CN112149678A - Character recognition method and device for special language and recognition model training method and device - Google Patents
Character recognition method and device for special language and recognition model training method and device Download PDFInfo
- Publication number
- CN112149678A CN112149678A CN202010981048.0A CN202010981048A CN112149678A CN 112149678 A CN112149678 A CN 112149678A CN 202010981048 A CN202010981048 A CN 202010981048A CN 112149678 A CN112149678 A CN 112149678A
- Authority
- CN
- China
- Prior art keywords
- special language
- language character
- image
- target
- phrase
- 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
- 238000012549 training Methods 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 80
- 238000012545 processing Methods 0.000 claims description 22
- 238000010586 diagram Methods 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 12
- 230000002457 bidirectional effect Effects 0.000 claims description 8
- 238000003709 image segmentation Methods 0.000 claims description 8
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 7
- 238000012015 optical character recognition Methods 0.000 description 7
- 238000010801 machine learning Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 4
- 239000013598 vector Substances 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000011022 operating instruction Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
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
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Character Discrimination (AREA)
Abstract
The specification discloses a character recognition method and a recognition model training method and device for a special language. The training method of the character recognition model of the special language comprises the following steps: segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character; generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters; and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method and an apparatus for character recognition in a special language.
Background
Currently, Optical Character Recognition (OCR) is a process of examining a Character printed on paper by an electronic device (e.g., a scanner or a digital camera), determining its shape by detecting dark and light patterns, and then translating the shape into a computer text by a Character Recognition method. Specifically, OCR is generally to convert characters in a paper document into an image file of black and white dot matrix optically for print characters, and convert the characters in the image into a text format through recognition software for further editing and processing by word processing software.
However, since there is a lack of a database of characters in a particular language, such as a particular language character, in the market, how to apply OCR technology to recognition of a particular language character, such as a Burma character, still needs to provide a further solution.
Disclosure of Invention
The embodiment of the specification provides a character recognition method, a recognition model training method and a recognition model training device for a special language, and aims to solve the problem that in the prior art, an OCR technology is difficult to apply to recognition of the special language characters due to the lack of a character database of the special language characters, such as Burma characters and other small languages.
In order to solve the above technical problem, the embodiments of the present specification are implemented as follows:
in a first aspect, a method for training a special language character recognition model is provided, including:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
In a second aspect, a method for recognizing a special language word is provided, which includes:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
In a third aspect, a training apparatus for a special language character recognition model is provided, which includes:
the image segmentation module is used for segmenting the handwritten notes of the special language characters of the users based on an image segmentation algorithm to obtain a plurality of special language characters in the handwritten notes of the special language characters of the users;
the sample image acquisition module is used for acquiring a plurality of sample images based on the plurality of special language characters, and one sample image comprises a special language character phrase;
and the model training module is used for training based on the plurality of sample images to obtain a special language character recognition model by taking special language character phrases contained in the plurality of sample images as labels.
In a fourth aspect, a device for recognizing special language characters is provided, which includes:
the image acquisition module is used for acquiring a target image to be identified, wherein the target image comprises special language characters;
the character recognition module is used for taking the target image as the input of a special language character recognition model so as to output special language characters contained in the target image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
In a fifth aspect, an electronic device is provided, which includes:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
In a sixth aspect, a computer-readable storage medium is presented, storing one or more programs that, when executed by an electronic device including a plurality of application programs, cause the electronic device to:
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
In a seventh aspect, an electronic device is provided, which includes:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
In an eighth aspect, a computer-readable storage medium is presented, the computer-readable storage medium storing one or more programs that, when executed by an electronic device that includes a plurality of application programs, cause the electronic device to:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
The embodiment of the specification can achieve at least the following technical effects by adopting the technical scheme:
when a special language character recognition model is trained, the handwritten notes corresponding to a plurality of special language characters can be respectively segmented to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character; then, generating a target special language word phrase image corresponding to the target special language word phrase based on the respective handwritten image sets of the target special language word phrase and the plurality of special language characters; and then training the special language character recognition model by taking the target special language character phrase image as a sample and taking the target special language character phrase corresponding to the target special language character phrase image as a label, so that not only is a handwritten image set of the special language characters constructed, but also the special language character recognition model is obtained by training in a machine learning mode, and the purpose of optimizing the accuracy and efficiency of recognizing the special language characters is achieved.
When the special language characters are identified, acquiring target special language character phrase images to be identified; taking the target special language character phrase image as the input of a special language character recognition model so as to output the special language character phrase in the target special language character phrase image; the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the sample images contain special language character phrases, the labels corresponding to the sample images are the special language character phrases contained in the sample images, recognition of special language characters such as Burma characters is achieved, recognition of the special language characters such as Burma characters is achieved by the special language character recognition model obtained by machine learning training, and recognition efficiency and accuracy of the special language characters are greatly improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the specification and are incorporated in and constitute a part of this specification, illustrate embodiments of the specification and together with the description serve to explain the document and are not to limit the specification in any way. In the drawings:
fig. 1 is a schematic flow chart illustrating an implementation of a training method for a special language character recognition model according to an embodiment of the present disclosure;
fig. 2 is a schematic flow chart illustrating an implementation of image segmentation and target special language word phrase image generation in a training method for a special language word recognition model according to an embodiment of the present specification;
FIG. 3 is a schematic diagram of an implementation process of a training method for a special language character recognition model according to an embodiment of the present disclosure;
fig. 4 is a schematic flow chart illustrating an implementation of a method for recognizing a special language word according to an embodiment of the present specification;
FIG. 5 is a schematic structural diagram of a training apparatus for a special language-character recognition model according to an embodiment of the present disclosure;
fig. 6 is a schematic structural diagram of a device for recognizing a special language character according to an embodiment of the present disclosure;
fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure;
fig. 8 is a schematic structural diagram of another electronic device provided in an embodiment of the present specification.
Detailed Description
In order to make the purpose, technical solutions and advantages of this document more clear, the technical solutions of this specification will be clearly and completely described below with reference to specific embodiments of this specification and the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present disclosure, and not all embodiments. All other embodiments obtained by a person skilled in the art without making any inventive step based on the embodiments in this description belong to the protection scope of this document.
The technical solutions provided by the embodiments of the present description are described in detail below with reference to the accompanying drawings.
In order to solve the problem that an OCR technology is difficult to apply to recognition of special language characters in a language database lacking special language characters such as Burma characters and the like in the prior art, the embodiment of the specification provides a training method of a special language character recognition model, and when the method provided by the embodiment of the specification is adopted to train the special language recognition model, handwritten notes corresponding to a plurality of Burma characters can be respectively segmented to obtain respective handwritten image sets of the Burma characters, wherein the handwritten note corresponding to one Burma character comprises a plurality of handwritten characters of the Burma character; then, generating a target special language word phrase image corresponding to the target special language word phrase based on the respective handwritten image sets of the target special language word phrase and a plurality of Burma characters; and then training the special language recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label, so that not only is a handwritten image set of the special language characters constructed, but also the special language character recognition model is obtained by training in a machine learning mode, and the purpose of optimizing the accuracy and efficiency of recognizing the special language characters is achieved.
The execution subject of the training method for the special language character recognition model provided in the embodiment of the present disclosure may be, but is not limited to, a server, a personal computer, and the like, which can be configured to execute at least one of the terminals of the method provided in the embodiment of the present disclosure.
For convenience of description, the following description will be made of an embodiment of the method, taking an execution subject of the method as a server capable of executing the method as an example. It is understood that the implementation of the method by the server is merely an exemplary illustration and should not be construed as a limitation of the method.
Specifically, an implementation flow diagram of a training method for a special language character recognition model provided in one or more embodiments of the present specification is shown in fig. 1, and includes:
It should be understood that in order to train a specific language character recognition model, a specific language character database needs to be established first, and the specific language character database needs to be acquired by a handwriting method. As shown in fig. 2, a schematic diagram of a process for acquiring Burma numeric characters is provided in the embodiment of the present specification. In fig. 2, firstly, an image segmentation algorithm can be adopted for the digits of the Burma body handwritten by the user to extract a handwritten image set of the digits of a plurality of Burma bodies handwritten by the user; then, acquiring the numbers of the Myanmar handwritten by different users for multiple times; and then, segmenting the handwritten notes corresponding to the acquired digits of the Burma bodies by an image segmentation algorithm to obtain respective handwritten image sets of the digits of the Burma bodies.
The special language words may be language words including a lack of word database, such as Burma words, Laos words, and other small languages. Of course, it is not excluded that other language characters with perfect character database can be used as special language characters in the present specification, which is not limited in the present specification.
After obtaining the respective handwritten image sets of the multiple special language characters, several categories of special language character combinations are selected from the respective handwritten image sets of the multiple special language characters through an image processing algorithm to obtain multiple target special language character phrase images, and the special language character phrases contained in the multiple target special language character phrase images can be consistent or inconsistent. As shown in fig. 2, after obtaining the handwritten image sets of each of the digits of a plurality of burma bodies, the image processing algorithm may select a combination of the digits of several burma bodies from the handwritten image sets of each of the digits of the plurality of burma bodies to obtain a target burma text phrase image, in which the Burma body digit corresponding to the target burma text phrase image is "1.2.2019".
It should be understood that, in order to enrich the sample library for training the special language character recognition model, there may be a plurality of target special language character phrase images in the embodiments of the present specification, and the target special language character phrases contained in the plurality of target special language character phrase images may be the same or different. When the target special language character phrases contained in the target special language character phrase images are the same, the target special language characters contained in the target special language character phrase images have different handwriting patterns, and the target special language characters can be obtained by handwriting of different users.
And step 130, training the special language character recognition model by taking the target special language character phrase image as a sample and taking the target special language character phrase corresponding to the target special language character phrase image as a label.
It should be understood that the sample for training the special language character recognition model may be stored in a sample library in advance, a plurality of target special language character phrase images may be stored in the sample library, a mapping relationship exists between the plurality of target special language character phrase images and the target special language character phrases corresponding to the target special language character phrase images, and the target special language character phrase corresponding to each target special language character phrase image, that is, the target special language character phrase included in each target special language character phrase image may be used as a label when the special language character recognition model is trained.
Optionally, training the special language character recognition model by using the target special language character phrase image as a sample and using the target special language character phrase corresponding to the target special language character phrase image as a tag, including:
sequentially extracting the features of the target special language character phrase image through a convolutional neural network to obtain a feature map of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image;
identifying the characteristic sequence of the target special language word phrase image through a bidirectional long-short term memory network (LSTM) to obtain a special language character corresponding to the characteristic sequence of the target special language word phrase image;
and optimizing the special language character recognition model based on the special language characters, the target special language character phrases and the preset loss function corresponding to the characteristic sequence of the target special language character phrase image until the special language character recognition model is obtained through training.
Optionally, optimizing the special language character recognition model based on the special language characters, the target special language character phrases and the preset loss function corresponding to the feature sequence of the target special language character phrase image until the special language character recognition model is obtained by training, including:
combining special language characters corresponding to the characteristic sequence of the target special language character phrase image to obtain a special language character phrase corresponding to the target special language character phrase image;
and optimizing the special language character recognition model based on the special language character phrase corresponding to the target special language character phrase image, the target special language character phrase and a preset loss function until the special language character recognition model is obtained through training.
The preset loss function is used for rewarding or punishing the special language character recognition model based on the special language character phrase and the target special language character phrase corresponding to the target special language character phrase image recognized by the special language character recognition model, and when the difference between the special language character phrase and the target special language character phrase corresponding to the target special language character phrase image recognized by the special language character recognition model is small, the preset loss function rewards the special language character recognition model, for example, the set step length can be reduced; when the difference between the special language character phrase corresponding to the target special language character phrase image identified by the special language character identification model and the target special language character phrase is large, the preset loss function punishs the special language character identification model, for example, the set step length can be increased.
Optionally, after the special language characters corresponding to the feature sequence of the target special language word phrase image are obtained, because there are a plurality of feature sequences of the target special language word phrase image, there are often two or more feature sequences representing the same character, and there may also be some spaces between the feature sequences representing the special language word characters, therefore, in order to improve the recognition accuracy of the special language word recognition model, before the target special language word phrase image is output, the special language characters corresponding to the feature sequence of the target special language word phrase image may also be subjected to preset processing. Specifically, combining the special language characters corresponding to the feature sequence of the target special language word phrase image to obtain the special language word phrase corresponding to the target special language word phrase image includes:
combining special language characters corresponding to the characteristic sequence of the target special language word phrase image to obtain a special language word phrase to be processed corresponding to the target special language word phrase image;
presetting the special language character phrase to be processed corresponding to the target special language character phrase image to obtain the special language character phrase corresponding to the target special language character phrase image;
the preset processing at least comprises removing blank characters and continuous repeated characters in the special language word phrase to be processed.
Taking Burma characters as an example of a special language character and one sample in the Burma character recognition model training process as an example, the training process is explained in detail by combining the schematic diagram of the Burma character recognition model training process shown in FIG. 3. As shown in fig. 3, includes:
s1, inputting an image.
For example, a target Burma text phrase image may be entered, the size of which is 32 (high) x 100 (wide) x 3 (channel);
and S2, extracting the features of the image.
For example, the features of the target Burma text phrase image can be extracted from the target Burma text phrase image through a convolutional neural network;
and S3, obtaining a characteristic map of the image.
For example, a feature map of the target Burma text phrase image may be obtained based on the features of the target Burma text phrase image, the feature map having a size of 1 (height) × 25 (width) × 512 (channel);
and S4, obtaining a characteristic sequence of the image.
For example, converting the feature map of the target Burma text phrase image to obtain a feature sequence of the target Burma text phrase image, wherein the size of the feature sequence is 25 × 512, specifically, the number of feature vectors contained in the feature sequence is 25, and each feature vector has 512 dimensions;
s5, the feature sequence is identified using bi-directional LSTM.
For example, the feature sequence of the target Burma text phrase image can be identified by using bidirectional LSTM, and the identification results of 25 feature vectors in S4 are obtained by sequentially identifying;
and S6, obtaining the recognized single character.
For example, the characters corresponding to the feature sequence of the target Burma text phrase image can be obtained based on the recognition result of the bidirectional LSTM. It should be understood that in recognition, recognition is performed on a single character basis.
And S7, outputting the recognized word phrase.
For example, the characters corresponding to the feature sequence of the target Burma text phrase image are subjected to preset processing (including removal of blank characters and removal of consecutive repeated characters), and combined to obtain the target Burma text phrase "1.2.2019" corresponding to the target Burma text phrase image.
When a special language character recognition model is trained, the handwritten notes corresponding to a plurality of special language characters can be respectively segmented to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character; then, generating a target special language word phrase image corresponding to the target special language word phrase based on the respective handwritten image sets of the target special language word phrase and the plurality of special language characters; and then training the special language character recognition model by taking the target special language character phrase image as a sample and taking the target special language character phrase corresponding to the target special language character phrase image as a label, so that not only is a handwritten image set of the special language characters constructed, but also the special language character recognition model is obtained by training in a machine learning mode, and the purpose of optimizing the accuracy and efficiency of recognizing the special language characters is achieved.
In order to solve the problem that an OCR technology is difficult to be applied to recognition of special language characters in a Chinese database which is lack of special language characters such as Burma characters and the like in the prior art, the embodiment of the specification further provides a recognition method of the special language characters, and when the special language characters are recognized by adopting the method provided by the embodiment of the specification, target special language character phrase images to be recognized are obtained; taking the target special language character phrase image as the input of a special language character recognition model so as to output the special language character phrase in the target special language character phrase image; the special language character recognition model is obtained based on training of a plurality of sample images and corresponding labels, the sample images contain special language character phrases, the labels corresponding to the sample images are the special language character phrases contained in the sample images, recognition of special language characters such as Burmese characters is achieved, the special language character recognition model obtained through machine learning training is used for recognizing the small language characters such as the special language characters, and recognition efficiency and accuracy of the special language characters are greatly improved.
The execution subject of the method for recognizing the special language characters provided by the embodiment of the present disclosure may be, but is not limited to, a server, a personal computer, a terminal device, and the like, which can be configured to execute at least one of the terminals of the method provided by the embodiment of the present disclosure.
For convenience of description, the following description will be made of an embodiment of the method taking as an example that an execution subject of the method is a terminal device capable of executing the method. It is understood that the implementation of the method by the terminal device is only an exemplary illustration, and should not be construed as a limitation of the method.
Specifically, an implementation flow diagram of a method for recognizing a special language word provided in one or more embodiments of the present specification is shown in fig. 4, and includes:
The special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
Optionally, the inputting the target special language character phrase image as an input of a special language character recognition model to output the special language character phrase in the target special language character phrase image includes:
extracting the characteristics of the target special language character phrase image through a convolutional neural network in a special language character recognition model, and respectively obtaining a characteristic diagram of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image through a special language character recognition model;
identifying the characteristic sequence of the target special language character phrase image through a bidirectional long-short term memory network (LSTM) in a special language character identification model to obtain a special language character corresponding to the characteristic sequence of the target special language character phrase image;
and processing the special language characters corresponding to the characteristic sequence of the target special language character phrase image through a special language character recognition model, and outputting to obtain the special language character phrases contained in the target special language character phrase image.
Optionally, the processing, by using the special language character recognition model, the special language characters corresponding to the feature sequence of the target special language character phrase image, and outputting the special language character phrase included in the target special language character phrase image includes:
processing the special language characters corresponding to the characteristic sequence of the target special language character phrase image through a special language character recognition model so as to remove blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language character phrase image;
and combining to obtain the special language characters contained in the target special language character phrase image based on the special language characters after blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language character phrase image are removed.
The specific implementation of the relevant steps in the embodiment shown in fig. 4 may refer to the specific implementation of the corresponding steps in the embodiments shown in fig. 1 to fig. 3, and one or more embodiments in this specification are not described herein again.
When the special language characters are identified, acquiring target special language character phrase images to be identified; taking the target special language character phrase image as the input of a special language character recognition model so as to output the special language character phrase in the target special language character phrase image; the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the sample images contain special language character phrases, the labels corresponding to the sample images are the special language character phrases contained in the sample images, recognition of special language characters such as Burma characters is achieved, recognition of the special language characters such as Burma characters is achieved by the special language character recognition model obtained by machine learning training, and recognition efficiency and accuracy of the special language characters are greatly improved.
Fig. 5 is a schematic structural diagram of a training apparatus 500 for a special language character recognition model provided in the present specification. Referring to fig. 5, in a software implementation, the training apparatus 500 for the special language character recognition model may include an image segmentation module 501, an image acquisition module 502 and a model training module 503, wherein:
an image segmentation module 501, configured to segment handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the plurality of special language characters, where a handwritten note corresponding to a special language character includes a plurality of handwritten characters of the special language character;
an image obtaining module 502, configured to generate a target special language word phrase image corresponding to a target special language word phrase based on the target special language word phrase and a respective handwritten image set of the plurality of special language characters;
the model training module 503 takes the target special language character phrase image as a sample, and takes the target special language character phrase corresponding to the target special language character phrase image as a label to train a special language character recognition model.
Optionally, in an embodiment, the model training module 503 is configured to:
sequentially extracting the features of the target special language character phrase image through a convolutional neural network to obtain a feature map of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image;
identifying the characteristic sequence of the target special language word phrase image through a bidirectional long-short term memory network (LSTM) to obtain a special language character corresponding to the characteristic sequence of the target special language word phrase image;
and optimizing the special language character recognition model based on the special language characters corresponding to the characteristic sequence of the target special language character phrase image, the target special language character phrase and a preset loss function until the special language character recognition model is obtained through training.
Optionally, in an embodiment, the model training module 503 is configured to:
combining special language characters corresponding to the characteristic sequence of the target special language word phrase image to obtain a special language word phrase corresponding to the target special language word phrase image;
and optimizing the special language character recognition model based on the special language character phrase corresponding to the target special language character phrase image, the target special language character phrase and a preset loss function until the special language character recognition model is obtained through training.
Optionally, in an embodiment, the model training module 503 is configured to:
combining special language characters corresponding to the characteristic sequence of the target special language word phrase image to obtain a special language word phrase to be processed corresponding to the target special language word phrase image;
presetting the special language character phrase to be processed corresponding to the target special language character phrase image to obtain the special language character phrase corresponding to the target special language character phrase image;
the preset processing at least comprises removing blank characters and continuous repeated characters in the special language word phrase to be processed.
The training apparatus 500 for the special language character recognition model can implement the method of the embodiment of the method shown in fig. 1 to 3, and specifically refer to the training method for the special language character recognition model of the embodiment shown in fig. 1 to 3, which is not described again.
Fig. 6 is a schematic structural diagram of a device 600 for recognizing a specific language character provided in the present specification. Referring to fig. 6, in a software implementation, the apparatus 600 for recognizing a specific language character may include an image obtaining module 601 and a character recognizing module 602, wherein:
the image acquisition module 601 is used for acquiring a target special language character phrase image to be identified;
a character recognition module 602, which takes the target special language character phrase image as an input of a special language character recognition model to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
Optionally, in an embodiment, the text recognition module 602 is configured to:
extracting the characteristics of the target special language character phrase image through a convolutional neural network in the special language character recognition model to respectively obtain a characteristic diagram of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image through the special language character recognition model;
identifying the characteristic sequence of the target special language character phrase image through a bidirectional long-short term memory network (LSTM) in the special language character identification model to obtain a special language character corresponding to the characteristic sequence of the target special language character phrase image;
and processing the special language characters corresponding to the characteristic sequence of the target special language character phrase image through the special language character recognition model, and outputting to obtain the special language character phrases contained in the target special language character phrase image.
Optionally, in an embodiment, the text recognition module 602 is configured to:
processing the special language characters corresponding to the characteristic sequence of the target special language word phrase image through the special language word recognition model so as to remove blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language word phrase image;
and combining to obtain the special language characters contained in the target special language character phrase image based on the special language characters after blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language character phrase image are removed.
The special language character recognition apparatus 600 can implement the method of the embodiment of the method shown in fig. 4, and specifically refer to the special language character recognition method shown in the embodiment shown in fig. 4, which is not described again.
Fig. 7 is a schematic structural diagram of an electronic device provided in an embodiment of the present specification. Referring to fig. 7, at a hardware level, the electronic device includes a processor, and optionally further includes an internal bus, a network interface, and a memory. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may further include a non-volatile Memory, such as at least 1 disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, the network interface, and the memory may be connected to each other via an internal bus, which may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 7, but this does not indicate only one bus or one type of bus.
And the memory is used for storing programs. In particular, the program may include program code comprising computer operating instructions. The memory may include both memory and non-volatile storage and provides instructions and data to the processor.
The processor reads the corresponding computer program from the nonvolatile memory to the memory and then runs the computer program to form the training device of the special language character recognition model on the logic level. The processor is used for executing the program stored in the memory and is specifically used for executing the following operations:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
The method for training the special language character recognition model disclosed in the embodiment of fig. 1 in this specification can be applied to or implemented by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components. The various methods, steps and logic blocks disclosed in one or more embodiments of the present specification may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with one or more embodiments of the present disclosure may be embodied directly in hardware, in a software module executed by a hardware decoding processor, or in a combination of the hardware and software modules executed by a hardware decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The electronic device may further perform the training method of the special language-character recognition model shown in fig. 1, which is not described herein again.
Of course, besides the software implementation, the electronic device in this specification does not exclude other implementations, such as logic devices or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or logic devices.
Embodiments of the present specification also provide a computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a portable electronic device comprising a plurality of application programs, enable the portable electronic device to perform the method of the embodiment shown in fig. 4, and in particular to perform the following operations:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
Fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present specification. Referring to fig. 8, at a hardware level, the electronic device includes a processor, and optionally further includes an internal bus, a network interface, and a memory. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may further include a non-volatile Memory, such as at least 1 disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, the network interface, and the memory may be connected to each other via an internal bus, which may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 8, but that does not indicate only one bus or one type of bus.
And the memory is used for storing programs. In particular, the program may include program code comprising computer operating instructions. The memory may include both memory and non-volatile storage and provides instructions and data to the processor.
The processor reads the corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to form the recognition device of the special language characters on the logic level. The processor is used for executing the program stored in the memory and is specifically used for executing the following operations:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
The method for recognizing the special language words disclosed in the embodiment shown in fig. 4 of the present specification can be applied to a processor, or can be implemented by the processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components. The various methods, steps and logic blocks disclosed in one or more embodiments of the present specification may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with one or more embodiments of the present disclosure may be embodied directly in hardware, in a software module executed by a hardware decoding processor, or in a combination of the hardware and software modules executed by a hardware decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The electronic device may also execute the method for recognizing the special language text shown in fig. 4, which is not described herein again.
Of course, besides the software implementation, the electronic device in this specification does not exclude other implementations, such as logic devices or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or logic devices.
Embodiments of the present specification also provide a computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a portable electronic device comprising a plurality of application programs, enable the portable electronic device to perform the method of the embodiment shown in fig. 4, and in particular to perform the following operations:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
In short, the above description is only a preferred embodiment of the present disclosure, and is not intended to limit the scope of the present disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of the present disclosure should be included in the scope of protection of one or more embodiments of the present disclosure.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
Claims (13)
1. A training method of a special language character recognition model comprises the following steps:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
2. The method of claim 1, wherein training a specific language character recognition model by using the target specific language character phrase image as a sample and the target specific language character phrase corresponding to the target specific language character phrase image as a tag comprises:
sequentially extracting the features of the target special language character phrase image through a convolutional neural network to obtain a feature map of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image;
identifying the characteristic sequence of the target special language word phrase image through a bidirectional long-short term memory network (LSTM) to obtain a special language character corresponding to the characteristic sequence of the target special language word phrase image;
and optimizing the special language character recognition model based on the special language characters corresponding to the characteristic sequence of the target special language character phrase image, the target special language character phrase and a preset loss function until the special language character recognition model is obtained through training.
3. The method of claim 2, wherein optimizing the recognition model based on the special language characters corresponding to the feature sequence of the image of the target special language text phrase, and a predetermined loss function until the recognition model is trained comprises:
combining special language characters corresponding to the characteristic sequence of the target special language word phrase image to obtain a special language word phrase corresponding to the target special language word phrase image;
and optimizing the special language character recognition model based on the special language character phrase corresponding to the target special language character phrase image, the target special language character phrase and a preset loss function until the special language character recognition model is obtained through training.
4. The method of claim 3, combining the special language characters corresponding to the feature sequence of the target special language word phrase image to obtain the special language word phrase corresponding to the target special language word phrase image, comprising:
combining special language characters corresponding to the characteristic sequence of the target special language word phrase image to obtain a special language word phrase to be processed corresponding to the target special language word phrase image;
presetting the special language character phrase to be processed corresponding to the target special language character phrase image to obtain the special language character phrase corresponding to the target special language character phrase image;
the preset processing at least comprises removing blank characters and continuous repeated characters in the special language word phrase to be processed.
5. A method for recognizing special language characters comprises the following steps:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
6. The method of claim 5, wherein inputting the target image as a special language character recognition model to output a special language character contained in the target image comprises:
extracting the characteristics of the target special language character phrase image through a convolutional neural network in the special language character recognition model to respectively obtain a characteristic diagram of the target special language character phrase image;
converting the characteristic graph of the target special language character phrase image into a characteristic sequence of the target special language character phrase image through the special language character recognition model;
identifying the characteristic sequence of the target special language character phrase image through a bidirectional long-short term memory network (LSTM) in the special language character identification model to obtain a special language character corresponding to the characteristic sequence of the target special language character phrase image;
and processing the special language characters corresponding to the characteristic sequence of the target special language character phrase image through the special language character recognition model, and outputting to obtain the special language character phrases contained in the target special language character phrase image.
7. The method according to claim 6, wherein the processing the special language characters corresponding to the feature sequence of the target special language word phrase image by the special language word recognition model, and outputting the special language word phrase included in the target special language word phrase image comprises:
processing the special language characters corresponding to the characteristic sequence of the target special language word phrase image through the special language word recognition model so as to remove blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language word phrase image;
and combining to obtain the special language characters contained in the target special language character phrase image based on the special language characters after blank characters and continuous repeated characters in the special language characters corresponding to the characteristic sequence of the target special language character phrase image are removed.
8. A training device for a special language character recognition model comprises:
the image segmentation module is used for segmenting the handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
the image acquisition module is used for generating a target special language word phrase image corresponding to a target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the special language characters;
and the model training module is used for training a special language character recognition model by taking the target special language character phrase image as a sample and taking the target special language character phrase corresponding to the target special language character phrase image as a label.
9. A device for recognizing a specific language character, comprising:
the image acquisition module is used for acquiring a target special language character phrase image to be identified;
the character recognition module is used for taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
10. An electronic device, comprising:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
11. A computer-readable storage medium storing one or more programs that, when executed by an electronic device including a plurality of application programs, cause the electronic device to:
segmenting handwritten notes corresponding to a plurality of special language characters respectively to obtain respective handwritten image sets of the special language characters, wherein the handwritten note corresponding to one special language character comprises a plurality of handwritten characters of the special language character;
generating a target special language word phrase image corresponding to the target special language word phrase based on the target special language word phrase and the respective handwritten image sets of the plurality of special language characters;
and training a special language character recognition model by taking the target special language character phrase image as a sample and the target special language character phrase corresponding to the target special language character phrase image as a label.
12. An electronic device, comprising:
a processor; and
a memory arranged to store computer executable instructions that, when executed, cause the processor to:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
13. A computer-readable storage medium storing one or more programs that, when executed by an electronic device including a plurality of application programs, cause the electronic device to:
acquiring a target special language word phrase image to be identified;
taking the target special language character phrase image as the input of a special language character recognition model so as to output a special language character phrase in the target special language character phrase image;
the special language character recognition model is obtained by training based on a plurality of sample images and corresponding labels, the plurality of sample images comprise special language character phrases, and the labels corresponding to the plurality of sample images are the special language character phrases contained in the plurality of sample images.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010981048.0A CN112149678A (en) | 2020-09-17 | 2020-09-17 | Character recognition method and device for special language and recognition model training method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010981048.0A CN112149678A (en) | 2020-09-17 | 2020-09-17 | Character recognition method and device for special language and recognition model training method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112149678A true CN112149678A (en) | 2020-12-29 |
Family
ID=73893920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010981048.0A Pending CN112149678A (en) | 2020-09-17 | 2020-09-17 | Character recognition method and device for special language and recognition model training method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112149678A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113420766A (en) * | 2021-07-05 | 2021-09-21 | 北京理工大学 | Low-resource language OCR method fusing language information |
| CN114926847A (en) * | 2021-12-06 | 2022-08-19 | 百度在线网络技术（北京）有限公司 | Image processing method, device, equipment and storage medium for minority language |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160328366A1 (en) * | 2015-05-04 | 2016-11-10 | King Fahd University Of Petroleum And Minerals | Systems and associated methods for arabic handwriting synthesis and dataset design |
| US20180033147A1 (en) * | 2016-07-26 | 2018-02-01 | Intuit Inc. | Label and field identification without optical character recognition (ocr) |
| CN109102037A (en) * | 2018-06-04 | 2018-12-28 | 平安科技（深圳）有限公司 | Chinese model training, Chinese image-recognizing method, device, equipment and medium |
| CN109858488A (en) * | 2018-12-28 | 2019-06-07 | 众安信息技术服务有限公司 | A kind of handwriting samples recognition methods and system based on sample enhancing |
-
2020
- 2020-09-17 CN CN202010981048.0A patent/CN112149678A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160328366A1 (en) * | 2015-05-04 | 2016-11-10 | King Fahd University Of Petroleum And Minerals | Systems and associated methods for arabic handwriting synthesis and dataset design |
| US20180033147A1 (en) * | 2016-07-26 | 2018-02-01 | Intuit Inc. | Label and field identification without optical character recognition (ocr) |
| CN109102037A (en) * | 2018-06-04 | 2018-12-28 | 平安科技（深圳）有限公司 | Chinese model training, Chinese image-recognizing method, device, equipment and medium |
| CN109858488A (en) * | 2018-12-28 | 2019-06-07 | 众安信息技术服务有限公司 | A kind of handwriting samples recognition methods and system based on sample enhancing |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113420766A (en) * | 2021-07-05 | 2021-09-21 | 北京理工大学 | Low-resource language OCR method fusing language information |
| CN113420766B (en) * | 2021-07-05 | 2022-09-16 | 北京理工大学 | Low-resource language OCR method fusing language information |
| CN114926847A (en) * | 2021-12-06 | 2022-08-19 | 百度在线网络技术（北京）有限公司 | Image processing method, device, equipment and storage medium for minority language |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110135411B (en) | Business card recognition method and device | |
| CN110046529B (en) | Two-dimensional code identification method, device and equipment | |
| CN110175609B (en) | Interface element detection method, device and equipment | |
| CN112070649B (en) | Method and system for removing specific character string watermark | |
| CN107341143B (en) | Sentence continuity judgment method and device and electronic equipment | |
| EP3881234A1 (en) | Automatically predicting text in images | |
| CN113963147B (en) | Key information extraction method and system based on semantic segmentation | |
| CN112861842A (en) | Case text recognition method based on OCR and electronic equipment | |
| CN112149678A (en) | Character recognition method and device for special language and recognition model training method and device | |
| CN112464845A (en) | Bill recognition method, equipment and computer storage medium | |
| CN112308046A (en) | Method, device, server and readable storage medium for positioning text region of image | |
| CN112149680A (en) | Wrong word detection and identification method and device, electronic equipment and storage medium | |
| CN110991303A (en) | Method and device for positioning text in image and electronic equipment | |
| CN114860905A (en) | Intention identification method, device and equipment | |
| Wicht et al. | Camera-based sudoku recognition with deep belief network | |
| CN111507250B (en) | Image recognition method, device and storage medium | |
| CN113129298A (en) | Definition recognition method of text image | |
| CN111339910A (en) | Text processing method and device and text classification model training method and device | |
| CN110825874A (en) | Chinese text classification method and device and computer readable storage medium | |
| Manzoor et al. | A novel system for image text recognition and classification using deep learning | |
| CN113221718B (en) | Formula identification method, device, storage medium and electronic equipment | |
| CN114299509A (en) | Method, device, equipment and medium for acquiring information | |
| CN113762160A (en) | Date extraction method and device, computer equipment and storage medium | |
| CN113343663A (en) | Bill structuring method and device | |
| CN112287130A (en) | Searching method, device and equipment for graphic questions |
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 |