CN106096601A - The method and system of character types in a kind of automatic detection bill - Google Patents
The method and system of character types in a kind of automatic detection bill Download PDFInfo
- Publication number
- CN106096601A CN106096601A CN201610396818.9A CN201610396818A CN106096601A CN 106096601 A CN106096601 A CN 106096601A CN 201610396818 A CN201610396818 A CN 201610396818A CN 106096601 A CN106096601 A CN 106096601A
- Authority
- CN
- China
- Prior art keywords
- character
- image
- bill
- character types
- automatic detection
- 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
Classifications
-
- 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
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Character Input (AREA)
Abstract
The invention discloses the method and system of character types in a kind of automatic detection bill, the method includes: A, the acquisition infrared image of bill, coloured image；B, binary conversion treatment infrared image and coloured image, the bianry image that output is corresponding；Character zone in C, location bianry image；D, the profile of detection character also generate boundary rectangle, judge character types according to the parameter of boundary rectangle.The system of the method is performed for correspondence.Character is extracted by the present invention by a series of process, carries out specificity analysis according to the parameter of the profile of character, distinguishes handwritten form and type-script, determination methods principle is simple, it is simple to realize, and saves time and resource, need not the early-stage preparations of complexity, improve identification efficiency simultaneously.
Description
Technical field
The present invention relates to the method and system of character types in a kind of automatic detection bill, belong to field of image recognition.
Background technology
Bank or unit can be manually to fill in beat for machine, in character recognition to some key element unit of bill
In, the recognition methods difference of type-script and handwritten form is very big, if the recognition methods of type-script character is known for handwritten character
Not, accuracy rate is very poor, and if the recognition methods of handwritten character is the most improper for type-script character recognition, and the time is multiple
Miscellaneous degree is high and wastes resource, and meanwhile, bank and the material of the unit pen to writing on bill and color also require, and uses and does not advises
What model pen was filled in must be for handwritten form, so the detection of specification is filled in the differentiation of handwritten character and type-script character for key element
Also there is great significance.
It is block letter or handwritten form that prior art distinguishes amount in Chinese character.Mainly extract and can distinguish printed character and hands
Write the up-and-down boundary flexibility feature of body character, stroke width characteristics, connected domain number feature, anyhow stroke feature totally 4
Component category feature, structural classification device is trained and identifies judging that check amount in Chinese character is block letter or handwritten form simultaneously.
The method has the disadvantage that (1) applicable surface is narrow, and its method is only applicable to judge that amount in Chinese character is block letter or handwritten form；
(2) after being partitioned into amount in Chinese character, it is judged that amount in Chinese character is block letter or the method for handwritten form relates to feature extraction, instruction
Practicing, more complicated, time complexity is high, wastes resource；(3) in terms of feature extraction, the method exists when image integral color
Time shallower, the value of stroke feature anyhow obtained can be made less；When cheque image color chin-deep, the connected domain number obtained can be made
Mesh eigenvalue is less；It is easily caused misclassification.
Summary of the invention
In order to solve the problems referred to above, the present invention is by providing the method for character types in a kind of automatic detection bill and being
System, improves discrimination power that is hand-written and that print word.
The technical solution used in the present invention is a kind of method of character types in automatic detection bill, it is characterised in that bag
Include following steps: A, obtain the infrared image of bill, coloured image；B, binary conversion treatment infrared image and coloured image, output
Corresponding bianry image；Character zone in C, location bianry image；D, the profile of detection character also generate boundary rectangle, according to
The parameter of boundary rectangle judges character types.
Preferably, step B also includes, by otsu algorithm, infrared image is carried out binary conversion treatment.
Preferably, step B also includes: obtain bill code name, is loaded into the general layout information of corresponding bill, root according to bill code name
It is red waterline district according to general layout information by color images and non-fills in district；Wherein, by retinex algorithm, red waterline district is entered
Row pretreatment, to strengthen image, then carries out binary conversion treatment by fixed threshold method to red waterline district；By otsu algorithm pair
Non-district of filling in carries out binary conversion treatment.
Preferably, step C also includes by connected component labeling algorithm location character region.
Preferably, the parameter of described boundary rectangle include Y coordinate, highly, width and boundary rectangle spacing.
Preferably, the threshold value of step D parameter and setting by comparing external matrix, it is judged that character types, character types
Including printing word and handwritten word.
Preferably, step D also includes that error judges, record prints word and the quantity of handwritten word, calculates both ratios, passes through
Contrast this ratio and error permissible value, it is judged that the character types of character string, wherein, error permissible value is for being used for defining judged result
Setting value, character string is boundary rectangle spacing character set within the specific limits.
The another aspect of the technical solution used in the present invention is the system of character types in a kind of automatic detection bill, bag
Include: image receiver module, for obtaining the infrared image of bill, coloured image；Image processing module, red for binary conversion treatment
Outer image and coloured image, the bianry image that output is corresponding；Character locating module, for positioning the character area in bianry image
Territory；Character processing module, for detecting the profile of character and generating boundary rectangle, judges character type according to the parameter of boundary rectangle
Type.
Beneficial effects of the present invention is for by providing the method and system of character types in a kind of automatic detection bill, utilizing
Bill is under infrared image, and the printed words of check own shows the strongest speciality, is extracted by character by a series of process, root
Carry out specificity analysis according to the parameter of the profile of character, distinguish handwritten form and type-script, it is judged that Method And Principle is simple, it is simple to be real
Existing, save time and resource, it is not necessary to complicated early-stage preparations, improve identification efficiency simultaneously.
Accompanying drawing explanation
Fig. 1 show the method flow signal of character types in a kind of automatic detection bill based on the embodiment of the present invention
Figure；
Fig. 2 show bill character machining flow chart based on the embodiment of the present invention；
Fig. 3 show the flow chart processing red waterline district based on the embodiment of the present invention.
Detailed description of the invention
Below in conjunction with embodiment, the present invention will be described.
Based on invention first embodiment, a kind of method of character types in automatic detection bill, including with
Lower step: A, the acquisition infrared image of bill, coloured image；B, binary conversion treatment infrared image and coloured image, output correspondence
Bianry image；Character zone in C, location bianry image；D, the profile of detection character also generate boundary rectangle, according to external
The parameter of rectangle judges character types.
Infrared ray radiation, on bill, by receiving and sense the light of reflection, forms the infrared image of bill；By general
Logical autochromy instrument obtains the coloured image of bill；The code name of bill, or feature identification identification is obtained by outside input
Bill.Binary conversion treatment image, orients the position of character by the attribute of pixel, uses opencv to provide
CvFindContours () function finds the profile of each character in bianry image, and extraction pattern uses and only extracts outermost wheel
Exterior feature, approach method use by be a little translated as point sequence form by chain code form, according to characters' property produce boundary rectangle,
Parameter according to boundary rectangle judges which kind of type these characters belong to.
Step B also includes, by otsu algorithm, infrared image is carried out binary conversion treatment.
Maximum variance between clusters is a kind of method that adaptive threshold value determines, is again Da-Jin algorithm, is called for short OTSU.It is by
The gamma characteristic of image, divides the image into background and target 2 part.Inter-class variance between background and target is the biggest, and structure is described
The difference becoming 2 parts of image is the biggest, when partial target mistake is divided into background or part background mistake to be divided into target all can cause 2 parts
Difference diminishes.Therefore, the segmentation making inter-class variance maximum means that misclassification probability is minimum.
As special product, check printing has special character, the character that check has been completed for printing originally as " in
State's bank money " etc., under conditions of infrared image, the character shown is the lightest, and client fills in or prints
Word up can clearly show that out, unique shortcoming be the character that shows in some cases may ratio shallower, this
Time be accomplished by light color character carry out image enhancement operation, in the case of not affecting background color, be allowed to blackening.For this
Situation, it is possible to use fuzzy defogging method strengthens the color of character in bill, reduces the impact of noise around character, then simultaneously
Re-use otsu algorithm and carry out binary conversion treatment.
Because for using infrared image processing to compare process coloured image, it is not necessary to processing interference, processing speed is more
Fast, therefore can preferentially carry out the process of infrared image, the most first obtain infrared image, after binary conversion treatment, carry out word
Symbol detection, if detection exists character, the most directly carries out character judgement, if can't detect character, then cancels at infrared image
Reason, changes carrying out the process of coloured image into.
As one embodiment of the present of invention, bill character machining flow chart as shown in Figure 2:
S21, beginning flow process；S22, binaryzation infrared image；S23, judge whether infrared image has word, have, enter S24, do not have
Then enter S25；S24, the character of extraction infrared image；S241, judge character types；S25, loading coloured image, and it is entered
Row binary conversion treatment；S251, extraction coloured image character；S252, judge character types；S26, output judged result.
Step B also includes obtaining bill code name, is loaded into the general layout information of corresponding bill according to bill code name, believes according to general layout
Cease color images and be red waterline district and non-fill in district；
Wherein, red waterline district carried out pretreatment to strengthen image by retinex algorithm, then by fixed threshold method to red
Waterline district carries out binary conversion treatment；By otsu algorithm, non-district of filling in is carried out binary conversion treatment.
Red waterline is the region including some red lines composition in the middle part of bill, and non-district of filling in is the sky not meeting and filling in specification
Between (on the printing type face of such as bill or edge), owing in red waterline district, red line can interfere by word thereon to identification,
Therefore using different processing methods from other regions, the non-purpose filling in district is the scope that reduction processes, by bill mould
The positional information that plate is had, is divided into red waterline district and non-district of filling in by pending image, and remainder is then for normal need
The region processed.
Retinex thought is that people perceives certain color put and brightness not merely depends on that this point enters the absolute of human eye
Light, the most relevant with brightness with color about.Substance theoretical for Retinex be the color of object be to length by object
That the reflectance of ripple (red), medium wave (green) and shortwave (blue) light determines rather than determined by the absolute value of intensity of reflected light
's；The color of object is not affected by illumination is heteropic, has concordance, i.e. Retinex theory is with color constancy (color
Constancy) based on.Incident illumination L reflection is obtained by the image S of the object that observer is seen by body surface, reflection
Rate R is determined by object itself, is not changed by incident illumination L.Its formula is S(x, y)=R(x, y) * L(x, y), and S(x in formula, y)
Represent the picture signal that observed or photographing unit receives；L(x, y) represents the irradiation component of ambient light；R(x, y) expression is carried
The reflecting component of the target object of image detail information, L(x, y).Core estimates illumination L exactly, estimates L * component from image S,
And remove L * component, obtain Primitive reflex components R.
As an embodiment of the present invention program, process the flow chart in red waterline district as shown in Figure 3:
S31, travel through pretreated image, if the blue color component value of certain pixel is less than threshold value Blue T and blue component, green
Colouring component, red component combination of two, when difference is respectively less than threshold value Each T, by meet above-mentioned under the conditions of the picture of correspondence position
Vegetarian refreshments pixel value is set to 255, and the pixel value of other positions is set to 0, is partitioned into the most roughly the black character in red waterline,
Wherein Blue T and Each T is the value set；S32, the image obtaining previous step carry out gaussian filtering；S33, use are admittedly
Determine threshold method and carry out binary conversion treatment；S34, the image obtaining previous step use oval structure to carry out 1 expansive working；
S35, the image obtaining previous step carry out opening operation operation；The image phase arrived of s36, the image that s31 is obtained and s35
With；Zonule independent in the image that s37, employing connected component labeling algorithm tag previous step obtain, and by removing of small regions,
The finest is partitioned into red underwater black character.
By otsu algorithm, non-district of filling in is carried out binary conversion treatment, because the background color of this subregion is thin, no
Need complicated process, it is possible to use self adaptation color range and contrast method carry out pretreatment operation to image at non-red waterline,
Obtain contrast and strengthen image, the pretreated image of gray processing, then carry out binary conversion treatment by otsu algorithm.
Step C also includes by connected component labeling algorithm location character region.
Connected region generally refers to have in image the image of the adjacent foreground pixel point composition of same pixel value and position
Region.Connected component analysis refers to find out each connected region in image and labelling, and a connected region is by having phase
With the sets of adjacent pixels pixel set of pixel value, therefore, we just can find connection by the two condition in the picture
Region, for each connected region found, we give one unique mark (Label), to distinguish other connected regions
Territory.
The most important method of binary image analysis is exactly connected component labeling, and it is the basis of all binary image analysises,
It, by white or the labelling of black picture element (target) in bianry image, allows each single connected region form one and is marked
The block known, further we just can obtain the geometric parameters such as the profile of these blocks, boundary rectangle, barycenter.
The parameter of boundary rectangle include Y coordinate, highly, width and boundary rectangle spacing.
Y coordinate i.e. rectangular top coordinate, the distance that height is rectangular base to top, width is rectangle left and right distance, square
Shape spacing is character rectangle spacing.
The threshold value of step D parameter and setting by comparing external matrix, it is judged that character types, character types include printing
Word and handwritten word.
The relatively y-coordinate (judging that character rectangle top is the most point-blank) of each character-circumscribed rectangle, highly (judgement
Whether character rectangular elevation the same), width (judge character rectangle width whether as), two intercharacter intervals (judge each character
Whether the spacing between rectangle is the same) difference that the most simultaneously meets boundary rectangle y-coordinate two-by-two is not more than L, and boundary rectangle is high two-by-two
The difference of degree is not more than H, and the difference of boundary rectangle width is less than W two-by-two, and in three characters, the difference of two spacing is less than S, and wherein L, H, W, S are
Setting value.
If meeting, then recording this monocase is type fount, and counts；If being unsatisfactory for, then it is hand-written for recording this monocase
Font, and count.
Step D also includes that error judges, record prints word and the quantity of handwritten word, calculates both ratios, should by contrast
Ratio and error permissible value, it is judged that the character types of character string, wherein, error permissible value is the setting for defining judged result
Value, character string is boundary rectangle spacing character set within the specific limits.
Due to the difference of the font of personal handwritten word, step-up error permissible value, i.e. in a string character, most character
It is judged as hand-written or prints, i.e. thinking that this string character belongs to hand-written or prints；And the judgement of character string, with character
The Distance Judgment of boundary rectangle, it is clear that when a last character is the biggest with the spacing of a series of characters before when,
This character the most probably belongs to another character string.
The system of character types in a kind of automatic detection bill, including: image receiver module, for obtaining the infrared of bill
Image, coloured image；Image processing module, for binary conversion treatment infrared image and coloured image, the binary map that output is corresponding
Picture；Character locating module, for positioning the character zone in bianry image；Character processing module, for detecting the profile of character
And generate boundary rectangle, judge character types according to the parameter of boundary rectangle.
The above, simply presently preferred embodiments of the present invention, the invention is not limited in above-mentioned embodiment, as long as
It reaches the technique effect of the present invention with identical means, all should belong to protection scope of the present invention.Protection model in the present invention
In enclosing, its technical scheme and/or embodiment can have various different modifications and variations.
Claims (8)
1. the method for character types in an automatic detection bill, it is characterised in that comprise the following steps:
A, the acquisition infrared image of bill, coloured image；
B, binary conversion treatment infrared image and coloured image, the bianry image that output is corresponding；
Character zone in C, location bianry image；
D, the profile of detection character also generate boundary rectangle, judge character types according to the parameter of boundary rectangle.
The method of character types in a kind of automatic detection bill the most according to claim 1, step B also includes passing through otsu
Algorithm carries out binary conversion treatment to infrared image.
The method of character types in a kind of automatic detection bill the most according to claim 1, step B also includes:
Obtain bill code name, be loaded into the general layout information of corresponding bill according to bill code name, according to general layout information, coloured image divided
It is segmented into red waterline district and non-fills in district；
Wherein, red waterline district carried out pretreatment to strengthen image by retinex algorithm, then by fixed threshold method to red
Waterline district carries out binary conversion treatment；
By otsu algorithm, non-district of filling in is carried out binary conversion treatment.
The method of character types in a kind of automatic detection bill the most according to claim 1, step C also includes by connection
Field mark algorithm location character region.
The method of character types, the parameter of described boundary rectangle in a kind of automatic detection bill the most according to claim 1
Including Y coordinate, highly, width and boundary rectangle spacing.
A kind of method of character types in automatic detection bill, outside step D is by comparison
Connecing the parameter of matrix and the threshold value of setting, it is judged that character types, character types include printing word and handwritten word.
The method of character types in a kind of automatic detection bill the most according to claim 6, step D also includes that error is sentenced
Fixed, record prints word and the quantity of handwritten word, calculates both ratios, by contrasting this ratio and error permissible value, it is judged that character
The character types of string, wherein, error permissible value is the setting value for defining judged result, and character string is that boundary rectangle spacing exists
A range of character set.
8. the system of character types in an automatic detection bill, it is characterised in that including:
Image receiver module, for obtaining the infrared image of bill, coloured image；
Image processing module, for binary conversion treatment infrared image and coloured image, the bianry image that output is corresponding；
Character locating module, for positioning the character zone in bianry image；
Character processing module, for detecting the profile of character and generating boundary rectangle, judges character according to the parameter of boundary rectangle
Type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610396818.9A CN106096601B (en) | 2016-06-06 | 2016-06-06 | The method and system of character types in a kind of automatic detection bill |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610396818.9A CN106096601B (en) | 2016-06-06 | 2016-06-06 | The method and system of character types in a kind of automatic detection bill |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106096601A true CN106096601A (en) | 2016-11-09 |
| CN106096601B CN106096601B (en) | 2019-06-18 |
Family
ID=57448675
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610396818.9A Expired - Fee Related CN106096601B (en) | 2016-06-06 | 2016-06-06 | The method and system of character types in a kind of automatic detection bill |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106096601B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952391A (en) * | 2017-02-27 | 2017-07-14 | 深圳怡化电脑股份有限公司 | One kind is stained Paper Currency Identification and device |
| CN107194400A (en) * | 2017-05-31 | 2017-09-22 | 北京天宇星空科技有限公司 | A kind of finance reimbursement unanimous vote is according to picture recognition processing method |
| CN107958205A (en) * | 2017-10-31 | 2018-04-24 | 北京艾克利特光电科技有限公司 | Gunnery training intelligent management system |
| CN108416765A (en) * | 2018-01-30 | 2018-08-17 | 华南理工大学 | A kind of character defect automatic testing method and system |
| CN109902563A (en) * | 2019-01-17 | 2019-06-18 | 桂林远望智能通信科技有限公司 | A kind of multi-angle model recognizing method and system |
| CN110032996A (en) * | 2018-01-11 | 2019-07-19 | 台达电子工业股份有限公司 | The character skewness correction and segmentation devices and methods therefor of basis of classification formula |
| CN111144191A (en) * | 2019-08-14 | 2020-05-12 | 广东小天才科技有限公司 | Font identification method and device, electronic equipment and storage medium |
| CN111340032A (en) * | 2020-03-16 | 2020-06-26 | 天津得迈科技有限公司 | Character recognition method based on application scene in financial field |
| CN112530079A (en) * | 2019-09-17 | 2021-03-19 | 深圳怡化电脑股份有限公司 | Method and device for detecting bill factors, terminal equipment and storage medium |
| CN112767430A (en) * | 2021-01-06 | 2021-05-07 | 维沃移动通信有限公司 | Information prompting method and device |
| CN113191360A (en) * | 2020-01-14 | 2021-07-30 | 深圳怡化电脑股份有限公司 | Method and device for positioning character outline of bill |
| CN114549504A (en) * | 2022-03-01 | 2022-05-27 | 安徽工业技术创新研究院六安院 | Appearance quality detection method based on machine vision |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751567A (en) * | 2008-12-12 | 2010-06-23 | 汉王科技股份有限公司 | Quick text recognition method |
| CN102779275A (en) * | 2012-07-04 | 2012-11-14 | 广州广电运通金融电子股份有限公司 | Paper characteristic identification method and relative device |
| US20120308127A1 (en) * | 2009-12-15 | 2012-12-06 | Fujitsu Frontech Limited | Character recognition method, character recognition apparatus, and character recognition program |
| CN104952143A (en) * | 2015-06-18 | 2015-09-30 | 辰通智能设备(深圳)有限公司 | Color change detection method and color change detection system for red waterlines of bill images |
| CN105320951A (en) * | 2014-06-23 | 2016-02-10 | 株式会社日立信息通信工程 | Optical character recognition apparatus and optical character recognition method |
-
2016
- 2016-06-06 CN CN201610396818.9A patent/CN106096601B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751567A (en) * | 2008-12-12 | 2010-06-23 | 汉王科技股份有限公司 | Quick text recognition method |
| US20120308127A1 (en) * | 2009-12-15 | 2012-12-06 | Fujitsu Frontech Limited | Character recognition method, character recognition apparatus, and character recognition program |
| CN102779275A (en) * | 2012-07-04 | 2012-11-14 | 广州广电运通金融电子股份有限公司 | Paper characteristic identification method and relative device |
| CN105320951A (en) * | 2014-06-23 | 2016-02-10 | 株式会社日立信息通信工程 | Optical character recognition apparatus and optical character recognition method |
| CN104952143A (en) * | 2015-06-18 | 2015-09-30 | 辰通智能设备(深圳)有限公司 | Color change detection method and color change detection system for red waterlines of bill images |
Non-Patent Citations (1)
| Title |
|---|
| 朱宁波 等: "支票大写金额的印刷体与手写体辨识", 《计算机工程》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952391A (en) * | 2017-02-27 | 2017-07-14 | 深圳怡化电脑股份有限公司 | One kind is stained Paper Currency Identification and device |
| CN107194400A (en) * | 2017-05-31 | 2017-09-22 | 北京天宇星空科技有限公司 | A kind of finance reimbursement unanimous vote is according to picture recognition processing method |
| CN107958205A (en) * | 2017-10-31 | 2018-04-24 | 北京艾克利特光电科技有限公司 | Gunnery training intelligent management system |
| CN107958205B (en) * | 2017-10-31 | 2021-11-16 | 贵州景浩科技有限公司 | Shooting training intelligent management system |
| CN110032996B (en) * | 2018-01-11 | 2021-06-04 | 台达电子工业股份有限公司 | Character inclination correcting device and method based on classification |
| CN110032996A (en) * | 2018-01-11 | 2019-07-19 | 台达电子工业股份有限公司 | The character skewness correction and segmentation devices and methods therefor of basis of classification formula |
| CN108416765A (en) * | 2018-01-30 | 2018-08-17 | 华南理工大学 | A kind of character defect automatic testing method and system |
| CN109902563A (en) * | 2019-01-17 | 2019-06-18 | 桂林远望智能通信科技有限公司 | A kind of multi-angle model recognizing method and system |
| CN111144191A (en) * | 2019-08-14 | 2020-05-12 | 广东小天才科技有限公司 | Font identification method and device, electronic equipment and storage medium |
| CN111144191B (en) * | 2019-08-14 | 2024-03-22 | 广东小天才科技有限公司 | Font identification method, font identification device, electronic equipment and storage medium |
| CN112530079A (en) * | 2019-09-17 | 2021-03-19 | 深圳怡化电脑股份有限公司 | Method and device for detecting bill factors, terminal equipment and storage medium |
| CN112530079B (en) * | 2019-09-17 | 2022-10-11 | 深圳怡化电脑股份有限公司 | Method, device, terminal equipment and storage medium for detecting bill factors |
| CN113191360A (en) * | 2020-01-14 | 2021-07-30 | 深圳怡化电脑股份有限公司 | Method and device for positioning character outline of bill |
| CN111340032A (en) * | 2020-03-16 | 2020-06-26 | 天津得迈科技有限公司 | Character recognition method based on application scene in financial field |
| CN112767430A (en) * | 2021-01-06 | 2021-05-07 | 维沃移动通信有限公司 | Information prompting method and device |
| WO2022148293A1 (en) * | 2021-01-06 | 2022-07-14 | 维沃移动通信有限公司 | Information prompting method and apparatus |
| CN114549504A (en) * | 2022-03-01 | 2022-05-27 | 安徽工业技术创新研究院六安院 | Appearance quality detection method based on machine vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106096601B (en) | 2019-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106096601A (en) | The method and system of character types in a kind of automatic detection bill | |
| CN104392205B (en) | A kind of recognition methods and system of improper license plate | |
| CN106845480A (en) | A kind of method that car plate is recognized from picture | |
| CN103310211B (en) | A kind ofly fill in mark recognition method based on image procossing | |
| CN102043950B (en) | Vehicle outline recognition method based on canny operator and marginal point statistic | |
| CN103116751B (en) | A kind of Method of Automatic Recognition for Character of Lcecse Plate | |
| CN107273896A (en) | A kind of car plate detection recognition methods based on image recognition | |
| CN104751142A (en) | Natural scene text detection algorithm based on stroke features | |
| CN106096610B (en) | A kind of file and picture binary coding method based on support vector machines | |
| CN103198315B (en) | Based on the Character Segmentation of License Plate of character outline and template matches | |
| CN104715252B (en) | A kind of registration number character dividing method of dynamic template combination pixel | |
| CN109284758A (en) | A kind of invoice seal removing method, device and computer storage medium | |
| CN108805018A (en) | Road signs detection recognition method, electronic equipment, storage medium and system | |
| CN106156768A (en) | The vehicle registration certificate detection method of view-based access control model | |
| CN109034145A (en) | Bank's card number recognition methods based on OpenCV | |
| CN101599125A (en) | The binarization method that the complex background hypograph is handled | |
| CN103413147A (en) | Vehicle license plate recognizing method and system | |
| CN104700092A (en) | Small-character number identification method based on template and feature matching | |
| CN105844242A (en) | Method for detecting skin color in image | |
| CN109784342A (en) | A kind of OCR recognition methods and terminal based on deep learning model | |
| CN104766344B (en) | Vehicle checking method based on movement edge extractor | |
| CN102542244A (en) | Face detection method and system and computer program product | |
| CN109191429A (en) | A kind of 3D printing nozzle detection method based on machine vision | |
| CN113158977A (en) | Image character editing method for improving FANnet generation network | |
| CN104573656B (en) | A kind of car plate color determination methods based on connected region information |
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 |
Granted publication date: 20190618 Termination date: 20200606 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |