CN103268490A - Digital recognition method based on two-side and three-width characteristic - Google Patents
Digital recognition method based on two-side and three-width characteristic Download PDFInfo
- Publication number
- CN103268490A CN103268490A CN201310209889XA CN201310209889A CN103268490A CN 103268490 A CN103268490 A CN 103268490A CN 201310209889X A CN201310209889X A CN 201310209889XA CN 201310209889 A CN201310209889 A CN 201310209889A CN 103268490 A CN103268490 A CN 103268490A
- Authority
- CN
- China
- Prior art keywords
- character
- envelope
- identification
- value
- picture
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Character Input (AREA)
- Character Discrimination (AREA)
Abstract
The invention discloses a digital recognition method based on the two-side and three-width characteristic and belongs to the field of image recognition. According to the digital recognition method based on the two-side and three-width characteristic, an acquired single-character image is normalized to a size suitable for processing, and then recognition of the ten digital characters from 0 to 9 is carried out based on the external physical characteristic of a digit. Specifically, the method comprises the steps of obtaining the left envelop EL and the right envelop ER of each character and calculating the head width WH, the middle width WM and WM0 and the foot width WF, namely two sides and three widths; recognizing the character 7 and the character 1 based on the three-width characteristics; recognizing the character 4 through three times of itemizing detection; obtaining opening parameters, L0 and L1, of the left envelop of each character and opening parameters, R0 and R1, of the right envelop of each character, and recognizing characters 0, 2, 3, 5, 6, 8 and 9 in a classified mode based on the opening and closing characteristics of the left envelops and the right envelops. The digital recognition method based on the two-side and three-width characteristic is used for recognition of print form digits, capable of achieving accurate and fast recognition of print form digits on notes with different fonts, different sizes and different definition degrees, and high in accuracy and real-time performance.
Description
Technical field
The invention belongs to the technical field of image recognition, relate in particular to the recognition methods to press figure.
Background technology
Along with popularizing of computer, the most work of people all will be finished on computers, and the informational needs of useful numeral manually is input in the computer on a lot of bills, as the date on the bill, the amount of money, voucher code etc.For the numeral on these bills, typing manually is power consuming time not only, and make mistakes easily, and therefore, the automatic identification of printing digital has very big researching value and application prospect on the bill.
Through prior art sum up is found that the printing digital recognition methods mainly contains two big classes: based on the method for architectural feature with based on the method for statistical nature.Use the classification of statistical nature to be easy to training, and on given training set, can obtain high recognition, yet outside training set, can't guarantee recognition effect.Use the method for architectural feature to utilize the fixed sturcture of character to extract feature, integrated structure information effectively in identifying has stronger adaptability with comparing based on the method for statistical nature.Tactic pattern identification can be divided at skeleton, profile, three kinds of methods of topography.
Digital recognition technology is widely used in every field in recent years, and digital recognition technology is ripe day by day, and the numeral identification of block letter has also reached very high recognition correct rate.But in actual applications, the method robustness of a lot of character recognition is relatively poor, can only be applicable to the identification of the printed character of single type, can descend much for its recognition accuracy of multiple font, and simultaneously, algorithm complex is higher.
Summary of the invention
Goal of the invention of the present invention is: provide a kind of digit recognition method that adopts both sides three quant's signs, to realize simple, the method for quickly identifying of printing digital to different fonts, different size, different sharpness.
The present invention carries out the following step to the monocase image that collects:
Step S1: character picture is carried out pre-service;
Step S2: extract the left envelope EL of character, right envelope ER, described left envelope EL be character leftmost side each point apart from the distance value of the left front scene vegetarian refreshments of character, right envelope ER is that character rightmost side each point is apart from the distance value of the right front scene vegetarian refreshments of character;
Step S3: based on the height H of character picture, by the head breadth WH of character, the wide WM of maximum waist, the wide WMO of minimum waist, the wide WF identification character of pin " 7 " or " 1 ",
Described WH be character picture (0, aH] in the zone, the breadth extreme of foreground pixel, wherein a ∈ [0.15,0.3];
Described WM and WMO are respectively in character picture [bH, the cH] zone, the breadth extreme of foreground pixel, minimum widith, b ∈ [04,0.55] wherein, c ∈ [0.65,0.75];
Described WF is in character picture [dH, the H] zone, the breadth extreme of foreground pixel, wherein c ∈ [0.75,0.85];
Step S4: based on detecting whether have upper left corner triangle white space and two belt-like zones that constitute the font style characteristic of character " 4 ", identification character " 4 " in the character picture;
Step S5: based on opening and the closing characteristics of the left and right sides envelope of character, Classification and Identification character " 0 ", " 2 ", " 3 ", " 5 ", " 6 ", " 8 ", " 9 ":
By last following certainly, left envelope first minimum point of getting character picture is L0, and the last minimum point of left envelope is L1, and right envelope first minimum point is R0, and the last minimum point of right envelope is R1; When L1≤L0, be left envelope closure; During R1≤R0, be right envelope closure; L1〉L0, be left envelope opening; R1〉during R0, be right envelope opening;
S501: if left and right envelope is all closed, then based on the pixel distribution of character, in conjunction with corresponding font style characteristic identification character based on corresponding font style characteristic identification character " 0 " or " 8 ";
S502: if left envelope opening, then based on corresponding font style characteristic identification character " 3 ";
S503: if left envelope opening and right envelope closure, then based on corresponding font style characteristic identification character " 0 ", " 3 ", " 8 " or " 9 ";
S504: if right envelope opening and left envelope closure, then based on corresponding font style characteristic identification character " 8 ", " 0 " or " 6 ";
S505: if the equal opening of left and right envelope, then based on corresponding font style characteristic identification character " 8 ", " 2 " or " 5 ".
In step S5 of the present invention, different openings and closed situation at the left and right sides envelope of character, when identifying kinds of characters based on corresponding font style characteristic, can adopt pixel distribution based on character, carry out the identification of kinds of characters in conjunction with the mode of corresponding font style characteristic, also can be other ways customary.
The invention has the beneficial effects as follows: compared with prior art, the present invention is based on the research to general character on most of printing digital form, extract common trait---both sides three quant's signs of the same character of different printing body, set up the model of character recognition, recognition accuracy is higher, strong robustness.
Description of drawings
The present invention will illustrate by example and with reference to the mode of accompanying drawing, wherein:
Fig. 1 is in the specific embodiment of the invention, the process flow diagram of numeral identification;
Fig. 2 is in the specific embodiment of the invention, image left and right sides envelope and three wide mark synoptic diagram;
Fig. 3 is in the specific embodiment of the invention, the upper left A test zone synoptic diagram of character " 4 ";
Fig. 4 is in the specific embodiment of the invention, opening parameter mark synoptic diagram.
Embodiment
Disclosed all features in this instructions, or the step in disclosed all methods or the process except mutually exclusive feature and/or step, all can make up by any way.
Disclosed arbitrary feature in this instructions (comprising any accessory claim, summary and accompanying drawing) is unless special narration all can be replaced by other equivalences or the alternative features with similar purpose.That is, unless special narration, each feature is an example in a series of equivalences or the similar characteristics.
Main thought of the present invention is: the resemblance discriminating digit that uses numeral, the width of the several key parts of the main index word of resemblance, left and right sides envelope forms etc. except 1,4,7, can clearly inerrably be differentiated other numerals according to the left and right sides envelope morphological feature of numeral.Method is as follows: ask for the left envelope EL of character, right envelope ER; Calculate head breadth WH, the wide WM of waist and WM0 and the wide WF of pin, namely both sides three are wide; Based on three quant's sign discriminating digits 7 and 1; 4 word features are comparatively special, detect by 3 subitems and finish judgement; Ask for left envelope opening parameter L 0, L1, right envelope opening parameters R 0, R1 are based on opening and the closing characteristics of left and right sides envelope, Classification and Identification numeral 2,3,5,6,8,9,0.This method has realized the accurately identification fast to the printing digital of different fonts, different size, different sharpness on the bill, has very high recognition correct rate and real-time in actual applications.
Referring to Fig. 1, specific implementation process of the present invention is, at first the monocase image (the monocase image normally obtains by the printing digital string image that collects is carried out Character segmentation) that collects is carried out pre-service, namely normalize to suitable dimension, the height of character picture is H, wide is W, and extracts character left and right sides envelope, carries out the identification of 0～90 numeral then.
At first extract character left and right sides envelope:
Referring to Fig. 2, in the present invention, left envelope EL is defined as character leftmost side each point apart from the distance value of the left front scene vegetarian refreshments of character, and right envelope ER is defined as character rightmost side each point apart from the distance value of the right front scene vegetarian refreshments of character.
Identification step (1), based on the three quant's sign discriminating digits 7 and 1 of character:
1. calculate head breadth WH, the wide WM of waist and WM0, the wide WF of pin, wherein,
WH be character picture (0, aH] in the zone, the breadth extreme of foreground pixel, wherein a ∈ [0.15,0.3];
WM and WMO are respectively in character picture [bH, the cH] zone, the breadth extreme of foreground pixel, minimum widith, b ∈ [04,0.55] wherein, c ∈ [0.65,0.75];
WF is in character picture [dH, the H] zone, the breadth extreme of foreground pixel, wherein c ∈ [0.75,0.85];
For example, head breadth WH of the present invention, the wide WM of waist and WM0, the wide WF of pin can ask for according to following formula:
WH=max(W-EL[i]-ER[i])；i=0,1,...,(0.3H-1)
WM=max(W-EL[i]-ER[i])；i=0.4H,...,0.7H
WM0=min(W-EL[i]-ER[i])；i=0.4H,...,0.7H
WF=max(W-EL[i]-ER[i])；i=0.8H,...,H-1
2. based on the WH that calculates, WM, WM0, WF value, identify the current character image:
If satisfy: WH ﹥ [0.55,0.65] W, WM 〉=[0.95,1.05] WF, (WH-WF) 〉=[0.2,0.3] WH, (WH-WM) 〉=[0.2,0.3] WM, WH ﹥ [1.95,2.05] WF; Judge that then current character is " 7 ", and finish identifying operation.
If satisfy: WH≤[0.55,0.65] W, (WM-WM0)≤[0.15,0.25] W, WM≤[0.4,0.5] W, WF 〉=[0.25,0.35] W judges that then current character is " 1 ", and finishes identifying operation.
If can not judge current character based on identification step (1), then enter identification step (2).
Identification step (2) by A, B and three grades of detections of C, namely detects whether have upper left corner triangle white space and two font style characteristics that constitute character " 4 " in the character picture, realizes the identification of character " 4 " is handled:
A detects: referring to Fig. 3, and the upper left corner of statistics character picture, the length of side is that pixel value is not 0 pixel summation SV, if SV≤T1 then detects upper left corner triangle white space, continues to carry out B and detects in the zone of isosceles right triangle Rt of Tran_w; Otherwise change next identification step over to, i.e. step (3);
Wherein the value of length of side Tran_w is [0.35,0.45] W; T1 is predetermined threshold value, and its value is the area of k1 times of isosceles right triangle Rt; The value of k1 is [0.0042,0.0156].
B detects: to the r row pixel of character picture, by last column of character picture scanning probe up, get the behavior C1 that first is not 0 pixel place, accumulative total (C1-1), (C1-2) and (C1-3) capable in the summation SHC of foreground pixel, if SHC 〉=T2, then detect article one belt-like zone, continue to carry out C and detect; Otherwise change identification step (3) over to;
Wherein, (C1-1) lastrow of expression C1, (C1-2), (C1-3) roughly the same, the value of the r of r row be [2,6], the value of predetermined threshold value T2 is for being [0.55,0.65] * 3W.
C detects: capable capable of the right envelope ER R1 that averages to c2 from the c1 of character picture, statistics (W-R1-4), (W-R1-3) and (W-R1-2) be listed as in the summation SVC of foreground pixel, if SVC 〉=T3, then detect the second belt-like zone, judge that current character is " 1 ", and finish identifying operation; Otherwise change identification step (3) over to;
Wherein the width of (W-R1-4) expression character picture deducts the value of R1 and 4, (W-R1-3), (W-R1-2) roughly the same, the value of predetermined threshold value T3 is [0.55,0.65] * 3H.
Identification step (3): based on opening and the closing characteristics of the left and right sides envelope of character, Classification and Identification character " 0 ", " 2 ", " 3 ", " 5 ", " 6 ", " 8 ", " 9 ":
At first, the opening parameter of calculating character image is judged left and right sides envelope opening and is closed and characteristic, as shown in Figure 4, is split shed parameter-definition figure of the present invention;
By last following certainly, left envelope first minimum point of getting character picture is L0, and the last minimum point of left envelope is L1, and right envelope first minimum point is R0, and the last minimum point of right envelope is R1; When L1≤L0, be left envelope closure; During R1≤R0, be right envelope closure; L1〉L0, be left envelope opening; R1〉during R0, be right envelope opening;
As L1〉during L0:
Calculate the left envelope area LM between opening L0～L1, if described LM≤0.1(H * W), left envelope aperture area grade LV=0 then; (H * W)＜LM≤[0.22,0.28] (H * W), then described LV=1 [if 0.08.1.02]; Otherwise, LV=2;
Calculate left envelope peak position LI, if described LI≤0.4H, then left peak grade LS=0; [if 0.38,0.42] H＜LI≤[0.68,0.72] H, then described LS=1; Otherwise LS=2;
As R1〉during R0:
Calculate the right envelope area RM between opening R0～R1, if RM≤0.1(H * W), left envelope aperture area grade RV=0; (H * W)＜RM≤[0.22,0.28] (H * W), then described RV=1 [if 0.08.1.02]; Otherwise, RV=2;
Calculate right envelope peak position RI, if RI≤0.4H, then left peak grade RS=0; [if 0.38,0.42] H＜RI≤[0.68,0.72] H, described then RS=1; Otherwise RS=2;
Based on above-mentioned parameter L0, L1, R0, R1, LV, LS, RV, RS, realize that the detailed process of character recognition is:
1. when L1≤L0 and R1≤R0:
Extracting the feature characteristic parameter of character " 0 " or " 8 ", is " 0 " or " 8 " with the identification current character:
Hole heart height H 0, the H1 in the upper and lower hole on detection and the definite character picture perpendicular bisector, and diameter D0, the D1 in upper and lower hole, judge whether the hole heart is less than or equal to [0.45 apart from ︱ H0-H1 ︱, 0.55] * (D0+D1), if then current character possesses the character property of the character property " 8 " of " 8 "; Otherwise be the character property of " 0 ".
2. as L1〉during L0:
At first from (L0-3) OK, to scope of (L1+3) row, calculate left envelope difference, find out first and last difference absolute value and be designated as i and j respectively greater than the point of [0.16,0.24] W; If less than [0.46,0.54] H, some j less than [0.56,0.64] H, judges then that current character is " 3 " apart from the distance of character picture top margin to i point, and the end identifying operation apart from the distance of character picture top margin; Otherwise carry out next identification again and judge (be L1〉L0 and R1≤R0 situation);
3. as L1〉when L0 and R1≤R0:
If LV＜1, then at first based on the character property of " 8 ", whether the identification current character is " 8 ", if, judge that then current character is " 8 ", finish identifying operation; Otherwise by the value of LV and LS, based on bivariate table I, obtain recognition result, finish identification and handle, described bivariate table I is:
4. as L1≤L0 and R1〉during R0:
If RV＜1, at first based on the character property of " 8 ", whether the identification current character is " 8 ", if then finish identification and handle; Otherwise by the value of RV, RS, based on two-dimentional Table II, obtain recognition result, finish identification and handle, described two-dimentional Table II is:
5. as L1〉L0 and R1〉during R0:
At first extract the feature characteristic parameter of character " 8 ", whether the identification current character is " 8 ", if then finish identification and handle;
If can not identify current character is " 8 ", then proceed following two condition judgment:
(i) judge whether to satisfy: LI＜RI and RI 0.6H, then by the value of RV, LV, based on two-dimentional Table III, obtain recognition result, finish identification and handle, described two-dimentional Table III is:
In the two dimension Table III, " 1 " expression current character is nonnumeric character;
(ii) judge whether to satisfy LI〉RI, and RI＜0.5H, judge that then current character is " 5 ", finish identifying operation.
The present invention is not limited to aforesaid embodiment.The present invention expands to any new feature or any new combination that discloses in this manual, and the arbitrary new method that discloses or step or any new combination of process.
Claims (6)
1. a digit recognition method that adopts both sides three quant's signs is characterized in that, the monocase image that collects is carried out the following step:
Step S1: character picture is carried out pre-service;
Step S2: extract the left envelope EL of character, right envelope ER, described left envelope EL be character leftmost side each point apart from the distance value of the left front scene vegetarian refreshments of character, right envelope ER is that character rightmost side each point is apart from the distance value of the right front scene vegetarian refreshments of character;
Step S3: based on the height H of character picture, by the head breadth WH of character, the wide WM of maximum waist, the wide WMO of minimum waist and the wide WF identification character of pin " 7 " or " 1 ",
Described WH be character picture (0, aH] in the zone, the breadth extreme of foreground pixel, wherein a ∈ [0.15,0.3];
Described WM and WMO are respectively in character picture [bH, the cH] zone, the breadth extreme of foreground pixel, minimum widith, b ∈ [04,0.55] wherein, c ∈ [0.65,0.75];
Described WF is in character picture [dH, the H] zone, the breadth extreme of foreground pixel, wherein c ∈ [0.75,0.85];
Step S4: based on detecting whether have upper left corner triangle white space and two belt-like zones that constitute the font style characteristic of character " 4 ", identification character " 4 " in the character picture;
Step S5: based on opening and the closing characteristics of the left and right sides envelope of character, Classification and Identification character " 0 ", " 2 ", " 3 ", " 5 ", " 6 ", " 8 ", " 9 ":
By last following certainly, left envelope first minimum point of getting character picture is L0, and the last minimum point of left envelope is L1, and right envelope first minimum point is R0, and the last minimum point of right envelope is R1; When L1≤L0, be left envelope closure; During R1≤R0, be right envelope closure; L1〉L0, be left envelope opening; R1〉during R0, be right envelope opening;
S501: if left and right envelope is all closed, then based on corresponding font style characteristic identification character " 0 " or " 8 ";
S502: if left envelope opening, then based on corresponding font style characteristic identification character " 3 ";
S503: if left envelope opening and right envelope closure, then based on corresponding font style characteristic identification character " 0 ", " 3 ", " 8 " or " 9 ";
S504: if right envelope opening and left envelope closure, then based on corresponding font style characteristic identification character " 8 ", " 0 " or " 6 ";
S505: if the equal opening of left and right envelope, then based on corresponding font style characteristic identification character " 8 ", " 2 " or " 5 ".
2. the method for claim 1 is characterized in that, among the described step S3,
If satisfy: WH ﹥ [0.55,0.65] W, WM 〉=[0.95,1.05] WF, (WH-WF) 〉=[0.2,0.3] WH, (WH-WM) 〉=[0.2,0.3] WM, WH ﹥ [1.95,2.05] WF, then current character is " 7 ";
If satisfy: WH≤[0.55,0.65] W, (WM-WM0)≤[0.15,0.25] W, WM≤[0.4,0.5] W, WF 〉=[0.25,0.35] W, then current character is " 1 ";
Wherein W is the width of character picture.
3. the method for claim 1 is characterized in that, among the described step S4, detects whether have upper left corner triangle white space and two belt-like zones that constitute the font style characteristic of character " 4 " in the character picture, is specially:
S301: the upper left corner of statistics character picture, the length of side are that pixel value is not 0 pixel summation SV in the zone of isosceles right triangle Rt of Tran_w, if described SV is less than or equal to threshold value T1, then detect upper left corner triangle white space;
The value of wherein said length of side Tran_w is [0.35,0.45] W, and threshold value T1 is the area of k1 times of isosceles right triangle Rt, and the value of described k1 is [0.0042,0.0156];
S302: to the r row pixel of character picture, by last column of character picture scanning probe up, get the behavior C1 that first is not 0 pixel place, accumulative total (C1-1), (C1-2) and (C1-3) capable in the summation SHC of foreground pixel, if described SHC more than or equal to threshold value T2, then detects article one belt-like zone;
The value of described r is [2,6], and the value of threshold value T2 is [0.55,0.65] * 3W;
S303: capable capable of the right envelope ER R1 that averages to c2 from the c1 of character picture, statistics (W-R1-4), (W-R1-3) and (W-R1-2) the summation SVC of foreground pixel in the row more than or equal to threshold value T3, then detect the second belt-like zone as if SVC;
Wherein said W is the width of character picture, and the value of threshold value T3 is [0.55,0.65] * 3H.
4. the method for claim 1 is characterized in that, among the described step S501, to the concrete identification step of character " 0 " and " 8 " is:
Hole heart height H 0, the H1 in the upper and lower hole on detection and the definite character picture perpendicular bisector, and diameter D0, the D1 in upper and lower hole, judge whether the hole heart is less than or equal to [0.45 apart from ︱ H0-H1 ︱, 0.55] * (D0+D1), if then current character possesses the character property of the character property " 8 " of " 8 "; Otherwise be the character property of " 0 ".
5. method as claimed in claim 4 is characterized in that, among the described step S502, is specially based on corresponding font style characteristic identification character " 3 ":
From (L0-3) OK, to scope of (L1+3) row, calculate left envelope difference, find out first and last difference absolute value and be designated as i and j respectively greater than the point of [0.16,0.24] W;
If i point apart from the distance of character picture top margin less than [0.46,0.54] H, put j apart from the distance of character picture top margin less than [0.56,0.64] H, then current character is " 3 ".
6. method as claimed in claim 5 is characterized in that, among the described step S5:
As L1〉during L0:
Calculate the left envelope area LM between opening L0～L1, if described LM≤0.1(H * W), left envelope aperture area grade LV=0 then; (H * W)＜LM≤[0.22,0.28] (H * W), then described LV=1 [if 0.08.1.02]; Otherwise, LV=2;
Calculate left envelope peak position LI, if described LI≤0.4H, then left peak grade LS=0; [if 0.38,0.42] H＜LI≤[0.68,0.72] H, then described LS=1; Otherwise LS=2;
As R1〉during R0:
Calculate the right envelope area RM between opening R0～R1, if RM≤0.1(H * W), left envelope aperture area grade RV=0;
(H * W)＜RM≤[0.22,0.28] (H * W), then described RV=1 [if 0.08.1.02]; Otherwise, RV=2;
Calculate right envelope peak position RI, if RI≤0.4H, then left peak grade RS=0; If
[0.38,0.42] H＜RI≤[0.68,0.72] H, described then RS=1; Otherwise RS=2;
Then described step S503-S505 is specially;
S503: if LV＜1, then at first based on the character property of " 8 ", whether the identification current character is " 8 ", if, judge that then current character is " 8 ", finish identifying operation; Otherwise by the value of LV and LS, based on bivariate table I, obtain recognition result, finish identification and handle, described bivariate table I is:
S504: if RV＜1, at first based on the character property of " 8 ", whether the identification current character is " 8 ", if then finish identification and handle; Otherwise by described RV, RS value, based on two-dimentional Table II, obtain recognition result, finish identification and handle, described two-dimentional Table II is:
S505: at first based on the character property of " 8 ", whether the identification current character is " 8 ", if then finish identification and handle;
Otherwise continue to judge whether to satisfy: LI＜RI and RI 0.6H, then by the value of described RV, LV, based on two-dimentional Table III, obtain recognition result, finish identification and handle, described two-dimentional Table III is:
In the two dimension Table III, " 1 " expression current character is nonnumeric character;
If LI〉RI, and RI＜0.5H, judge that then current character is " 5 ", finish identifying operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310209889.XA CN103268490B (en) | 2013-05-30 | 2013-05-30 | A kind of digit recognition method adopting both sides three quant's sign |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310209889.XA CN103268490B (en) | 2013-05-30 | 2013-05-30 | A kind of digit recognition method adopting both sides three quant's sign |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103268490A true CN103268490A (en) | 2013-08-28 |
| CN103268490B CN103268490B (en) | 2016-01-13 |
Family
ID=49012117
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310209889.XA Expired - Fee Related CN103268490B (en) | 2013-05-30 | 2013-05-30 | A kind of digit recognition method adopting both sides three quant's sign |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103268490B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106257495A (en) * | 2015-06-19 | 2016-12-28 | 阿里巴巴集团控股有限公司 | A kind of digit recognition method and device |
| CN108734680A (en) * | 2018-05-29 | 2018-11-02 | 电子科技大学 | A kind of image repair method of adaptively selected sample block size |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030190077A1 (en) * | 1999-03-19 | 2003-10-09 | Raf Technology, Inc. | Rollup functions and methods |
| CN101916364A (en) * | 2010-07-06 | 2010-12-15 | 浙江大学 | Adaptive dial digital identification device and method |
| CN102629322A (en) * | 2012-03-12 | 2012-08-08 | 华中科技大学 | Character feature extraction method based on stroke shape of boundary point and application thereof |
-
2013
- 2013-05-30 CN CN201310209889.XA patent/CN103268490B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030190077A1 (en) * | 1999-03-19 | 2003-10-09 | Raf Technology, Inc. | Rollup functions and methods |
| CN101916364A (en) * | 2010-07-06 | 2010-12-15 | 浙江大学 | Adaptive dial digital identification device and method |
| CN102629322A (en) * | 2012-03-12 | 2012-08-08 | 华中科技大学 | Character feature extraction method based on stroke shape of boundary point and application thereof |
Non-Patent Citations (2)
| Title |
|---|
| 耿西伟等: "基于结构特征分类BP网络的手写数字识别", 《计算机技术与发展》, vol. 17, no. 1, 31 January 2007 (2007-01-31) * |
| 马斌等: "特征信息的新数字图像识别方法", 《沈阳建筑大学学报（自然科学版）》, vol. 27, no. 6, 30 November 2011 (2011-11-30) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106257495A (en) * | 2015-06-19 | 2016-12-28 | 阿里巴巴集团控股有限公司 | A kind of digit recognition method and device |
| CN108734680A (en) * | 2018-05-29 | 2018-11-02 | 电子科技大学 | A kind of image repair method of adaptively selected sample block size |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103268490B (en) | 2016-01-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103049750B (en) | Character identifying method | |
| Rodriguez et al. | Local gradient histogram features for word spotting in unconstrained handwritten documents | |
| Pal et al. | Segmentation of Bangla unconstrained handwritten text | |
| CN100533466C (en) | Font recognizing method and system | |
| CN101957919B (en) | Character recognition method based on image local feature retrieval | |
| CN102298606B (en) | Random walking image automatic annotation method and device based on label graph model | |
| CN103824091B (en) | A kind of licence plate recognition method for intelligent transportation system | |
| TW201636900A (en) | Method and apparatus for generating text line classifier | |
| CN104809481A (en) | Natural scene text detection method based on adaptive color clustering | |
| Alvaro et al. | A shape-based layout descriptor for classifying spatial relationships in handwritten math | |
| CN104123531B (en) | A kind of method and apparatus of palmmprint interest zone location | |
| CN102663382A (en) | Video image character recognition method based on submesh characteristic adaptive weighting | |
| CN105930836A (en) | Identification method and device of video text | |
| Tian et al. | Natural scene text detection with MC–MR candidate extraction and coarse-to-fine filtering | |
| CN104751148B (en) | A kind of method of cognitive science formula in layout files | |
| CN108664474A (en) | A kind of resume analytic method based on deep learning | |
| CN106127265B (en) | A kind of text in picture identification error correction method based on activating force model | |
| CN104484645A (en) | Human-computer interaction-oriented '1' gesture-recognition method and system | |
| Belaïd et al. | Handwritten and printed text separation in real document | |
| CN110738092B (en) | Invoice text detection method | |
| Patel et al. | Gujarati handwritten character recognition using hybrid method based on binary tree-classifier and k-nearest neighbour | |
| CN105117704B (en) | A kind of text image consistency comparison method based on multiple features | |
| CN104992454A (en) | Regionalized automatic-cluster-change image segmentation method | |
| Adak et al. | Named entity recognition from unstructured handwritten document images | |
| CN105224945A (en) | A kind of automobile logo identification method based on joint-detection and identification algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160113 Termination date: 20190530 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |