CN112651323A - Chinese handwriting recognition method and system based on text line detection - Google Patents
Chinese handwriting recognition method and system based on text line detection Download PDFInfo
- Publication number
- CN112651323A CN112651323A CN202011526382.3A CN202011526382A CN112651323A CN 112651323 A CN112651323 A CN 112651323A CN 202011526382 A CN202011526382 A CN 202011526382A CN 112651323 A CN112651323 A CN 112651323A
- Authority
- CN
- China
- Prior art keywords
- image
- text
- character
- line
- neural network
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/30—Writer recognition; Reading and verifying signatures
- G06V40/33—Writer recognition; Reading and verifying signatures based only on signature image, e.g. static signature recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
-
- 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
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06V10/30—Noise filtering
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Biology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Human Computer Interaction (AREA)
- Character Input (AREA)
Abstract
The invention provides a Chinese handwriting recognition method and system based on text line detection, which comprises the following steps: segmenting an image to be identified into text lines; identifying the information of the inserted characters according to the average height of the text lines; extracting image features of the text lines by utilizing a pre-constructed multilayer convolutional neural network model; extracting character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model; and identifying the text content of the image to be identified according to the inserted character information and the character sequence characteristics. The invention can accurately identify the text content in the handwritten Chinese image and has high identification efficiency.
Description
Technical Field
The invention relates to the technical field of image recognition, in particular to a Chinese handwriting recognition method and system based on text line detection.
Background
Chinese recognition is an important issue in the field of computer vision and is one of the most challenging problems in the field of OCR. Chinese handwriting recognition and Chinese print recognition are the main research content of Chinese recognition. The printed matter recognition is developed to date, has higher recognition rate and is widely applied. Chinese handwriting recognition can be subdivided into offline handwriting recognition and online handwriting recognition. The on-line handwriting recognition needs to write on a specific electronic device, the electronic device senses stroke tracks, the number of strokes and the writing speed when a user writes Chinese characters, and the machine processes writing information captured by an instrument in real time to recognize the Chinese characters. While offline Chinese handwriting recognition can only acquire images of Chinese characters by using image acquisition equipment and recognize characters by analyzing the images.
Because Chinese characters have various categories, complicated structures and more characters with similar shapes, the writing randomness is large, the normalization is lacked, and the writing style difference of different people is large; and because the influence of the Chinese character structure and the writing are not standard, the Chinese character handwriting recognition method is difficult to segment for the adhered characters, the inserted characters and the like, so that the accuracy of the off-line Chinese handwriting character recognition can not be well applied to the industry.
In recent years, with the development of deep learning and the improvement of hardware computing performance, the computer vision field has achieved good results in a plurality of fields of pattern recognition, so that the use of deep learning technology has great significance for solving the recognition of Chinese handwriting.
Disclosure of Invention
In view of the above-mentioned deficiencies of the prior art, the present invention provides a method and system for Chinese handwriting recognition based on text line detection, so as to solve the above-mentioned technical problems.
In a first aspect, the present invention provides a method for Chinese handwriting recognition based on text line detection, comprising:
segmenting an image to be identified into text lines;
identifying the information of the inserted characters according to the average height of the text lines;
extracting image features of the text lines by utilizing a pre-constructed multilayer convolutional neural network model;
extracting character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model;
and identifying the text content of the image to be identified according to the inserted character information and the character sequence characteristics.
Further, the method further comprises:
removing Gaussian noise of the image to be identified in the scanning process by utilizing a Gaussian filtering technology;
and removing long straight lines appearing in the image to be recognized by adopting a straight line detection algorithm.
Further, the segmenting the image to be recognized into text lines includes:
segmenting the examinee answer image by a line unit through image transverse projection and a water drop algorithm;
longitudinally projecting and segmenting the writing inclined image of the line with the height exceeding the average height, and removing redundant blanks on the upper side and the lower side to obtain a text line;
all text lines are integrated in sequence into an image sequence.
Further, the identifying the position of the inserted character according to the average height of the text line comprises:
calculating the line average height of all text lines;
screening out target text lines with the height exceeding the line average height;
searching a region of the target text line, in which the inserted characters possibly exist, through text projection, and taking the region as a candidate region of interest;
extracting haar features of the candidate interesting regions, and clustering the candidate interesting regions according to the haar features of the candidate interesting regions;
inputting the clustering result into a support vector machine classifier, and screening out an interested region;
and identifying the inserted character and the position of the inserted character in the region of interest by utilizing a character recognition technology.
Further, the method for constructing the multilayer convolutional neural network model comprises the following steps:
setting the number of the convolution layers as six layers;
setting the convolution kernel size to 3 x 3;
selecting a maximum pooling layer;
adding batch normalization layers to the fourth convolution layer and the fifth convolution layer;
and collecting historical text lines, constructing a first training set by using the historical text lines, and training the multilayer convolutional neural network model by using the first training set.
Further, the construction method of the bidirectional recurrent neural network model comprises the following steps:
constructing a bidirectional cyclic neural network model;
collecting historical image features as a second training set;
introducing null characters to non-character positions in the historical image features of the second training set;
and training the bidirectional cyclic neural network model by utilizing a time sequence classification algorithm and the second training set.
Further, the recognizing the text content of the image to be recognized according to the inserted character information and the character sequence features includes:
according to the inserted character information and the character sequence characteristics, recognizing the content of the text line by utilizing a character recognition technology, and integrating the content of the text line into the text content of the image to be recognized;
and removing the scratch-out characters and the spaces in the text content, and outputting a final recognition result.
In a second aspect, the present invention provides a Chinese handwriting recognition system based on text line detection, comprising:
the image segmentation unit is configured to segment the image to be identified into text lines;
the symbol recognition unit is configured for recognizing the inserted character information according to the average height of the text line;
the characteristic identification unit is configured to extract the image characteristics of the text line by utilizing a pre-constructed multilayer convolutional neural network model;
the sequence recognition unit is configured to extract character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model;
and the content identification unit is configured to identify the text content of the image to be identified according to the inserted character information and the character sequence characteristics.
Further, the image segmentation unit includes:
the segmentation execution module is configured for segmenting the examinee answer image in a row unit through image transverse projection and a water drop algorithm;
the blank removing module is configured for longitudinally projecting and segmenting the writing inclined image of the line with the height exceeding the average height, and removing redundant blanks on the upper side and the lower side to obtain a text line;
and the sequence integration module is configured for sequentially integrating all text lines into an image sequence.
Further, the symbol recognition unit includes:
a height calculation module configured to calculate a line average height of all text lines;
the height comparison module is configured to screen out target text lines with heights exceeding the line average height;
the projection searching module is configured to search a region of the target text line, where the inserted character may exist, through text projection, and take the region as a candidate region of interest;
the characteristic clustering module is configured for extracting the haar characteristics of the candidate interesting regions and clustering the candidate interesting regions according to the haar characteristics of the candidate interesting regions;
the clustering screening module is configured for inputting the clustering result into the support vector machine classifier and screening out the region of interest;
and the insertion recognition module is configured for recognizing the insertion character and the insertion character position of the region of interest by utilizing a character recognition technology.
The beneficial effect of the invention is that,
the Chinese handwriting recognition method and system based on text line detection can accurately recognize text contents in handwritten Chinese images and have high recognition efficiency.
In addition, the invention has reliable design principle, simple structure and very wide application prospect.
Drawings
In order to more clearly illustrate the embodiments or technical solutions in the prior art of the present invention, the drawings used in the description of the embodiments or prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without creative efforts.
FIG. 1 is a schematic flow diagram of a method of one embodiment of the invention.
FIG. 2 is a schematic flow diagram of the insertion character recognition of a method of one embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention, and it is obvious that the described embodiment is only a part of the embodiment of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The following explains key terms appearing in the present invention.
Specifically, referring to fig. 1, the method for recognizing chinese handwriting based on text line detection includes:
step (1): and preprocessing the answer image of the examinee.
And preprocessing the answer image of the examinee, wherein Gaussian filtering is adopted in the preprocessing process to remove Gaussian noise of the image in the scanning process, and a straight line detection algorithm is adopted to remove long straight lines such as text boxes appearing in the partial image, so that the whole image I is obtained.
Step (2): and cutting the preprocessed image and removing upper and lower blank areas.
Segmenting the examinee answer image through image transverse projection and water drop algorithm to obtain a segmented image IiOf size hi×wi. Then, longitudinally projecting and cutting the writing inclined image with the height of the cut image far larger than the average height Th, removing redundant blanks on the upper side and the lower side, and then merging and cutting the images again to obtain a line cut image IjOf size hj×wiFinally obtaining nxhj×wiThe lines of (a) are cut into the image sequence. Finally, the n line segmentation image sequences are converted into n multiplied by h multiplied by w with fixed heightiImage sequence In。
And (3): as shown in fig. 2, the text line image suspected of having the insertion symbol is placed in an insertion character processing module for insertion symbol recognition and processing.
For line-slicing image sequences InSince the image height of the existing inserted character is much larger than the average height of the n line-segmented images, the method is used for detecting the inserted character of the line-segmented image suspected of having the inserted character. In order to reduce the number of the interested regions and combine with practical situations, through projection, a region where the inserted characters are possibly existed is found to be used as a candidate interested region, haar features are extracted from all the candidate regions, after clustering, an SVM classifier is input, and candidate region images are screened. If the insertion symbol exists, the image above or below the insertion symbol is transmitted to a character recognition module, and a recognition result is output.
And (4): and extracting image features by adopting a multilayer convolutional neural network.
The invention uses 6 layers of convolution neural networks, the convolution kernels are all small convolution kernels of 3 x 3, the largest pooling layer is selected after the convolution layer, and the BN layer is added after the pooling layers of the 4 th layer and the 5 th layer, thereby accelerating the training and convergence speed of the network and playing the role of preventing overfitting.
This step yields a 4-dimensional image feature data matrix X ═ X1,x2,...,xn]∈Rn×dAnd the width, the height, the depth and the like of the feature matrix are consistent.
And (5): and introducing the extracted image features into a bidirectional recurrent neural network to continuously extract character sequence features.
The circulating network layer in the invention is a two-layer bidirectional LSTM network, and character sequence characteristics are continuously extracted on the basis of convolution characteristics. And (4) transmitting the characteristic data matrix obtained in the step (4) into an LSTM network, and extracting the sequence characteristics of the characters.
And (6): and processing the recognition result into blank spaces, wrongly written characters, inserted characters and the like, and adjusting the corresponding sequence.
After the network recognition is finished, post-processing of the recognition result is carried out, namely, characters and spaces in the recognition result are removed and replaced, so that the semantics of the recognition result are not influenced.
The embodiment provides a Chinese handwriting recognition system based on text line detection, which comprises:
the image segmentation unit is configured to segment the image to be identified into text lines;
the symbol recognition unit is configured for recognizing the inserted character information according to the average height of the text line;
the characteristic identification unit is configured to extract the image characteristics of the text line by utilizing a pre-constructed multilayer convolutional neural network model;
the sequence recognition unit is configured to extract character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model;
and the content identification unit is configured to identify the text content of the image to be identified according to the inserted character information and the character sequence characteristics.
Optionally, as an embodiment of the present invention, the image segmentation unit includes:
the segmentation execution module is configured for segmenting the examinee answer image in a row unit through image transverse projection and a water drop algorithm;
the blank removing module is configured for longitudinally projecting and segmenting the writing inclined image of the line with the height exceeding the average height, and removing redundant blanks on the upper side and the lower side to obtain a text line;
and the sequence integration module is configured for sequentially integrating all text lines into an image sequence.
Optionally, as an embodiment of the present invention, the symbol recognition unit includes:
a height calculation module configured to calculate a line average height of all text lines;
the height comparison module is configured to screen out target text lines with heights exceeding the line average height;
the projection searching module is configured to search a region of the target text line, where the inserted character may exist, through text projection, and take the region as a candidate region of interest;
the characteristic clustering module is configured for extracting the haar characteristics of the candidate interesting regions and clustering the candidate interesting regions according to the haar characteristics of the candidate interesting regions;
the clustering screening module is configured for inputting the clustering result into the support vector machine classifier and screening out the region of interest;
and the insertion recognition module is configured for recognizing the insertion character and the insertion character position of the region of interest by utilizing a character recognition technology.
Although the present invention has been described in detail by referring to the drawings in connection with the preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made on the embodiments of the present invention by those skilled in the art without departing from the spirit and scope of the present invention, and these modifications or substitutions are within the scope of the present invention/any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A Chinese handwriting recognition method based on text line detection is characterized by comprising the following steps:
segmenting an image to be identified into text lines;
identifying the information of the inserted characters according to the average height of the text lines;
extracting image features of the text lines by utilizing a pre-constructed multilayer convolutional neural network model;
extracting character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model;
and identifying the text content of the image to be identified according to the inserted character information and the character sequence characteristics.
2. The method of claim 1, further comprising:
removing Gaussian noise of the image to be identified in the scanning process by utilizing a Gaussian filtering technology;
and removing long straight lines appearing in the image to be recognized by adopting a straight line detection algorithm.
3. The method according to claim 1, wherein the segmenting the image to be recognized into text lines comprises:
segmenting the examinee answer image by a line unit through image transverse projection and a water drop algorithm;
longitudinally projecting and segmenting the writing inclined image of the line with the height exceeding the average height, and removing redundant blanks on the upper side and the lower side to obtain a text line;
all text lines are integrated in sequence into an image sequence.
4. The method of claim 1, wherein identifying an inserted character position according to an average height of a text line comprises:
calculating the line average height of all text lines;
screening out target text lines with the height exceeding the line average height;
searching a region of the target text line, in which the inserted characters possibly exist, through text projection, and taking the region as a candidate region of interest;
extracting haar features of the candidate interesting regions, and clustering the candidate interesting regions according to the haar features of the candidate interesting regions;
inputting the clustering result into a support vector machine classifier, and screening out an interested region;
and identifying the inserted character and the position of the inserted character in the region of interest by utilizing a character recognition technology.
5. The method of claim 1, wherein the method for constructing the multilayer convolutional neural network model comprises:
setting the number of the convolution layers as six layers;
setting the convolution kernel size to 3 x 3;
selecting a maximum pooling layer;
adding batch normalization layers to the fourth convolution layer and the fifth convolution layer;
and collecting historical text lines, constructing a first training set by using the historical text lines, and training the multilayer convolutional neural network model by using the first training set.
6. The method of claim 1, wherein the method for constructing the bidirectional recurrent neural network model comprises:
constructing a bidirectional cyclic neural network model;
collecting historical image features as a second training set;
introducing null characters to non-character positions in the historical image features of the second training set;
and training the bidirectional cyclic neural network model by utilizing a time sequence classification algorithm and the second training set.
7. The method of claim 1, wherein the recognizing the text content of the image to be recognized according to the inserted character information and the character sequence features comprises:
according to the inserted character information and the character sequence characteristics, recognizing the content of the text line by utilizing a character recognition technology, and integrating the content of the text line into the text content of the image to be recognized;
and removing the scratch-out characters and the spaces in the text content, and outputting a final recognition result.
8. A chinese handwriting recognition system based on text line detection, comprising:
the image segmentation unit is configured to segment the image to be identified into text lines;
the symbol recognition unit is configured for recognizing the inserted character information according to the average height of the text line;
the characteristic identification unit is configured to extract the image characteristics of the text line by utilizing a pre-constructed multilayer convolutional neural network model;
the sequence recognition unit is configured to extract character sequence features from the image features by utilizing a pre-constructed bidirectional recurrent neural network model;
and the content identification unit is configured to identify the text content of the image to be identified according to the inserted character information and the character sequence characteristics.
9. The system of claim 8, wherein the image segmentation unit comprises:
the segmentation execution module is configured for segmenting the examinee answer image in a row unit through image transverse projection and a water drop algorithm;
the blank removing module is configured for longitudinally projecting and segmenting the writing inclined image of the line with the height exceeding the average height, and removing redundant blanks on the upper side and the lower side to obtain a text line;
and the sequence integration module is configured for sequentially integrating all text lines into an image sequence.
10. The system of claim 8, wherein the symbol recognition unit comprises:
a height calculation module configured to calculate a line average height of all text lines;
the height comparison module is configured to screen out target text lines with heights exceeding the line average height;
the projection searching module is configured to search a region of the target text line, where the inserted character may exist, through text projection, and take the region as a candidate region of interest;
the characteristic clustering module is configured for extracting the haar characteristics of the candidate interesting regions and clustering the candidate interesting regions according to the haar characteristics of the candidate interesting regions;
the clustering screening module is configured for inputting the clustering result into the support vector machine classifier and screening out the region of interest;
and the insertion recognition module is configured for recognizing the insertion character and the insertion character position of the region of interest by utilizing a character recognition technology.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011526382.3A CN112651323B (en) | 2020-12-22 | 2020-12-22 | Chinese handwriting recognition method and system based on text line detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011526382.3A CN112651323B (en) | 2020-12-22 | 2020-12-22 | Chinese handwriting recognition method and system based on text line detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112651323A true CN112651323A (en) | 2021-04-13 |
| CN112651323B CN112651323B (en) | 2022-12-13 |
Family
ID=75358958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011526382.3A Active CN112651323B (en) | 2020-12-22 | 2020-12-22 | Chinese handwriting recognition method and system based on text line detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112651323B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554012A (en) * | 2021-09-22 | 2021-10-26 | 江西博微新技术有限公司 | Primitive model classification method, system, equipment and storage medium in three-dimensional engineering |
| CN115862045A (en) * | 2023-02-16 | 2023-03-28 | 中国人民解放军总医院第一医学中心 | Case automatic identification method, system, equipment and storage medium based on image-text identification technology |
| CN117894033A (en) * | 2024-03-14 | 2024-04-16 | 山东山大鸥玛软件股份有限公司 | Method and system for checking consistency of answer sheets based on OCR (optical character recognition) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101452444A (en) * | 2007-12-04 | 2009-06-10 | 哈尔滨工业大学深圳研究生院 | Rapid editing and typesetting method for handwriting information and edition symbol identification method |
| CN108351745A (en) * | 2015-08-25 | 2018-07-31 | 迈思慧公司 | The system and method for digital notes record |
| CN108509881A (en) * | 2018-03-22 | 2018-09-07 | 五邑大学 | A kind of the Off-line Handwritten Chinese text recognition method of no cutting |
| CN110287952A (en) * | 2019-07-01 | 2019-09-27 | 中科软科技股份有限公司 | A kind of recognition methods and system for tieing up sonagram piece character |
| CN110298343A (en) * | 2019-07-02 | 2019-10-01 | 哈尔滨理工大学 | A kind of hand-written blackboard writing on the blackboard recognition methods |
| CN111008594A (en) * | 2019-12-04 | 2020-04-14 | 科大讯飞股份有限公司 | Error correction evaluation method, related equipment and readable storage medium |
| CN111091124A (en) * | 2019-12-04 | 2020-05-01 | 吉林大学 | Spine character recognition method |
| CN111222589A (en) * | 2018-11-27 | 2020-06-02 | 中国移动通信集团辽宁有限公司 | Image text recognition method, device, equipment and computer storage medium |
| CN111368632A (en) * | 2019-12-27 | 2020-07-03 | 上海眼控科技股份有限公司 | Signature identification method and device |
| CN111652332A (en) * | 2020-06-09 | 2020-09-11 | 山东大学 | Deep learning handwritten Chinese character recognition method and system based on two classifications |
| CN112686345A (en) * | 2020-12-31 | 2021-04-20 | 江南大学 | Off-line English handwriting recognition method based on attention mechanism |
| CN112990175A (en) * | 2021-04-01 | 2021-06-18 | 深圳思谋信息科技有限公司 | Method and device for recognizing handwritten Chinese characters, computer equipment and storage medium |
-
2020
- 2020-12-22 CN CN202011526382.3A patent/CN112651323B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101452444A (en) * | 2007-12-04 | 2009-06-10 | 哈尔滨工业大学深圳研究生院 | Rapid editing and typesetting method for handwriting information and edition symbol identification method |
| CN108351745A (en) * | 2015-08-25 | 2018-07-31 | 迈思慧公司 | The system and method for digital notes record |
| CN108509881A (en) * | 2018-03-22 | 2018-09-07 | 五邑大学 | A kind of the Off-line Handwritten Chinese text recognition method of no cutting |
| CN111222589A (en) * | 2018-11-27 | 2020-06-02 | 中国移动通信集团辽宁有限公司 | Image text recognition method, device, equipment and computer storage medium |
| CN110287952A (en) * | 2019-07-01 | 2019-09-27 | 中科软科技股份有限公司 | A kind of recognition methods and system for tieing up sonagram piece character |
| CN110298343A (en) * | 2019-07-02 | 2019-10-01 | 哈尔滨理工大学 | A kind of hand-written blackboard writing on the blackboard recognition methods |
| CN111008594A (en) * | 2019-12-04 | 2020-04-14 | 科大讯飞股份有限公司 | Error correction evaluation method, related equipment and readable storage medium |
| CN111091124A (en) * | 2019-12-04 | 2020-05-01 | 吉林大学 | Spine character recognition method |
| CN111368632A (en) * | 2019-12-27 | 2020-07-03 | 上海眼控科技股份有限公司 | Signature identification method and device |
| CN111652332A (en) * | 2020-06-09 | 2020-09-11 | 山东大学 | Deep learning handwritten Chinese character recognition method and system based on two classifications |
| CN112686345A (en) * | 2020-12-31 | 2021-04-20 | 江南大学 | Off-line English handwriting recognition method based on attention mechanism |
| CN112990175A (en) * | 2021-04-01 | 2021-06-18 | 深圳思谋信息科技有限公司 | Method and device for recognizing handwritten Chinese characters, computer equipment and storage medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554012A (en) * | 2021-09-22 | 2021-10-26 | 江西博微新技术有限公司 | Primitive model classification method, system, equipment and storage medium in three-dimensional engineering |
| CN115862045A (en) * | 2023-02-16 | 2023-03-28 | 中国人民解放军总医院第一医学中心 | Case automatic identification method, system, equipment and storage medium based on image-text identification technology |
| CN117894033A (en) * | 2024-03-14 | 2024-04-16 | 山东山大鸥玛软件股份有限公司 | Method and system for checking consistency of answer sheets based on OCR (optical character recognition) |
| CN117894033B (en) * | 2024-03-14 | 2024-05-28 | 山东山大鸥玛软件股份有限公司 | Method and system for checking consistency of answer sheets based on OCR (optical character recognition) |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112651323B (en) | 2022-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112651323B (en) | Chinese handwriting recognition method and system based on text line detection | |
| Mahmoud | Recognition of writer-independent off-line handwritten Arabic (Indian) numerals using hidden Markov models | |
| Singh et al. | Offline script identification from multilingual indic-script documents: a state-of-the-art | |
| Chaudhuri et al. | An approach for detecting and cleaning of struck-out handwritten text | |
| CN108664975B (en) | Uyghur handwritten letter recognition method and system and electronic equipment | |
| CN110458158B (en) | Text detection and identification method for assisting reading of blind people | |
| Roy et al. | Date-field retrieval in scene image and video frames using text enhancement and shape coding | |
| Al-Dmour et al. | Word extraction from Arabic handwritten documents based on statistical measures | |
| Verma et al. | A novel approach for structural feature extraction: contour vs. direction | |
| CN107103289B (en) | Method and system for handwriting identification by using handwriting outline characteristics | |
| Saabni | Efficient recognition of machine printed Arabic text using partial segmentation and Hausdorff distance | |
| Chacko et al. | Handwritten character recognition in malayalam scripts-a review | |
| Bains et al. | Dynamic features based stroke recognition system for signboard images of Gurmukhi text | |
| Nath et al. | Improving various offline techniques used for handwritten character recognition: a review | |
| Houtinezhad et al. | Off-line signature verification system using features linear mapping in the candidate points | |
| Aravinda et al. | South Indian character recognition using statistical feature extraction and distance classifier | |
| Zanwar et al. | A comprehensive survey on soft computing based optical character recognition techniques | |
| Zaghloul et al. | Recognition of Hindi (Arabic) handwritten numerals | |
| Jain | Unconstrained Arabic & Urdu text recognition using deep CNN-RNN hybrid networks | |
| AlKhateeb | Word-based handwritten Arabic scripts recognition using dynamic Bayesian network | |
| Rehman | Neural computing for online Arabic handwriting recognition using hard stroke features mining | |
| Rehman | Neural computing for online Arabic handwriting character recognition using hard stroke features mining | |
| CN116597458B (en) | Handwritten letter recognition method, system and application | |
| AlKhateeb et al. | Interactive knowledge discovery for baseline estimation and word segmentation in handwritten Arabic text | |
| Islam et al. | An enhanced MSER pruning algorithm for detection and localization of bangla texts from scene images. |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |