CN106874911A - The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing - Google Patents
The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing Download PDFInfo
- Publication number
- CN106874911A CN106874911A CN201710123187.8A CN201710123187A CN106874911A CN 106874911 A CN106874911 A CN 106874911A CN 201710123187 A CN201710123187 A CN 201710123187A CN 106874911 A CN106874911 A CN 106874911A
- Authority
- CN
- China
- Prior art keywords
- digital
- area ratio
- carries out
- printing
- zonule
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Abstract
It is the main differentiation point of numeral contrast that the present invention uses the area ratio shared by digital stroke, because the main rope for contrasting is cited as area ratio rather than the absolute position of digital stroke or digital pen cuttings product in itself, and contrasted after area ratio is quantified in this course, different fonts are thus counteracted to a certain extent, the influence that scanning process and clear background degree bring, so as to improve the accuracy rate of numeral identification.At the same time, because overall process only carries out simple mathematical operation step, thus reduce the work of processor it is negative but, improve the speed of printing digit recognizing and contrast.
Description
Technical field
The invention belongs to a kind of technical field that automatic identification is carried out to the printing digital that is scanned through obtaining, particularly
Apply the printing digital on to image to carry out during automatic identification, it is adaptable to which check is achieved, car plate scanning, postcode identification,
The related technical fields such as the automatic identifications of digital picture such as digital instrument monitoring.
Background technology
Automatic identification is carried out to printing digital to be widely used, how more simply and effectively to recognize and scan at present
To image in blocks letter Arabic numeral be automatic identification technology next step institute facing challenges.Due to printing digital
Font difference, while the uncertainty of the angle and orientation when being scanned every time after scanning again, therefore different journeys can be brought
The deformation of degree, while blurred background and the not good situation of actual print quality and objective reality, therefore how effectively to process this
Numeral in the problem that series of factors is brought, such as image is contaminated, break pen and deformation the problems such as, improve digital identification whereby
Accuracy rate be automatic identification technology key point.
The content of the invention
1st, the area ratio that a kind of application partitioning is obtained is the method that carries out printing digit recognizing, it is characterized in that:Will
Region where each printing digital being identified is divided into three rows three and arranges equal-sized nine zonule blocks, Ran Houying
Automatic identification is carried out for printing digital.
2nd, area ratio that a kind of application partitioning according to claim 1 is obtained carries out printing digit recognizing
Method, it is characterized in that:Calculate digital stroke occupied area in each zonule block.
3rd, area ratio that a kind of application partitioning according to claim 1 is obtained carries out printing digit recognizing
Method, it is characterized in that:Calculate the ratio that digital stroke occupied area in each zonule block accounts for this zonule block gross area
And record.
4th, area ratio that a kind of application partitioning according to claim 1 is obtained carries out printing digit recognizing
Method, it is characterized in that:This zonule block area shared by digital stroke in each zonule block will be recorded as the above method
Ratio is compared one by one with the ratio of identical order and known digital template.
The area ratio obtained the invention provides a kind of application partitioning the method that carries out printing digit recognizing, its
Purpose is the identification and the process for matching in order to accurately and quickly realize printing digital.
The technical solution adopted for the present invention to solve the technical problems is:
The area ratio that a kind of application partitioning is obtained the method that carries out printing digit recognizing, block letter number that will be identified
Word is input in computer by scanner, same decile treatment is then carried out to each numeral that scanning is obtained, i.e., not
The information for directly being included in itself using each digital stroke is identified, but each numeral is divided into the nine of equalization first
Part, then calculate and the ratio of this zonule block gross area accounted for per stroke occupied area in portion, then with identical order according to
It is secondary to be contrasted with known digital template, if the absolute value of the difference of the stroke occupied area ratio value of a certain region unit is less than
The numerical value of thresholding 1, then it is assumed that this region Block- matching is simultaneously recorded, then carries out the numeral identification of next step.
Think that the number of the region unit of matching reaches door set in advance in nine region units of each described numeral
Limit value 2 then thinks that this identified numeral and the numeral corresponding to digital template are same numerals.
Described two threshold values set in advance can be adjusted, and discrimination is also just changed along with the regulation of threshold value
The two main digital distinguishing indexes with accuracy rate.
The beneficial effects of the invention are as follows:
Identified printing digital is divided into zonule block by the present invention, followed by each cell shared by digital stroke
The area ratio of domain block come carry out numeral identification contrast, simply and reliablely pre-process each identified printing digital and included
Information, by using stroke occupied area ratio rather than stroke in itself come carry out data identification counteract to a certain extent
The change of the stroke absolute position that the factors such as different fonts and different definition are brought, which improves printing digit recognizing
Accuracy rate.With traditional direct control methods of digital stroke comparatively, due to being compared using relative quantity, therefore
Deformation resolving capacity to numeral increases, while improve the accuracy rate of identification.
Specific embodiment
The present invention is that a kind of area ratio value for calculating each isolated area block shared by stroke carries out printing digital
Know method for distinguishing, including by scanning after identified printing digital be divided into three rows three first and arrange nine zonule blocks altogether,
Then the stroke occupied area ratio value of each region unit this key element is contrasted again, and nine described area ratio sequences are represented
Each digital basic feature.
Identified printing digital of the present invention after first by scanning is divided into three rows three and arranges nine zonules altogether
The relevant parameters such as block, all length of sides of this nine zonule blocks and area are identical, and this nine zonule blocks will be every
In individual identified printing digital is completely contained in(As shown in Figure 1).
Brief description of the drawings
Fig. 1 is digital zoning schematic diagram.
Then independent calculating is carried out for each zonule BOB(beginning of block), the stroke that numeral is identified in the block of one's respective area is obtained first
Shared region, then calculates stroke occupied area, then calculates stroke occupied area and account for this zonule block gross area
Ratio, the comparison for next step is stored by this ratio.In this step, at the same complete to area ratio this
The quantizing process of analog quantity.
Contrasted with known digital template successively with identical zonule block order again, if a certain region unit
Stroke occupied area ratio value difference absolute value less than thresholding 1 numerical value, then it is assumed that this region Block- matching is simultaneously recorded
Come, then carry out the numeral identification of next step.The acquisition of wherein digital template is to choose several common block letter font, will
After they are processed according to above-mentioned identical method, the mathematical mean of the stroke occupied area ratio of each zonule block is taken
As template count.The size of thresholding 1 be can adjust, and identification different printing body font can be changed by the numerical value for adjusting thresholding 1
When allowable error between the block letter and digital template, discrimination this numerical value is also just changed simultaneously.
The number of all region units for thinking matching in nine region units of each above-mentioned numeral is recorded it
After sue for peace, think if this numerical value reaches threshold value set in advance 2 this identified numeral and digital template it is right
The numeral answered is same numeral.The size of thresholding 2 be can adjust, and identification different printing can be changed by the numerical value for adjusting thresholding 2
Permission tolerance limit during body font between the block letter and template, also just changes the accuracy rate of identification this numerical value simultaneously.
Described two threshold values set in advance can be adjusted, and discrimination is also just changed along with the regulation of threshold value
The two main digital distinguishing indexes with accuracy rate.
In the identification process of the above-mentioned numeral of execution, numeral is missed carefully due to that can exist during actual digital scan
The situation of background noise or miscellaneous point is saved or exists, in this case, by the threshold value 1 and threshold value 2 that adjust matching
The digital purpose with known digital template matching degree is identified to reach adjustment, that is, changes discrimination and standard simultaneously
True the two main digital distinguishing indexes of rate.Similarly, for different block letter fonts, as long as the threshold value for setting is more than being somebody's turn to do
Error between font and template, also just adapts to the identification of corresponding font, it is also possible to same raising printing digit recognizing
Accuracy rate.
It is the main differentiation point of numeral contrast that the present invention uses the area ratio shared by digital stroke, due to the main rope for contrasting
Area ratio is cited as rather than the absolute position of digital stroke or digital pen cuttings product in itself, and in this course by area
Ratio is contrasted after being quantified, and different fonts, scanning process and clear background degree are thus counteracted to a certain extent
The influence for bringing, so as to improve the accuracy rate of numeral identification.At the same time, because overall process only carries out simple mathematical operation
Step, thus reduce the work of processor it is negative but, improve the speed of printing digit recognizing and contrast.
Claims (4)
1. the area ratio for obtaining in a kind of application partitioning the method that carries out printing digit recognizing, it is characterized in that:Will be by
Region where each printing digital of identification is divided into three rows three and arranges equal-sized nine zonule blocks, then applies
Automatic identification is carried out in printing digital.
2. the area ratio that a kind of application partitioning according to claim 1 is obtained carries out the side of printing digit recognizing
Method, it is characterized in that:Calculate digital stroke occupied area in each zonule block.
3. the area ratio that a kind of application partitioning according to claim 1 is obtained carries out the side of printing digit recognizing
Method, it is characterized in that:Digital stroke occupied area accounts for the ratio of this zonule block gross area and remembers in calculating each zonule block
Record is got off.
4. the area ratio that a kind of application partitioning according to claim 1 is obtained carries out the side of printing digit recognizing
Method, it is characterized in that:The ratio of this zonule block area shared by digital stroke in each zonule block will be recorded as the above method
Compared one by one with the ratio of identical order and known digital template.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710123187.8A CN106874911A (en) | 2017-03-03 | 2017-03-03 | The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710123187.8A CN106874911A (en) | 2017-03-03 | 2017-03-03 | The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106874911A true CN106874911A (en) | 2017-06-20 |
Family
ID=59169827
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710123187.8A Pending CN106874911A (en) | 2017-03-03 | 2017-03-03 | The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106874911A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107491422A (en) * | 2017-08-14 | 2017-12-19 | 广州阿里巴巴文学信息技术有限公司 | Character displaying method, device and electronic equipment |
CN109271986A (en) * | 2018-07-02 | 2019-01-25 | 国家海洋局东海标准计量中心(国家海洋计量站上海分站) | A kind of digit recognition method based on Second-Confirm |
US11301709B2 (en) * | 2018-02-16 | 2022-04-12 | Kabushiki Kaisha Toshiba | Reading system, reading device, and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101025789A (en) * | 2007-03-29 | 2007-08-29 | 上海大学 | Digital identification method for extracting digital image selective bone area features based on computation |
CN101038626A (en) * | 2007-04-25 | 2007-09-19 | 上海大学 | Method and device for recognizing test paper score |
CN103902981A (en) * | 2014-04-02 | 2014-07-02 | 浙江师范大学 | Method and system for identifying license plate characters based on character fusion features |
WO2015183015A1 (en) * | 2014-05-30 | 2015-12-03 | 삼성에스디에스 주식회사 | Character recognition method and apparatus therefor |
CN105184265A (en) * | 2015-09-14 | 2015-12-23 | 哈尔滨工业大学 | Self-learning-based handwritten form numeric character string rapid recognition method |
CN105574531A (en) * | 2015-12-11 | 2016-05-11 | 中国电力科学研究院 | Intersection point feature extraction based digital identification method |
-
2017
- 2017-03-03 CN CN201710123187.8A patent/CN106874911A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101025789A (en) * | 2007-03-29 | 2007-08-29 | 上海大学 | Digital identification method for extracting digital image selective bone area features based on computation |
CN101038626A (en) * | 2007-04-25 | 2007-09-19 | 上海大学 | Method and device for recognizing test paper score |
CN103902981A (en) * | 2014-04-02 | 2014-07-02 | 浙江师范大学 | Method and system for identifying license plate characters based on character fusion features |
WO2015183015A1 (en) * | 2014-05-30 | 2015-12-03 | 삼성에스디에스 주식회사 | Character recognition method and apparatus therefor |
CN105184265A (en) * | 2015-09-14 | 2015-12-23 | 哈尔滨工业大学 | Self-learning-based handwritten form numeric character string rapid recognition method |
CN105574531A (en) * | 2015-12-11 | 2016-05-11 | 中国电力科学研究院 | Intersection point feature extraction based digital identification method |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107491422A (en) * | 2017-08-14 | 2017-12-19 | 广州阿里巴巴文学信息技术有限公司 | Character displaying method, device and electronic equipment |
US11301709B2 (en) * | 2018-02-16 | 2022-04-12 | Kabushiki Kaisha Toshiba | Reading system, reading device, and storage medium |
US20220189182A1 (en) * | 2018-02-16 | 2022-06-16 | Kabushiki Kaisha Toshiba | Reading system, reading device, and storage medium |
US11915498B2 (en) * | 2018-02-16 | 2024-02-27 | Kabushiki Kaisha Toshiba | Reading system, reading device, and storage medium |
CN109271986A (en) * | 2018-07-02 | 2019-01-25 | 国家海洋局东海标准计量中心(国家海洋计量站上海分站) | A kind of digit recognition method based on Second-Confirm |
CN109271986B (en) * | 2018-07-02 | 2021-11-26 | 国家海洋局东海标准计量中心(国家海洋计量站上海分站) | Digital identification method based on Second-Confirm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106874911A (en) | The area ratio that a kind of application partitioning is obtained is come the method that carries out printing digit recognizing | |
CN110728225B (en) | High-speed face searching method for attendance checking | |
CN109859227B (en) | Method and device for detecting flip image, computer equipment and storage medium | |
CN1035904C (en) | Estimation of baseline, line spacing and character height for handwriting recognition | |
CN107067044A (en) | A kind of finance reimbursement unanimous vote is according to intelligent checks system | |
CN106096610B (en) | A kind of file and picture binary coding method based on support vector machines | |
US10311322B2 (en) | Character information recognition method based on image processing | |
CN110619274A (en) | Identity verification method and device based on seal and signature and computer equipment | |
CN107194400A (en) | A kind of finance reimbursement unanimous vote is according to picture recognition processing method | |
EP2383678A1 (en) | Handwritten character recognition method and system | |
CN109840524B (en) | Text type recognition method, device, equipment and storage medium | |
CN109858476B (en) | Tag expansion method and electronic equipment | |
CN107014291B (en) | A kind of vision positioning method of material precision transfer platform | |
CN1149508C (en) | On-line character recognition system | |
CN112052852A (en) | Character recognition method of handwritten meteorological archive data based on deep learning | |
CN111091124A (en) | Spine character recognition method | |
CN108960235A (en) | A method of answer card filling block for identification | |
CN1135492C (en) | Handwriting verification device | |
CN111626280A (en) | Method and device for identifying answer sheet without positioning point | |
CN103761518A (en) | Method for processing vote information | |
CN107622562A (en) | Extracting method, device, equipment and the storage medium of banknote image | |
CN114463425B (en) | Workpiece surface featureless point positioning method based on probability Hough straight line detection | |
CN110147792B (en) | Medicine package character high-speed detection system and method based on memory optimization | |
CN107729863B (en) | Human finger vein recognition method | |
CN106599910B (en) | Mimeograph documents discrimination method based on texture recombination |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170620 |
|
WD01 | Invention patent application deemed withdrawn after publication |