105120A1 (en) | Text detection method and apparatus from image, computer device and storage medium | |
| CN104007886B (en) | The method and electronic equipment of a kind of information processing | |
| WO2020210975A1 (en) | Sub-trajectory selection method, electronic device, and computer readable storage medium | |
| CN112819925A (en) | Method and device for processing focus labeling, electronic equipment and medium |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20191226 Address after: 314400 Room 101, workshop 20, Shidai Road, Haining Economic Development Zone, Haining City, Jiaxing City, Zhejiang Province Applicant after: Zhejiang everything workshop Intelligent Technology Co.,Ltd. Address before: 200233 A109 room A, building 3, 7 GUI Qing Road, Xuhui District, Shanghai. Applicant before: SHANGHAI SIYU INTELLIGENT TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201211 Address after: Room 1003, building 18, 2555 xiupu Road, Pudong New Area, Shanghai, 200120 Applicant after: Shanghai Guangai Information Technology Co.,Ltd. Address before: Room 101, No.20 Shidai Road, Haining Economic Development Zone, Jiaxing City, Zhejiang Province Applicant before: Zhejiang everything workshop Intelligent Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180629 |
|
| RJ01 | Rejection of invention patent application after publication |