CN108364036A - A kind of modeling method, recognition methods, device, storage medium and equipment - Google Patents
A kind of modeling method, recognition methods, device, storage medium and equipment Download PDFInfo
- Publication number
- CN108364036A CN108364036A CN201711459765.1A CN201711459765A CN108364036A CN 108364036 A CN108364036 A CN 108364036A CN 201711459765 A CN201711459765 A CN 201711459765A CN 108364036 A CN108364036 A CN 108364036A
- Authority
- CN
- China
- Prior art keywords
- waybill
- written
- hand
- sample
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/24—Character recognition characterised by the processing or recognition method
- G06V30/242—Division of the character sequences into groups prior to recognition; Selection of dictionaries
- G06V30/244—Division of the character sequences into groups prior to recognition; Selection of dictionaries using graphical properties, e.g. alphabet type or font
- G06V30/2455—Discrimination between machine-print, hand-print and cursive writing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Abstract
The invention discloses a kind of modeling method, recognition methods, device, storage medium and equipment, the modeling method includes：Sample is subjected to path scanning, the processing of time recurrent neural network, convolutional neural networks processing, generates full connection layer data；The full connection layer data obtains the model by CTC graders, optimization.The model realization that the identification usage is built with the modeling method, the storage medium and equipment are respectively used to realize the method.The process that the technology of the present invention can replace artificial typewriting record single, reduces a large amount of human capital, and the single accuracy rate of record is greatly improved, and technical solution can not only identify the handwritten text on waybill.
Description
Technical field
The present invention relates to handwritten Kanji recognition technical field more particularly to a kind of modeling method, recognition methods, device, deposit
Storage media and equipment.
Background technology
The accurate identification to single handwritten Chinese character has may be implemented in current manual's smart field, so by hand-written
Address text dividing the identification that full address originally may be implemented is identified at individual Chinese character and successively, still, at present " cutting
The technology of Chinese character " can ensure the degree of accuracy there are no developing to, and then cause recognition result accuracy very low.
In addition, express delivery, before being transported on road, each express delivery will undergo the single process of record, it is therefore an objective to will be on waybill
The correspondence of Quick Response Code and the hand-written address information of sender be stored in computer, and then transmitting-receiving node later only leads to
Quick Response Code is over-scanned to obtain the destination address of the express mail.Moreover, our record list process is all to rely on manually to strike at present
Keystroke, in the hand-written address typing computer system on waybill, not only to consume a large amount of human cost in this way, can not also protect
The single accuracy of card record.Once the address error of typing, what corresponding express delivery will have no suspense is mailed to the place of mistake.
As stated above, problem of the existing technology is：" technology of cutting Chinese character " there are no develop to ensure just
The degree of true rate, and then cause recognition result accuracy very low, and in delivery industry, the single process of artificial typewriting record can consume
A large amount of human cost, and the work of uninteresting repetition, can allow manual identified address often to be malfunctioned, and since hand-written address is
Do not have spaced Chinese character for a string, while Chinese character is often left and right or up-down structure, so address is cut into one by one
Chinese character be the process got half the result with twice the effort.
Invention content
In order to solve above-mentioned deficiency in the prior art, the purpose of the present invention is to provide a kind of modeling method, identification sides
Method, device, storage medium and equipment.It improves the identification accuracy of hand-written waybill, improves recognition efficiency.
To achieve the goals above, the technical solution adopted in the present invention is：
A method of hand-written waybill text identification model is established, including：
Sample is subjected to path scanning, the processing of time recurrent neural network, convolutional neural networks processing, generates full articulamentum
Data；
The full connection layer data obtains the model by CTC graders, optimization.
The path scanning is that four direction paths are scanned.
The time recurrent neural network processing, convolutional neural networks are handled in the process of implementation, including：
Sample set is equally divided into several sample sets；
Each sample in the sample set, duplicate paths scanning, LSTM processing and convolution summation process several times,
Generate full connection layer data.
It is described to optimize in the process of implementation, including：
The full connection layer data is compared and is changed with the truthful data of sample by the data that CTC graders obtain
In generation, carries out the optimization of whole network parameter, recycles training set, test set and verification collection, is optimized using gradient descent method, and be based on
GPU accelerator iteration carries out model training.
A kind of device for establishing hand-written waybill text identification model, including：
Data generating unit is configured to sample carrying out path scanning, the processing of time recurrent neural network, convolutional Neural
Network processes generate full connection layer data；
Optimize unit, is configured to the full connection layer data and obtains the model by CTC graders, optimization.
A kind of equipment, the equipment include：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of places
It manages device and executes the method for establishing hand-written waybill text identification model.
A kind of computer readable storage medium being stored with computer program, when which is executed by processor described in realization
The method for establishing hand-written waybill text identification model.
A kind of hand-written waybill text recognition method, including：
The handwritten text region in hand-written waybill to be identified is extracted, several letters are extracted in the handwritten text region
Cease picture；
Described information picture is identified by hand-written waybill text identification model, and the model establishes hand using described
The method for writing waybill text identification model is established.
A kind of hand-written waybill text identification device, which is characterized in that including：
Extraction unit is configured to extract the handwritten text region in hand-written waybill to be identified, in the handwritten text area
Several information pictures are extracted in domain；
Recognition unit is configured to described information picture and is identified by hand-written waybill text identification model, the mould
Type is established using the method for establishing hand-written waybill text identification model.
A kind of equipment, the equipment include：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of places
It manages device and executes the hand-written waybill text recognition method.
A kind of computer readable storage medium being stored with computer program, when which is executed by processor described in realization
Hand-written waybill text recognition method.
Compared with prior art, the invention has the advantages that：
1, the technical program can not only identify the handwritten text on waybill, by preparing different training samples, the party
Case can also be transplanted to the identification of the handwritten Chinese character under other scenes.
2, the single process of traditional artificial typewriting record can consume a large amount of human capitals, and in the work of uninteresting repetition,
Manual identified address is often malfunctioned, and the technical program can partly replace artificial typewriting record list by extraction target information picture
Process not only reduces a large amount of human capital, but also the single accuracy rate of record is greatly improved.
3, hand-written address be it is a string do not have a spaced Chinese character, while Chinese character is often left and right or up-down structure, institute
It is the process got half the result with twice the effort so that address is cut into Chinese character one by one, the technical program has evaded the recognition methods of traditional OCR,
A kind of text fragment recognition methods that exempting from cutting used, i.e., need not carry out Chinese character paragraph the cutting of individual Chinese character, therefore
The exemplary technical solution of the present invention can ensure Chinese Character Recognition accuracy.
Description of the drawings
Fig. 1 is the method flow diagram for establishing hand-written waybill text identification model shown in the present embodiment.
Fig. 2 is to carry out the structural schematic diagram that four direction scanning obtains one-dimensional sequence to every pictures.
Fig. 3 is the hand-written waybill text recognition method flow chart shown in the present embodiment.
Fig. 4 is the schematic diagram that hand-written address area is extracted from whole Zhang Yundan address informations picture.
Fig. 5 is the effective information region of the waybill of interception.
Fig. 6 is blank waybill.
Fig. 7 is the waybill picture after comparison, counteracting.
Specific implementation mode
The application 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 related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, is illustrated only in attached drawing and invent relevant part.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Embodiment：
As shown in Figure 1, a kind of method for establishing hand-written waybill text identification model, including：
Sample is subjected to path scanning, the processing of time recurrent neural network, convolutional neural networks processing, generates full articulamentum
Data.
S01 extracts handwritten Chinese character region from hand-written waybill picture, then extracts from the handwritten Chinese character region several
Open information picture；Using information picture as sample, sample set is created.This sample set is made of correspondence a group by a group,
That is { (picture 1, the content on picture 1), (picture 2, the content on picture 2) ... }, then by all combinations of these sample sets
It is grouped as three sample sets according to a certain percentage, the embodiment of the present invention is according to 8:1：1 ratio is grouped as training set, test
Collection and verification collection.
The step S01 includes step：
S011 is named information picture whole in step S01, and each described information picture corresponds to unique life
Name.Naming method is：Such as " 00000001.jpg ", " 00000356.jpg " etc., i.e. number total bit 8, insufficient front
Zero padding.
S012, using the word sequence of every described information picture as the label of the pictures；
Whole labels is stored in a text file by S013 in sequence, and each label accounts for a line；
S014 creates the sample set of handwritten Kanji recognition using described information picture and its label as sample.
The sample set is equally divided into several sample sets by S02；
S03 carries out path scanning to every information picture in the sample set and obtains corresponding subset with LSTM processing；
S04 carries out convolution and summation process to every pictures in the corresponding subset, obtains fisrt feature image；
S05 carries out fisrt feature image caused by same information picture at path scanning and LSTM successively respectively
Reason, convolution and summation process obtain second feature image；
S06 carries out second feature image caused by same information picture at path scanning and LSTM successively respectively
Reason, obtains intermediate image subset；
The intermediate image subset is established parameter matrix, then by relevant parameter matrix phase by S07 by full articulamentum
Add summation, finally obtains the full connection layer data of every information picture；
S08 carries out parameter normalization processing to each full connection layer data, obtains normalized parameter matrix；
S09 passes through CTC graders to each normalized parameter matrix, obtains feature vector；
Whole sample sets is carried out the processing of S03-S09 by S10, and the result of acquisition and true result are carried out
Compare the optimization that simultaneously iteration carries out whole network parameter.
The path scanning and LSTM processing include step：Four kinds of paths are carried out to every pictures and scan (four kinds of paths
Scanning is as shown in Figure 2), obtain the scanned picture of four order informations per pictures；Every scanned picture is passed through one layer
LSTM layers containing N cores respectively obtain N corresponding pictures.
LSTM (Long-Short Term Memory, shot and long term memory) is a kind of time recurrent neural network.Usually
LSTM is suitable for the critical event being spaced in processing and predicted time sequence and delay is very long.
The performance of LSTM usually more preferably than time recurrent neural network and Hidden Markov Model (HMM), for example is used in not
On zonal cooling handwriting recognition.2009, the match of ICDAR handwriting recognitions was won with the LSTM artificial nerve network models built
Champion.LSTM is also commonly used for autonomous speech recognition, and database of giving a lecture naturally with TIMIT for 2013 reaches 17.7% error rate
Record.As nonlinear model, LSTM can be used as complicated non-linear unit for constructing larger deep neural network.
The convolution and summation process include step：Every pictures are by the convolutional layer that one includes M convolution kernel
Eigenmatrix is obtained by filtration, the eigenmatrix of the picture corresponding to same information picture is correlation eigen matrix, the correspondence
It refers to by same information picture the obtained picture after above-mentioned processing, the correlation eigen matrix is summed simultaneously
It carries out nonlinear operation and obtains characteristic image.
In the step S03, N=2；In the step S05, N=10；N=50 in the step S35.
In the step S03, M=6；In the step S05, M=20.
The full connection layer data obtains initial model by CTC graders.
This programme uses recurrent neural network (Recurrent Neural on tensorflow deep learning platforms
Network LSTM (the Long-Short Term Memory) algorithms in) and CTC (Connectionist Temporal
Classification) method that algorithm combines builds deep learning network.By taking specific implementation mode as an example, initial model is established
Comprise the concrete steps that：
A. all sample decompositions are divided into 20 and separately included at the set of several same sizes, such as 10000 pictures
Then the set of 500 samples is each gathered while for carrying out parameter operation below.
B. four kinds of path scannings are carried out to each pictures in set, and then obtains four scannings for carrying order information
Then all scanned pictures are passed through the LSTM layers of one layer of core containing there are two by picture respectively, have one to carry order information respectively
Picture become two corresponding pictures.
C. each set from 500 pictures becomes 4000 pictures (this set is divided into four according to four kinds of scan paths
A subclass), these pictures are passed through to a convolutional layer comprising 6 convolution kernels respectively, corresponding eigenmatrix is obtained by filtration,
By relevant four eigenmatrixes, (four eigenmatrixes herein are to be directed in four subclass of same information picture relatively
The eigenmatrix answered, for example first, second, third, fourth subclass their first figure is exactly so-called relevant four
Eigenmatrix) it is summed and carries out nonlinear operation, and then each set is converted into 500*6 characteristic images (first again
Characteristic image).
D. it opens image by each group 6 and repeats step b as object, wherein each LSTM layers of check figure becomes 10, then
Step c is repeated, wherein the check figure of each convolutional layer is set as 20, each collection credit union obtains 500*20 characteristic images (the at this time
Two characteristic images).
E. each group of 20 images being repeated into step b as an object, wherein each LSTM layers of check figure is 50,
Then it is one group by every 50 obtained, it is four groups every (to be derived from the image with group by four acquired in four kinds of scanpaths
Group) object pass through a full articulamentum, establish 50:Then relevant four groups of additions are summed, are obtained by 3755 parameter matrix
Each original image corresponds to a full connection layer parameter.
F. flattening is carried out to the data of each full articulamentum and passes through softmax progress parameter normalizations respectively.
G. each normalized parameter matrix is passed through into a CTC grader, obtains one-dimensional including 3755 variables
Feature vector, it illustrates all words shown in this pictures.
H. all pictures are all handled by sequence described above, and the result of all acquisitions and true result are compared
Pair and iteration carry out whole network parameter optimization.
Initial model obtains the model by optimization.Network structure is mainly the final one layer of CTC compositions of three layers of LSTM, phase
Feature is extracted by one layer of CNN between two layers of LSTM of neighbour and carries out down-sampling.The input of wherein first layer LSTM is by each
It opens sample and carries out the sequence composition that four kinds of scanning directions scannings generate, such as Fig. 1.
By in the sample information picture and corresponding label according to 8:1:1 ratio is established training library, test library and is tested
Demonstrate,prove library.
The flow of training pattern is to be made iteratively to operate：1. taking a certain amount of training set sample by matching some ginsengs
It counts to establish a grader；2. taking model of a certain amount of verification collection sample to study out, the parameter of grader is adjusted；3. taking
The recognition capability of a certain amount of trained model of test set test sample is known to decide whether to carry out next iteration to improve
Not rate.
Training set, test set and verification collection based on acquisition optimize handwritten Chinese character text identification model using gradient descent method
Parameter, and model training is carried out based on GPU accelerator iteration, final obtain establishes hand-written waybill text identification model.
A kind of device for establishing hand-written waybill text identification model, including：
Data generating unit is configured to sample carrying out path scanning, the processing of time recurrent neural network, convolutional Neural
Network processes generate full connection layer data；
Optimize unit, is configured to the full connection layer data and obtains the model by CTC graders, optimization.
A kind of equipment, the equipment include：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of places
It manages device and executes the method for establishing hand-written waybill text identification model.
A kind of computer readable storage medium being stored with computer program, when which is executed by processor described in realization
The method for establishing hand-written waybill text identification model.
As shown in Figure 3-4, a kind of hand-written waybill text recognition method, includes the following steps：
S1 extracts handwritten Chinese character region from hand-written waybill picture to be identified, then is extracted from the handwritten Chinese character region
Go out several information pictures；
The step S1 includes step：
S11 obtains hand-written waybill picture to be identified；
S12 adjusts the hand-written waybill picture to be identified, makes the hand-written waybill picture to be identified and object reference picture
Between angle of inclination in preset error range；The object reference picture is by hand-written waybill picture to be identified according to system
The picture that predetermined manner is put.The concept at angle of inclination is：Appoint in hand-written waybill picture to be identified and takes (A, B) at 2 points even
It is in alignment to be defined as first straight line, since hand-written waybill picture to be identified is that the same picture is with object reference picture
Disposing way is different, so can find (A1, B1) with corresponding A in images to be recognized, B, A1 in object reference picture at 2 points
It draws a straight line with 2 points of B1 and is defined as target line, which is target line defined in object reference picture, described
Angle between first straight line and target line is angle of inclination.
S13 matches adjusted images to be recognized in step S12 with background picture, obtains in images to be recognized
Handwritten Chinese character region；The background image is the image behind hand-written waybill picture removal handwriting region to be identified；It will the back of the body
In scape image white space (i.e. blank fortune is defined as locking the region in handwritten Chinese character region in hand-written waybill picture to be identified
Destination address region in list), the handwritten Chinese character area in target image is gone out using the feature recognition of white space in background image
Domain.I.e. according in blank waybill picture the characteristics of destination address region, the characteristics of with destination address region upper and lower, left and right edge
It is matched with waybill picture to be identified, to lock the destination address region in waybill.
S14 obtains the picture in effective information region from handwritten Chinese character region；
S15 extracts several information pictures from the picture in the effective information region.
As shown in Fig. 2, adjust the waybill picture, make the bottom edge of the waybill picture with horizontal misalignment angle pre-
If error range in；
As shown in fig. 6-7, adjusted waybill picture in step S12 compared with blank waybill picture, offset, obtained
Obtain the fill substance in waybill picture.
As shown in figure 5, intercepting the 45%-70% of whole Zhang Yundan pictures using the waybill picture left side edge as initial edge
Region obtains the picture in effective information region as effective information region (value 60% when specific implementation).
S15 extracts several information pictures from the picture in the effective information region.
S2, described information picture are identified by hand-written waybill text identification model, and the model is built using described
The method for founding hand-written waybill text identification model is established.
The step S2 includes step：
Whole information pictures is sequentially input the hand-written waybill text identification model to be identified；It will be above-mentioned whole
Recognition result, which is spliced and rectified a deviation, obtains final recognition result.
A kind of hand-written waybill text identification device, which is characterized in that including：
Extraction unit is configured to extract the handwritten text region in hand-written waybill to be identified, in the handwritten text area
Several information pictures are extracted in domain；
Recognition unit is configured to described information picture and is identified by hand-written waybill text identification model, the mould
Type is established using the method for establishing hand-written waybill text identification model.
A kind of equipment, the equipment include：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of places
It manages device and executes the hand-written waybill text recognition method.
A kind of computer readable storage medium being stored with computer program, when which is executed by processor described in realization
Hand-written waybill text recognition method.
The single process of traditional artificial typewriting record can consume a large amount of human capitals, and in the work of uninteresting repetition, people
Work identification address is often malfunctioned, and the process that technical solution of the present invention can partly replace artificial typewriting record single not only reduces a large amount of
Human capital, and the single accuracy rate of record is greatly improved.Moreover, hand-written address be it is a string do not have spaced Chinese character,
Chinese character is often left and right or up-down structure simultaneously, so it is the mistake got half the result with twice the effort that address, which is cut into Chinese character one by one,
Journey, the technical program have evaded the recognition methods of traditional OCR, a kind of text fragment recognition methods for exempting from cutting of use, i.e., not
The cutting to Chinese character paragraph progress individual Chinese character is needed, therefore this programme can ensure Chinese Character Recognition accuracy.
By preparing different training samples, the program can also be transplanted to the identification of the handwritten Chinese character under other scenes
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.People in the art
Member should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from the inventive concept, it is carried out by above-mentioned technical characteristic or its equivalent feature
Other technical solutions of arbitrary combination and formation.Such as features described above has similar work(with (but not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
In addition to the technical characteristic described in specification, remaining technical characteristic is the known technology of those skilled in the art, is prominent
Go out the innovative characteristics of the present invention, details are not described herein for remaining technical characteristic.
Claims (11)
1. a kind of method for establishing hand-written waybill text identification model, which is characterized in that including：
Sample is subjected to path scanning, the processing of time recurrent neural network, convolutional neural networks processing, generates the full connection number of plies
According to；
The full connection layer data obtains the model by CTC graders, optimization.
2. according to the method described in claim 1, it is characterized in that, path scanning is the scanning of four direction paths.
3. according to the method described in claim 1, it is characterized in that, time recurrent neural network processing, convolutional Neural net
Network is handled in the process of implementation, including：
Sample set is equally divided into several sample sets；
Each sample in the sample set, duplicate paths scanning, LSTM processing and convolution summation process several times, generate
Full connection layer data.
4. according to the method described in claim 1, it is characterized in that, it is described optimization in the process of implementation, including：
By the data that the full connection layer data is obtained by CTC graders be compared with the truthful data of sample and iteration into
The optimization of row whole network parameter is recycled training set, test set and verification collection, is optimized using gradient descent method, and added based on GPU
Fast device iteration carries out model training.
5. a kind of device for establishing hand-written waybill text identification model, which is characterized in that including：
Data generating unit is configured to sample carrying out path scanning, the processing of time recurrent neural network, convolutional neural networks
Processing generates full connection layer data；
Optimize unit, is configured to the full connection layer data and obtains the model by CTC graders, optimization.
6. a kind of equipment, which is characterized in that the equipment includes：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of processors
Execute the method as described in any one of claim 1-4.
7. a kind of computer readable storage medium being stored with computer program, which is characterized in that the program is executed by processor
Methods of the Shi Shixian as described in any one of claim 1-4.
8. a kind of hand-written waybill text recognition method, which is characterized in that including：
The handwritten text region in hand-written waybill to be identified is extracted, several hum patterns are extracted in the handwritten text region
Piece；
Described information picture is identified by hand-written waybill text identification model, and the model is any using claim 1-4
The method is established.
9. a kind of hand-written waybill text identification device, which is characterized in that including：
Extraction unit is configured to extract the handwritten text region in hand-written waybill to be identified, in the handwritten text region
Extract several information pictures；
Recognition unit is configured to described information picture and is identified by hand-written waybill text identification model, the model profit
It is established with any methods of claim 1-4.
10. a kind of equipment, which is characterized in that the equipment includes：
One or more processors；
Memory, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of processors
Execute method as claimed in claim 8.
11. a kind of computer readable storage medium being stored with computer program, which is characterized in that the program is executed by processor
Shi Shixian methods as claimed in claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711459765.1A CN108364036A (en) | 2017-12-28 | 2017-12-28 | A kind of modeling method, recognition methods, device, storage medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711459765.1A CN108364036A (en) | 2017-12-28 | 2017-12-28 | A kind of modeling method, recognition methods, device, storage medium and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108364036A true CN108364036A (en) | 2018-08-03 |
Family
ID=63010481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711459765.1A Pending CN108364036A (en) | 2017-12-28 | 2017-12-28 | A kind of modeling method, recognition methods, device, storage medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108364036A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512692A (en) * | 2015-11-30 | 2016-04-20 | 华南理工大学 | BLSTM-based online handwritten mathematical expression symbol recognition method |
| CN106570456A (en) * | 2016-10-13 | 2017-04-19 | 华南理工大学 | Handwritten Chinese character recognition method based on full-convolution recursive network |
| CN106991374A (en) * | 2017-03-07 | 2017-07-28 | 中国矿业大学 | Handwritten Digit Recognition method based on convolutional neural networks and random forest |
| CN107220655A (en) * | 2016-03-22 | 2017-09-29 | 华南理工大学 | A kind of hand-written, printed text sorting technique based on deep learning |
| CN107239733A (en) * | 2017-04-19 | 2017-10-10 | 上海嵩恒网络科技有限公司 | Continuous hand-written character recognizing method and system |
-
2017
- 2017-12-28 CN CN201711459765.1A patent/CN108364036A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512692A (en) * | 2015-11-30 | 2016-04-20 | 华南理工大学 | BLSTM-based online handwritten mathematical expression symbol recognition method |
| CN107220655A (en) * | 2016-03-22 | 2017-09-29 | 华南理工大学 | A kind of hand-written, printed text sorting technique based on deep learning |
| CN106570456A (en) * | 2016-10-13 | 2017-04-19 | 华南理工大学 | Handwritten Chinese character recognition method based on full-convolution recursive network |
| CN106991374A (en) * | 2017-03-07 | 2017-07-28 | 中国矿业大学 | Handwritten Digit Recognition method based on convolutional neural networks and random forest |
| CN107239733A (en) * | 2017-04-19 | 2017-10-10 | 上海嵩恒网络科技有限公司 | Continuous hand-written character recognizing method and system |
Non-Patent Citations (2)
| Title |
|---|
| RONALDO MESSINA ET AL.: "Segmentation Handwritten Chinese Text Recognition with LSTM-RNN", 《2015 13TH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION(ICDAR)》 * |
| 金连文 等: "深度学习在手写汉字识别中的应用综述", 《自动化学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2737720C1 (en) | Retrieving fields using neural networks without using templates | |
| CN109543690B (en) | Method and device for extracting information | |
| CN108364037A (en) | Method, system and the equipment of Handwritten Chinese Character Recognition | |
| US10963685B2 (en) | Generating variations of a known shred | |
| Davis et al. | Text and style conditioned GAN for generation of offline handwriting lines | |
| US20090116755A1 (en) | Systems and methods for enabling manual classification of unrecognized documents to complete workflow for electronic jobs and to assist machine learning of a recognition system using automatically extracted features of unrecognized documents | |
| CN110619059B (en) | Building marking method based on transfer learning | |
| US20170076152A1 (en) | Determining a text string based on visual features of a shred | |
| CN108681735A (en) | Optical character recognition method based on convolutional neural networks deep learning model | |
| CN112508011A (en) | OCR (optical character recognition) method and device based on neural network | |
| CN110114776A (en) | Use the system and method for the character recognition of full convolutional neural networks | |
| Thompson et al. | finFindR: Automated recognition and identification of marine mammal dorsal fins using residual convolutional neural networks | |
| CN111767390A (en) | Skill word evaluation method and device, electronic equipment and computer readable medium | |
| CN110197140A (en) | Material checking method and equipment based on Text region | |
| CN104899551B (en) | A kind of form image sorting technique | |
| WO2022134580A1 (en) | Method and apparatus for acquiring certificate information, and storage medium and computer device | |
| US20230206676A1 (en) | Systems and Methods for Generating Document Numerical Representations | |
| CN108364036A (en) | A kind of modeling method, recognition methods, device, storage medium and equipment | |
| Brewer et al. | Reading PDFs using Adversarially trained Convolutional Neural Network based optical character recognition | |
| Bhagat et al. | Complex document classification and integration with indexing | |
| US20230409644A1 (en) | Systems and method for generating labelled datasets | |
| CN112966685B (en) | Attack network training method and device for scene text recognition and related equipment | |
| Thambi et al. | Offline text document authorization on the basis SIFT and SURF | |
| Grieggs et al. | The Paleographer's Eye ex machina: Using Computer Vision To Assist Humanists in Scribal Hand Identification | |
| Koushik et al. | Equation Detection in the Camera Captured Handwritten Document |
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: 20180803 |
|
| RJ01 | Rejection of invention patent application after publication |