CN104915648B - document/document direction detection method and detection device - Google Patents
document/document direction detection method and detection device Download PDFInfo
- Publication number
- CN104915648B CN104915648B CN201510296307.5A CN201510296307A CN104915648B CN 104915648 B CN104915648 B CN 104915648B CN 201510296307 A CN201510296307 A CN 201510296307A CN 104915648 B CN104915648 B CN 104915648B
- Authority
- CN
- China
- Prior art keywords
- chinese character
- image
- document
- character image
- printed
- 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.)
- Active - Reinstated
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
- G06V30/333—Preprocessing; Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Input (AREA)
Abstract
The present invention relates to a kind of document/document direction detection method and detection devices.Wherein, document/document direction detection method includes:Pending image is pre-processed, vertical/horizontal image is generated;To vertical/horizontal image binaryzation, piecemeal, multiple block images are formed;The printed Chinese character image of predetermined quantity is extracted in multiple block images;0 degree of printed Chinese character image rotation that each is extracted, 90 degree, 180 degree, 270 degree, form the first printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;Judgement is identified to every group of first printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;Judge the direction of document/document.Using document of the present invention/document direction detection method and detection device, can convenient, efficient automatic identification document/document digitized video direction, and recognition accuracy is high.
Description
Technical field
The present invention relates to a kind of document/document angle detecting technology, especially a kind of document/document direction detection method and
Detection device.
Background technology
If the direction of existing document/document digitized video storage and word correct direction are not inconsistent, need manually dry
In advance, become a full member.
Not only efficiency is low for the method for manual intervention, but also is easy error, to influence the treatment effeciency of subsequent processing.
Invention content
The brief overview about the present invention is given below, in order to provide the basic reason about certain aspects of the invention
Solution.It should be appreciated that this general introduction is not the exhaustive general introduction about the present invention.It is not intended to determine the key of the present invention
Or pith, nor is it intended to limit the scope of the present invention.Its purpose only provides certain concepts in simplified form, with
This is as the preamble in greater detail discussed later.
A primary object of the present invention is to provide a kind of document/document direction detection method and detection device, can be with
The correct direction of automatic decision digitized video, and become a full member.
According to an aspect of the present invention, a kind of document/document direction detection method, including:
Pending image is pre-processed, vertical/horizontal image is generated;
To the vertical/horizontal image binaryzation, piecemeal, multiple block images are formed;
The printed Chinese character image of predetermined quantity is extracted in multiple block images;
0 degree of printed Chinese character image rotation that each is extracted, 90 degree, 180 degree, 270 degree, form the first block letter
Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;
Respectively to the first printed Chinese character image of printed Chinese character image described in each, the second printed Chinese character figure
Judgement is identified in picture, third printed Chinese character image and the 4th printed Chinese character image, obtains recognition result;
The direction of the document/document is judged based on the recognition result.
According to the second aspect of the invention, a kind of document/document direction detection device, including:
Preprocessing module generates vertical/horizontal image for being pre-processed to pending image;
Piecemeal module, for the vertical/horizontal image binaryzation, piecemeal, forming multiple block images;
Extraction module, the printed Chinese character image for extracting predetermined quantity in multiple block images;
Rotary module, 0 degree of printed Chinese character image rotation for being extracted to each, 90 degree, 180 degree, 270 degree,
Form the first printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character
Image;
Chinese character walking direction module, for respectively to the first printed Chinese character figure of printed Chinese character image described in each
Judgement is identified in picture, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image, obtains
Recognition result;
Document/document walking direction module, the direction for judging the document/document based on the recognition result.
Document using the present invention/document direction detection method and detection device, can convenient, efficient automatic identification number
The direction of word image, and recognition accuracy is high.
Description of the drawings
Below with reference to the accompanying drawings illustrate embodiments of the invention, the above of the present invention and its can be more readily understood that
Its objects, features and advantages.Component in attached drawing is intended merely to show the principle of the present invention.In the accompanying drawings, identical or similar
Technical characteristic or component will be indicated using same or similar reference numeral.
Fig. 1 is a kind of flow chart of embodiment of document/document direction detection method of the present invention;
Fig. 2 is a kind of structure chart of embodiment of document/document direction detection device of the present invention.
Specific implementation mode
Illustrate the embodiment of the present invention with reference to the accompanying drawings.It is retouched in the attached drawing of the present invention or a kind of embodiment
The elements and features stated can be combined with elements and features shown in one or more other attached drawings or embodiment.It answers
When note that for purposes of clarity, being omitted known to unrelated to the invention, those of ordinary skill in the art in attached drawing and explanation
Component and processing expression and description.
It is shown in Figure 1, for a kind of flow chart of embodiment of document/document direction detection method of the present invention.
In the present embodiment, document/document direction detection method includes:
S10:Pending image is pre-processed, vertical/horizontal image is generated;
S20:To vertical/horizontal image binaryzation, piecemeal, multiple block images are formed;
S30:The printed Chinese character image of predetermined quantity is extracted in multiple block images;
S40:0 degree of printed Chinese character image rotation that each is extracted, 90 degree, 180 degree, 270 degree, form the first print
Brush body Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;
S50:Respectively to the first printed Chinese character image, the second printed Chinese character of printed Chinese character image described in each
Judgement is identified in image, third printed Chinese character image and the 4th printed Chinese character image, obtains recognition result;
S60:The direction of the document/document is judged based on the recognition result.
By above step, the correct direction of document can be automatically obtained, consequently facilitating subsequent processing is smooth, high
Effect is implemented.
In one embodiment, step S10 can be specifically included:
S11:Seek the angle of inclination of literal line in pending image;
S12:It rotates pending image and generates vertical/horizontal image so that the angle of inclination of literal line is 0 degree or 90 degree.
The vertical/horizontal image generated after the processing for completing step S11 and step S12, the vertical angle of literal line
Degree or level angle for example can be within ± 5 ° of deviations.
In one embodiment, step S30 can be specifically included:
S31:In each block image, the connected region of predetermined quantity is sought.For example, labelling method may be used to ask
Take connected region.
S32:Judge whether each connected region is printed character.
S33:If printed character, then further judge whether each printed character is printed Chinese character, if
Printed Chinese character then saves as printed Chinese character image.
When printed character extracts, symmetry, effective content ratio, stroke width and backbone length pair can be first passed through
Image in connected region is screened.If content is symmetrical above and below in image, symmetrical, subsequent conversion be 90 degree, 180 degree,
When 270 degree, character and artwork are close, are unfavorable for walking direction, therefore in the connected region for seeking predetermined number, abandon this
Class connected region image.
In addition, when the effective content of image is excessive, i.e. the effective content pixel point of character stroke, in character Outside Dimensions frame
When proportion is excessively high, it is meant that charactor comparison is complicated, if it is Chinese character, stroke may be caused unintelligible due to image zooming-out,
At this point, recognition result accuracy rate is relatively low, it is easy to influence follow-up ballot decision-making mechanism, also abandons such connected region image.
Further, it for connected region image is not belonging to the effective character of two kinds of situations described above, answers
It is less than character size certain proportion when meeting stroke width, backbone length can not be long.If conditions are not met, may be non-character
Region or impurity interference, also abandon such connected region image.
When whether judge each printed character is printed Chinese character, in one embodiment, can by from
The supporting vector machine model of line training determines whether character, and the feature that training when takes has 0 degree of a character, 90 degree, 45 degree,
Ratio shared by 135 degree of direction gradients and 8 kinds of basic strokes (point, horizontal, vertical, left, flick, folding, curved, hook).
In one embodiment, step S40 can be specifically included:
S41:The first printed Chinese character image to each printed Chinese character image, the second printed Chinese character figure respectively
Judgement is identified in picture, third printed Chinese character image and the 4th printed Chinese character image, respectively obtains each block letter
First recognition result of Chinese character image and the first confidence level, each second printed Chinese character image the second recognition result and
Second confidence level, the third recognition result of each third printed Chinese character image and third confidence level, each described 4th
4th recognition result of printed Chinese character image and the 4th confidence level;
S42:First confidence level, the second confidence level, the third for being respectively compared each printed Chinese character image are set
The numerical values recited of reliability and the 4th confidence level;
S43:In maximum first confidence level of the numerical value of each printed Chinese character image, the second confidence level,
It is counted on the first accumulator, the second accumulator, third accumulator or the 4th accumulator corresponding to three confidence levels or the 4th confidence level
Number plus one.
In one embodiment, step S50 can be specifically included:
S51:Compare the numerical values recited of the first accumulator, the second accumulator, third accumulator and the 4th accumulator;
S52:With corresponding with numerical value the maximum in the first accumulator, the second accumulator, third accumulator and the 4th accumulator
The first printed Chinese character image, the second printed Chinese character image, third printed Chinese character image or the 4th printed Chinese character figure
Direction of the direction of picture as the document/document.
In other words, judge the direction of each Chinese character, and vote and obtain the direction of document.
In one embodiment, can also include before step S10:
Load document image;
Document image is converted to gray scale/coloured image as pending image.
It is shown in Figure 2, for a kind of structure chart of embodiment of document/document direction detection device of the present invention.
In the present embodiment, document/document direction detection device includes preprocessing module 10, piecemeal module 20, extraction
Module 30, rotary module 40, Chinese character walking direction module 50 and document/document walking direction module 60.
Wherein, preprocessing module 10 generates vertical/horizontal image for being pre-processed to pending image.
For example, in one embodiment, preprocessing module 10 can seek the angle of inclination of literal line in pending image,
And rotate pending image and generate vertical/horizontal image so that the angle of inclination of literal line is 0 degree or 90 degree.
Piecemeal module 20 is used to, to vertical/horizontal image binaryzation, piecemeal, form multiple block images.Extraction module 30
Printed Chinese character image for extracting predetermined quantity in multiple block images.
For example, in one embodiment, extraction module 30 can seek the connection of predetermined quantity in each block image
Region;Judge whether each connected region is printed character;If printed character, then each block letter is further judged
Whether character is printed Chinese character, if printed Chinese character, then saves as printed Chinese character image.
0 degree of printed Chinese character image rotation that rotary module 40 is used to extract each, 90 degree, 180 degree, 270 degree,
Form the first printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character
Image.
Chinese character walking direction module 50 is used for respectively to the first printed Chinese character of printed Chinese character image described in each
Judgement is identified in image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image, obtains
To recognition result.
In one embodiment, the document/document direction detection device can also include the first accumulator 71, second
Accumulator 72, third accumulator 73 and the 4th accumulator 74.
Chinese character walking direction module 50 is respectively to the first printed Chinese character image of each printed Chinese character image,
Judgement is identified in two printed Chinese character images, third printed Chinese character image and the 4th printed Chinese character image, respectively obtains
First recognition result of each first printed Chinese character image and the first confidence level, each second printed Chinese character figure
Second recognition result of picture and the second confidence level, the third recognition result of each third printed Chinese character image and third are set
The 4th recognition result and the 4th confidence level of reliability, each 4th printed Chinese character image, are respectively compared each print
First confidence level, the numerical values recited of the second confidence level, third confidence level and the 4th confidence level of brush body Chinese character image, every
Maximum first confidence level of numerical value of a printed Chinese character image, the second confidence level, third confidence level or the 4th set
It is counted on the first accumulator 71, the second accumulator 72, third accumulator 73 or the 4th accumulator 74 corresponding to reliability and adds one.
Document/document walking direction module 60 is used to judge based on the recognition result direction of the document/document.
In one embodiment, it is cumulative can to compare the first accumulator 71, second for document/document walking direction module 60
The numerical values recited of device 72, third accumulator 73 and the 4th accumulator 74.And with the first accumulator 71, the second accumulator 72,
The corresponding first printed Chinese character image of numerical value the maximum, the second printed Chinese character figure in three accumulators 73 and the 4th accumulator 74
Direction of the direction of picture, third printed Chinese character image or the 4th printed Chinese character image as the document/document.
As an implementation, document/document direction detection device can also include loading unit 80.Loading unit 80
For loading document image and converting document image to gray scale/coloured image as pending image.
Document using the present invention/document direction detection method and detection device, can convenient, efficient automatic identification number
The direction of word image, and recognition accuracy is high.
Some embodiments of the present invention are described in detail above.Such as those skilled in the art institute energy
Understand, the whole either any steps or component of methods and apparatus of the present invention, can any computing device (including place
Reason device, storage medium etc.) either realized with hardware, firmware, software or combination thereof in the network of computing device,
This is that those of ordinary skill in the art can be real with their basic programming skill in the case where understanding present disclosure
Existing, therefore be not required to illustrate herein.
It is further clear that when arriving possible peripheral operation involved in explanation above, undoubtedly to use
Any display equipment being connected with any computing device and any input equipment, corresponding interface and control program.To sum up,
It related hardware, software in computer, computer system or computer network and realizes each in the preceding method of the present invention
Hardware, firmware, software or the combination thereof of kind operation, that is, constitute the equipment of the present invention and its each building block.
Therefore, it is based on above-mentioned understanding, the purpose of the present invention can also be by running one on any information processing equipment
Program or batch processing are realized.Described information processing equipment can be well known common apparatus.Therefore, the purpose of the present invention
It can also be realized only by the program product comprising the program code for realizing the method or equipment is provided.Namely
It says, such program product also constitutes the present invention, and the medium for storing or transmitting such program product also constitutes this hair
It is bright.Obviously, either transmission medium can be known to the skilled in the art or what be developed in the future appoints for the storage
The storage of what type either transmission medium therefore also It is not necessary to be enumerated to various storages or transmission medium at this.
In the device and method of the present invention, it is clear that each component or each step are after can decomposing, combine and/or decomposing
It reconfigures.These decompose and/or reconfigure the equivalent scheme that should be regarded as the present invention.It may also be noted that on executing
The step of stating series of processes can execute according to the sequence of explanation in chronological order naturally, but do not need to centainly according to when
Between sequentially execute.Certain steps can execute parallel or independently of one another.Meanwhile the specific embodiment of the invention is retouched above
In stating, the feature for describing and/or showing for a kind of embodiment can be in a manner of same or similar one or more
It is used in other embodiment, it is combined with the feature in other embodiment, or substitute the feature in other embodiment.
It should be emphasized that term "comprises/comprising" refers to the presence of feature, element, step or component when being used herein, but simultaneously
It is not excluded for the presence or additional of one or more other features, element, step or component.
Although the present invention and its advantage has been described in detail it should be appreciated that without departing from by the attached claims
Defined by can carry out various changes, replacement and transformation in the case of the spirit and scope of the present invention.Moreover, the model of the application
Enclose the specific embodiment for being not limited only to process, equipment, means, method and steps described in specification.In the art is common
Technical staff executes and corresponding reality described herein from the disclosure it will be readily understood that can be used according to the present invention
Apply the essentially identical function of example or obtain the result essentially identical with it, existing and process to be developed in future, equipment,
Means, method or step.Therefore, the attached claims are intended to include such process, equipment, hand in the range of them
Section, method or step.
Claims (10)
1. a kind of document/document direction detection method, which is characterized in that including:
Pending image is pre-processed, generates vertical/horizontal image, the vertical angle or level angle of literal line ±
Within 5 ° of deviation;
To the vertical/horizontal image binaryzation, piecemeal, multiple block images are formed;
The printed Chinese character image of predetermined quantity is extracted in multiple block images;
0 degree of printed Chinese character image rotation that each is extracted, 90 degree, 180 degree, 270 degree, form the first printed Chinese character
Image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;
Respectively to the first printed Chinese character image of printed Chinese character image described in each, the second printed Chinese character image,
Judgement is identified in three printed Chinese character images and the 4th printed Chinese character image, obtains recognition result;
The direction of the document/document is judged based on the recognition result;
" the printed Chinese character image that predetermined quantity is extracted in multiple block images " specifically includes:
In each block image, the connected region of predetermined quantity is sought by labelling method;
Judge whether each connected region is printed character;
If printed character, then further judge whether each described printed character is printed Chinese character, if printing
Body Chinese character then saves as printed Chinese character image;
" connected region that predetermined quantity is sought by labelling method " includes:
When printed character extracts, by symmetry, effective content ratio, stroke width and backbone length in connected region
Image screened, if the picture material in connected region is symmetrical above and below, symmetrical, abandon such connected region;Or
Person,
If the effective content pixel point of character stroke is then abandoned when proportion is higher by certain proportion in character Outside Dimensions frame
Such connected region;Alternatively,
Stroke width is more than character size certain proportion, and backbone length is more than certain length, then abandons such connected region;
" the judging whether each described printed character is printed Chinese character " includes:
Determining whether character by off-line training supporting vector machine model, the feature that training when takes has 0 degree of a character, and 90
Degree, 45 degree, the ratio shared by 135 degree of direction gradients and 8 kinds of basic strokes, 8 kinds of basic strokes include point, horizontal, vertical, slash,
Right-falling stroke, folding, curved, hook.
2. document according to claim 1/document direction detection method, which is characterized in that it is described " to pending image into
Row pretreatment, generates vertical/horizontal image " it specifically includes:
Seek the angle of inclination of literal line in the pending image;
It rotates the pending image and generates vertical/horizontal image so that the angle of inclination of the literal line is 0 degree or 90 degree.
3. document according to claim 1/document direction detection method, which is characterized in that described " respectively to each institute
State the first printed Chinese character image of printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and
Judgement is identified in 4th printed Chinese character image, obtains recognition result " it specifically includes:
Respectively to the first printed Chinese character image, the second printed Chinese character image, third of each printed Chinese character image
Judgement is identified in printed Chinese character image and the 4th printed Chinese character image, respectively obtains each printed Chinese character image
The first recognition result and the first confidence level, each second printed Chinese character image the second recognition result and the second confidence
Degree, the third recognition result of each third printed Chinese character image and third confidence level, each 4th block letter Chinese
4th recognition result of word image and the 4th confidence level;
It is respectively compared first confidence level, the second confidence level, third confidence level and of each printed Chinese character image
The numerical values recited of four confidence levels;
In maximum first confidence level of the numerical value of each printed Chinese character image, the second confidence level, third confidence level
Or it is counted on the 4th the first accumulator, the second accumulator, third accumulator or the 4th accumulator corresponding to confidence level plus one.
4. document according to claim 3/document direction detection method, which is characterized in that described " to be tied based on the identification
Fruit judges the direction of the document/document " it specifically includes:
Compare the numerical values recited of first accumulator, the second accumulator, third accumulator and the 4th accumulator;
With corresponding with numerical value the maximum in first accumulator, the second accumulator, third accumulator and the 4th accumulator
One printed Chinese character image, the second printed Chinese character image, third printed Chinese character image or the 4th printed Chinese character image
Direction of the direction as the document/document.
5. document according to claim 1/document direction detection method, which is characterized in that described " to pending image
Pre-processed, generate vertical/horizontal image " before, further include:
Load document/document image;
The document/document image is converted to gray scale/coloured image as the pending image.
6. a kind of document/document direction detection device, which is characterized in that including:
Preprocessing module generates vertical/horizontal image, the vertical angle of literal line for being pre-processed to pending image
Degree or level angle are within ± 5 ° of deviation;
Piecemeal module, for the vertical/horizontal image binaryzation, piecemeal, forming multiple block images;
Extraction module, the printed Chinese character image for extracting predetermined quantity in multiple block images;
Rotary module, 0 degree of printed Chinese character image rotation for being extracted to each, 90 degree, 180 degree, 270 degree, formed
First printed Chinese character image, the second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image;
Chinese character walking direction module, for respectively to the first printed Chinese character image of printed Chinese character image described in each,
Judgement is identified in second printed Chinese character image, third printed Chinese character image and the 4th printed Chinese character image, is known
Other result;
Document/document walking direction module, the direction for judging the document/document based on the recognition result;
The extraction module is specifically used for:
In each block image, the connected region of predetermined quantity is sought by labelling method;
Judge whether each connected region is printed character;
If printed character, then further judge whether each described printed character is printed Chinese character, if printing
Body Chinese character then saves as printed Chinese character image;
The stroke width of image is less than character size certain proportion in the wherein described connected region, and backbone length is no more than a fixed length
Degree.
7. document according to claim 6/document direction detection device, which is characterized in that the preprocessing module is specific
For:
Seek the angle of inclination of literal line in the pending image;
It rotates the pending image and generates vertical/horizontal image so that the angle of inclination of the literal line is 0 degree or 90 degree.
8. document according to claim 6/document direction detection device, which is characterized in that further include the first accumulator,
Two accumulators, third accumulator and the 4th accumulator;
The Chinese character walking direction module is specifically used for:
Respectively to the first printed Chinese character image, the second printed Chinese character image, third of each printed Chinese character image
Judgement is identified in printed Chinese character image and the 4th printed Chinese character image, respectively obtains each first printed Chinese character
The second recognition result and second of first recognition result of image and the first confidence level, each second printed Chinese character image
Confidence level, the third recognition result of each third printed Chinese character image and third confidence level, each 4th printing
4th recognition result of body Chinese character image and the 4th confidence level;
It is respectively compared first confidence level, the second confidence level, third confidence level and of each printed Chinese character image
The numerical values recited of four confidence levels;
In maximum first confidence level of the numerical value of each printed Chinese character image, the second confidence level, third confidence level
Or it is counted on the 4th the first accumulator, the second accumulator, third accumulator or the 4th accumulator corresponding to confidence level plus one.
9. document according to claim 8/document direction detection device, which is characterized in that sentence in the document/document direction
Disconnected module is specifically used for:
Compare the numerical values recited of first accumulator, the second accumulator, third accumulator and the 4th accumulator;
With corresponding with numerical value the maximum in first accumulator, the second accumulator, third accumulator and the 4th accumulator
One printed Chinese character image, the second printed Chinese character image, third printed Chinese character image or the 4th printed Chinese character image
Direction of the direction as the document/document.
10. document according to claim 9/document direction detection device, it is characterised in that further include:
Loading unit is made for loading document/document image and converting the document/document image to gray scale/coloured image
For the pending image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510296307.5A CN104915648B (en) | 2015-06-02 | 2015-06-02 | document/document direction detection method and detection device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510296307.5A CN104915648B (en) | 2015-06-02 | 2015-06-02 | document/document direction detection method and detection device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104915648A CN104915648A (en) | 2015-09-16 |
CN104915648B true CN104915648B (en) | 2018-07-20 |
Family
ID=54084700
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510296307.5A Active - Reinstated CN104915648B (en) | 2015-06-02 | 2015-06-02 | document/document direction detection method and detection device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104915648B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108345827B (en) * | 2017-01-24 | 2021-11-30 | 富士通株式会社 | Method, system and neural network for identifying document direction |
CN108681738A (en) * | 2018-03-21 | 2018-10-19 | 江苏善壶网络科技有限公司 | A kind of seal recognition methods and system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5020117A (en) * | 1988-01-18 | 1991-05-28 | Kabushiki Kaisha Toshiba | Handwritten character string recognition system |
CN1205485A (en) * | 1997-07-15 | 1999-01-20 | 三星电子株式会社 | Method for correcting direction of document image |
CN102402686A (en) * | 2011-12-07 | 2012-04-04 | 北京云星宇交通工程有限公司 | Method for dividing license plate characters based on connected domain analysis |
CN103870799A (en) * | 2012-12-17 | 2014-06-18 | 北京千橡网景科技发展有限公司 | Character direction judging method and device |
CN104008384A (en) * | 2013-02-26 | 2014-08-27 | 山东新北洋信息技术股份有限公司 | Character identification method and character identification apparatus |
-
2015
- 2015-06-02 CN CN201510296307.5A patent/CN104915648B/en active Active - Reinstated
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5020117A (en) * | 1988-01-18 | 1991-05-28 | Kabushiki Kaisha Toshiba | Handwritten character string recognition system |
CN1205485A (en) * | 1997-07-15 | 1999-01-20 | 三星电子株式会社 | Method for correcting direction of document image |
CN102402686A (en) * | 2011-12-07 | 2012-04-04 | 北京云星宇交通工程有限公司 | Method for dividing license plate characters based on connected domain analysis |
CN103870799A (en) * | 2012-12-17 | 2014-06-18 | 北京千橡网景科技发展有限公司 | Character direction judging method and device |
CN104008384A (en) * | 2013-02-26 | 2014-08-27 | 山东新北洋信息技术股份有限公司 | Character identification method and character identification apparatus |
Also Published As
Publication number | Publication date |
---|---|
CN104915648A (en) | 2015-09-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11455525B2 (en) | Method and apparatus of open set recognition and a computer readable storage medium | |
US10817741B2 (en) | Word segmentation system, method and device | |
WO2018166116A1 (en) | Car damage recognition method, electronic apparatus and computer-readable storage medium | |
CN103714327A (en) | Method and system for correcting image direction | |
CN110689658A (en) | Taxi bill identification method and system based on deep learning | |
US11893773B2 (en) | Finger vein comparison method, computer equipment, and storage medium | |
CN111274848A (en) | Image detection method and device, electronic equipment and storage medium | |
CN104915648B (en) | document/document direction detection method and detection device | |
Krishnan et al. | Detection of alphabets for machine translation of sign language using deep neural net | |
CN108229313A (en) | Face identification method and device, electronic equipment and computer program and storage medium | |
CN115797336A (en) | Fault detection method and device of photovoltaic module, electronic equipment and storage medium | |
CN110942067A (en) | Text recognition method and device, computer equipment and storage medium | |
CN111414889B (en) | Financial statement identification method and device based on character identification | |
CN110909578A (en) | Low-resolution image recognition method and device and storage medium | |
CN108334602B (en) | Data annotation method and device, electronic equipment and computer storage medium | |
CN110991307B (en) | Face recognition method, device, equipment and storage medium | |
CN113627442A (en) | Medical information input method, device, equipment and storage medium | |
CN106446902B (en) | Non-legible image-recognizing method and device | |
CN115688083B (en) | Method, device and equipment for identifying image-text verification code and storage medium | |
CN112861841B (en) | Training method and device for bill confidence value model, electronic equipment and storage medium | |
KR20220034327A (en) | Smart phone application automated control method based on deep-learning object detection | |
CN113688835A (en) | System based on object detection and OCR and method for extracting information of application fault elastic picture | |
Napitupulu et al. | Covid-19 Mask Detector Based on YOLO and MATLAB | |
CN114077479A (en) | Method for detecting malicious codes of client virtual machine in cloud platform | |
CN117853440A (en) | Training method for text defect detection model and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180720 Termination date: 20190602 |
|
RR01 | Reinstatement of patent right | ||
RR01 | Reinstatement of patent right |
Former decision: termination of patent right due to unpaid annual fee Former decision publication date: 20200522 |