CN102930277A - Character picture verification code identifying method based on identification feedback - Google Patents
Character picture verification code identifying method based on identification feedback Download PDFInfo
- Publication number
- CN102930277A CN102930277A CN201210349375XA CN201210349375A CN102930277A CN 102930277 A CN102930277 A CN 102930277A CN 201210349375X A CN201210349375X A CN 201210349375XA CN 201210349375 A CN201210349375 A CN 201210349375A CN 102930277 A CN102930277 A CN 102930277A
- Authority
- CN
- China
- Prior art keywords
- image
- character
- point
- pixel
- 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)
Abstract
The invention provides a character picture verification code identifying method based on identification feedback. The character picture verification code identifying method is characterized by comprising the following steps of: firstly, converting an original colorful image into a gray level image and carrying out binaryzation treatment to obtain binaryzation image data; then, repairing a binaryzation image, removing image edge burrs and filling a central blank; finishing the connection of broken strokes to obtain a repaired image; removing a background and an interference point or line of the repaired image to obtain a noise-free image; thinning the noise-free image to obtain a thinned image with a single pixel; then, cutting the thinned image to obtain a single-character image which only contains a single character; and finally, carrying out normalization on the single-character image and identifying the character. According to the character picture verification code identifying method based on the identification feedback, a corresponding algorithm is designed to process a character picture identification code including various types of noises and having the phenomena that the character is stuck, inclined, rotated, deformed and the like, so as to finish high-efficiency and high-precision identification. The character picture verification code identifying method based on the identification feedback can be applied to verification code identification in an automatic program.
Description
Technical field
The present invention relates to a kind of method for recognizing verification code, particularly relate to a kind of image authentication code recognition methods based on Recognition feedback.
Background technology
The basic norm of identifying code design is to reduce the problem of hard artificial intelligence, if i.e. not general-purpose computers program solution of problem then can be used as identifying code.The implementation of identifying code comprises character picture, speech recognition, dynamic image etc. at present, and wherein character picture is because the advantages such as volume is little, easy transmission are the most widely used a kind of.Use identifying code can prevent other people to the website carry out batch registration, repeat to post, the mass-sending of violence password cracking, spam, use software that the automatic test of website, the release quickly of information, the robotization of transaction such as are carried out at the operation yet identifying code has also hindered the tester.
Many decades in the past, researchers have proposed a lot of method for recognizing verification codes, and these methods can be divided into: based on the method for template matches, based on method and the machine learning method of charcter topology.
Template matching method adopts the character in the pre-designed template matches image, and is relatively simple, implement easily, yet this method can't be processed and has noise, character position and big or small unfixed identifying code.
Based on the method for charcter topology according to character design feature separately as recognition feature, self-defined different recognizer, however can't the processing character adhesion, the situation such as distortion.
Machine learning method is by extracting feature, and training classifier can be realized the efficient identification of character.
The subject matter of character picture identifying code identification is to include various types of noises in the identifying code, and character exists adhesion, inclination, rotation, distortion, and these phenomenons have affected the accuracy rate of identification.
Summary of the invention
The objective of the invention is to provide a kind of character picture method for recognizing verification code based on Recognition feedback, overcome the method for recognizing verification code in the automated procedures of defective there are the None-identifieds such as noise, Characters Stuck, inclination, rotation, distortion in to(for) complex characters image authentication code, by the identification to the character picture identifying code, realize the smooth work of automated procedures, improve people's work efficiency.
The present invention is achieved in that the operation steps of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention is as follows to achieve the above object:
(1) the character picture is loaded in the internal memory, obtains the color of each pixel, color adopts the RGB form to represent, represents respectively redness, green, the blue component of this color;
(2) binaryzation
Convert coloured image to gray level image, and adopt following greyscale transformation formula,
Y=0.299R+0.587G+0.114B
And adopt threshold value that greyscale image transitions is become bianry image, choosing of threshold value adopted large Tianjin method and is OSTU, namely selects so that the gray-scale value of variance maximum is as threshold value between two class samples, and the choosing method of threshold value is in the binaryzation:
If
w 0 Be the ratio of the total pixel of the shared image of foreground pixel, order
u 0 For the average gray of all foreground pixels, establish
w 1 Be the count ratio of the total pixel of shared image of background, order
u 1 Be the average gray of all background pixels, then the average gray of all pixels of image is
u=
w 0 u 0 +
w 1 u 1 . when carrying out sequential operation,
tValue can the maximum gradation value from the minimum gradation value of image to image travel through successively, when
tWhen getting certain value, the inter-class variance formula
b=
w 0 (
u 0 -
u)
2 +
w 1 (
u 1 -
u)
2 Can obtain maximum, at this moment
tBe the threshold value of binaryzation;
(3) repair
Adopt following template that image is repaired, to remove burrs on edges and to plug a gap
0?0?0 0?0 0?0?0 1 1
0?1?0 0?1?1 0?1?1 1?0?1 1?0?1
0?0?0 0?0?1 0?0 1 0?0?1
Template T1 template T2 template T3 template T4 template T5
Wherein 0 represent the background pixel point, 1 represents the foreground pixel point;? can be background or prospect picture element, template T1-T3 becomes the background pixel point with center foreground pixel point, template T4 and T5 become the foreground pixel point with center background pixel point, and each template successively dextrorotation turn 90 degrees, 180 the degree and 270 the degree, form new template, and act on successively original image;
(4) denoising sound
Adopt corresponding eliminating noise method for different noise types, comprising:
Connected domain filtering: adjacent foreground pixel point is expanded, obtained connected domain, remove the connected domain area less than the zone of certain critical value,
Gaussian filtering: establish
F (i, j)Presentation video (i, j) is located the gray scale of pixel, through obtaining filtered image such as down conversion:
Super curve filtering: the continuous smooth long curve that exists in the detected image also filters, adopt dijkstra's algorithm, and for the Di Jiesitela algorithm is sought the shortest path between any two summits among the figure, thereby determine curve, select length greater than the curve of picture traverse 80% as super curve, and remove vertical pixel run length on the curve less than the pixel of average stroke width, thereby remove super curve;
(5) refinement
Lines in the image are peeled off from the edge to the center layer by layer, only contained the image of the wide lines of single pixel, algorithmic procedure is:
1) 8 neighborhoods of consideration centered by frontier point, the note central point is p1,8 points of its neighborhood are designated as respectively p2 around central point clockwise, p3.., p9, at first mark satisfies the frontier point of following condition simultaneously:
a)1<?N(?p1)<7
b)S(?p1)?=?1
c)p2?*?p4!*?p6=?0
d)p4?*?p6?*p8=?0
Wherein N (p1) is the number of the non-zero adjoint point of p1, and S (p1) is with p2, p3.., the number of times that the value of these points from 0 to 1 changed when p9 was order, when all frontier points are all checked complete after, with all marks point remove;
2) same step 1) is only with front condition c) P4 * p8=0; Condition d) change condition p2* p6*p8=0 into, equally when all frontier points are all checked complete after, with all marks point remove, more than the operation of two steps consist of an iteration, until point does not satisfy flag condition again, then algorithm termination;
(6) Character segmentation
Image after the refinement is carried out cutting, so that each subgraph only contains single character, then candidate's cut-off after at first definite refinement in the image adopts the dynamic programming algorithm searching from an optimal path of origin-to-destination, and the criterion of optimization is the probability sum of character recognition in all subgraphs;
(7) identification
To the character after cutting apart, adopt identification module to carry out character recognition, identification module adopts support vector machine as sorter;
(8) sorter
Select some character pictures as training sample, carry out vectorization by feature extraction, adopt support vector machine to make up sorter, the feature of wherein extracting comprises image:
Thick meshed feature: picture is divided into the 4*4 grid, adds up and deceive the number percent that pixel accounts for whole sub-grid in each grid, obtain 16 dimensional features;
Hand over and cut feature: at horizontal and vertical set direction 10 pixels, 20 pixels, three lines of 30 pixels, the intersection point number of statistics picture prospect and these lines obtains 6 dimensional features;
Framework characteristic: the quantity of end points, triradius, four crunodes in the statistics picture foreground point obtains 3 dimensional features;
Projection properties: picture is divided into 4 zones, and these 4 zones are comprised of 12 limits, and the projection number of statistics picture foreground point on these limits obtains 12 dimensional features;
Position, first foreground point: along level, vertical, eight the direction ecto-entads that tilt, add up the coordinate of position, first foreground point, obtain 8 dimensional features;
Peripheral characteristic: picture according to ranks 4 five equilibriums, successively along the scanning of four direction ecto-entad, is recorded the area of the non-character part that forms when every row runs into the foreground point pixel for the first time, obtain 4 dimensional features, along four direction scanning, obtain altogether 16 dimensional features;
The Zernike moment characteristics: the repetition rate of calculating picture is the n rank Zernike square of m:
Get Z00-Z33, totally 16 dimensions are as feature.
The invention provides a kind of character picture method for recognizing verification code based on Recognition feedback, mainly comprise step: convert original color image to gray level image first, and carry out binary conversion treatment, obtain the binary image data; Then described binary image is repaired, removed the image border burr, fill up central space, and finish the connection of fracture stroke, obtain repairing image; Again described repairing image is removed background and noise spot or line, obtain noise-free picture; And noise-free picture carried out refinement, obtain the refined image of single pixel; Then refined image is carried out cutting, only contained the monocase image of single character; At last the monocase image is carried out normalization and identification character wherein.
A kind of character picture method for recognizing verification code based on Recognition feedback of the present invention has following features:
1, a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention is carrying out binaryzation and is adopting the OSTU method dynamically to determine.
2, a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention is when removing picture noise, and denoising method can be removed ground unrest, Gaussian noise, curve noise.
3, a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention is when identification, and identification module is oneself to train out.
4, a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention also comprises repairing, refinement, 3 steps of cutting, in the especially cutting step cutting and identification is merged, and can guarantee to find globally optimal solution in conjunction with dynamic programming algorithm.
Generally speaking, the present invention can exist the character picture identifying code of the phenomenons such as adhesion, inclination, rotation and distortion to design corresponding algorithm to process, finish high-level efficiency and high-precision identification to containing various types of noises, character.
Description of drawings
Concrete structure of the present invention is provided by following embodiment and accompanying drawing thereof.
Fig. 1 is the schematic flow sheet of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 2 is the identifying code image synoptic diagram of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 3 is the identifying code image synoptic diagram after the binary conversion treatment of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 4 is the identifying code image synoptic diagram after the repairing of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 5 is the identifying code image synoptic diagram after the denoising of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 6 is the identifying code image synoptic diagram after the refinement of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Fig. 7 is the identifying code image synoptic diagram after the Character segmentation of a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention.
Embodiment
Below with reference to accompanying drawing a kind of character picture method for recognizing verification code based on Recognition feedback of the present invention is described in further detail.
With reference to Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6 and Fig. 7, a kind of character picture method for recognizing verification code based on Recognition feedback of this invention comprises following operation steps:
(1) the character picture is loaded in the internal memory, obtains the color of each pixel, color adopts the RGB form to represent that represent respectively redness, green, the blue component of this color, image as shown in Figure 2;
(2) binaryzation
Convert coloured image to gray level image, and adopt following greyscale transformation formula,
Y=0.299R+0.587G+0.114B
Adopt threshold value that greyscale image transitions is become bianry image, choosing of threshold value adopted large Tianjin method and is OSTU, namely selects so that the gray-scale value of variance maximum is as threshold value between two class samples, and the choosing method of threshold value is in the binaryzation:
If
w 0 Be the ratio of the total pixel of the shared image of foreground pixel, order
u 0 For the average gray of all foreground pixels, establish
w 1 Be the count ratio of the total pixel of shared image of background, order
u 1 Be the average gray of all background pixels, then the average gray of all pixels of image is
u=
w 0 u 0 +
w 1 u 1 . when carrying out sequential operation,
tValue can the maximum gradation value from the minimum gradation value of image to image travel through successively, when
tWhen getting certain value, the inter-class variance formula
b=
w 0 (
u 0 -
u)
2 +
w 1 (
u 1 -
u)
2 Can obtain maximum, at this moment
tBe the threshold value of binaryzation, the image after the binaryzation as shown in Figure 3;
(3) repair
Adopt following template that image is repaired, to remove burrs on edges and to plug a gap
0?0?0 0?0 0?0?0 1 1
0?1?0 0?1?1 0?1?1 1?0?1 1?0?1
0?0?0 0?0?1 0?0 1 0?0?1
Template T1 template T2 template T3 template T4 template T5
Wherein 0 represent the background pixel point, 1 represents the foreground pixel point;? can be background or prospect picture element, template T1-T3 becomes the background pixel point with center foreground pixel point, template T4 and T5 become the foreground pixel point with center background pixel point, and each template successively dextrorotation turn 90 degrees, 180 the degree and 270 the degree, form new template, and acting on successively original image, the image after the repairing is as shown in Figure 4;
(4) denoising sound
Adopt corresponding eliminating noise method for different noise types, comprising:
Connected domain filtering: adjacent foreground pixel point is expanded, obtained connected domain, remove the connected domain area less than the zone of certain critical value,
Gaussian filtering: establish
F (i, j)Presentation video (i, j) is located the gray scale of pixel, through obtaining filtered image such as down conversion:
Super curve filtering: the continuous smooth long curve that exists in the detected image also filters, adopt Dijkstra (Di Jiesitela) algorithm to seek the shortest path between any two summits among the figure, thereby determine curve, select length greater than the curve of picture traverse 80% as super curve, and remove vertical pixel run length on the curve less than the pixel of average stroke width, thereby remove super curve, the image after the denoising as shown in Figure 5;
(5) refinement
Lines in the image are peeled off from the edge to the center layer by layer, only contained the image of the wide lines of single pixel, algorithmic procedure is:
1) 8 neighborhoods of consideration centered by frontier point, the note central point is p1,8 points of its neighborhood are designated as respectively p2 around central point clockwise, p3.., p9, at first mark satisfies the frontier point of following condition simultaneously:
a)1<?N(?p1)<7
b)S(?p1)?=?1
c)p2?*?p4!*?p6=?0
d)p4?*?p6?*p8=?0
Wherein N (p1) is the number of the non-zero adjoint point of p1, and S (p1) is with p2, p3.., the number of times that the value of these points from 0 to 1 changed when p9 was order, when all frontier points are all checked complete after, with all marks point remove;
2) same step 1) is only with front condition c) P4 * p8=0; Condition d) change condition p2* p6*p8=0 into, equally when all frontier points are all checked complete after, with all marks point remove, more than two step operations consist of an iteration, until point does not satisfy flag condition again, then algorithm stops, and the image after the refinement as shown in Figure 6;
(6) Character segmentation
Image after the refinement is carried out cutting, so that each subgraph only contains single character, candidate's cut-off after at first definite refinement in the image, then adopt the dynamic programming algorithm searching from an optimal path of origin-to-destination, the criterion of optimizing is the probability sum of character recognition in all subgraphs, and the image after cutting apart as shown in Figure 7;
(7) identification
To the character after cutting apart, adopt identification module to carry out character recognition, identification module adopts support vector machine as sorter;
(8) sorter
Select some character pictures as training sample, carry out vectorization by feature extraction, adopt support vector machine to make up sorter, the feature of wherein extracting comprises image:
Thick meshed feature: picture is divided into the 4*4 grid, adds up and deceive the number percent that pixel accounts for whole sub-grid in each grid, obtain 16 dimensional features;
Hand over and cut feature: at horizontal and vertical set direction 10 pixels, 20 pixels, three lines of 30 pixels, the intersection point number of statistics picture prospect and these lines obtains 6 dimensional features;
Framework characteristic: the quantity of end points, triradius, four crunodes in the statistics picture foreground point obtains 3 dimensional features;
Projection properties: picture is divided into 4 zones, and these 4 zones are comprised of 12 limits, and the projection number of statistics picture foreground point on these limits obtains 12 dimensional features;
Position, first foreground point: along level, vertical, eight the direction ecto-entads that tilt, add up the coordinate of position, first foreground point, obtain 8 dimensional features;
Peripheral characteristic: picture according to ranks 4 five equilibriums, successively along the scanning of four direction ecto-entad, is recorded the area of the non-character part that forms when every row runs into the foreground point pixel for the first time, obtain 4 dimensional features, along four direction scanning, obtain altogether 16 dimensional features;
The Zernike moment characteristics: the repetition rate of calculating picture is the n rank Zernike square of m:
Get Z00-Z33, totally 16 dimensions are as feature.
Claims (1)
1. character picture method for recognizing verification code based on Recognition feedback is characterized in that the operation steps of the method is as follows:
(1) the character picture is loaded in the internal memory, obtains the color of each pixel, color adopts the RGB form to represent, represents respectively redness, green, the blue component of this color;
(2) binaryzation
Convert coloured image to gray level image, and adopt following greyscale transformation formula,
Y=0.299R+0.587G+0.114B
Adopt threshold value that greyscale image transitions is become bianry image, choosing of threshold value adopted large Tianjin method and is OSTU, namely selects so that the gray-scale value of variance maximum is as threshold value between two class samples, and the choosing method of threshold value is in the binaryzation:
If
w 0 Be the ratio of the total pixel of the shared image of foreground pixel, order
u 0 For the average gray of all foreground pixels, establish
w 1 Be the count ratio of the total pixel of shared image of background, order
u 1 Be the average gray of all background pixels, then the average gray of all pixels of image is
u=
w 0 u 0 +
w 1 u 1 . when carrying out sequential operation,
tValue can the maximum gradation value from the minimum gradation value of image to image travel through successively, when
tWhen getting certain value, the inter-class variance formula
b=
w 0 (
u 0 -
u)
2 +
w 1 (
u 1 -
u)
2 Can obtain maximum, at this moment
tBe the threshold value of binaryzation;
(3) repair
Adopt following template that image is repaired, to remove burrs on edges and to plug a gap
0?0?0 0?0 0?0?0 1 1
0?1?0 0?1?1 0?1?1 1?0?1 1?0?1
0?0?0 0?0?1 0?0 1 0?0?1
Template T1 template T2 template T3 template T4 template T5
Wherein 0 represent the background pixel point, 1 represents the foreground pixel point;? can be background or prospect picture element, template T1-T3 becomes the background pixel point with center foreground pixel point, template T4 and T5 become the foreground pixel point with center background pixel point, and each template successively dextrorotation turn 90 degrees, 180 the degree and 270 the degree, form new template, and act on successively original image;
(4) denoising sound
Adopt corresponding eliminating noise method for different noise types, comprising:
Connected domain filtering: adjacent foreground pixel point is expanded, obtained connected domain, remove the connected domain area less than the zone of certain critical value,
Gaussian filtering: establish
F (i, j)Presentation video (i, j) is located the gray scale of pixel, through obtaining filtered image such as down conversion:
Super curve filtering: the continuous smooth long curve that exists in the detected image also filters, adopt dijkstra's algorithm and be the shortest path between any two summits among the Di Jiesitela algorithm searching figure, thereby determine curve, select length greater than the curve of picture traverse 80% as super curve, and remove vertical pixel run length on the curve less than the pixel of average stroke width, thereby remove super curve;
(5) refinement
Lines in the image are peeled off from the edge to the center layer by layer, only contained the image of the wide lines of single pixel, algorithmic procedure is:
1) 8 neighborhoods of consideration centered by frontier point, the note central point is p1,8 points of its neighborhood are designated as respectively p2 around central point clockwise, p3.., p9, at first mark satisfies the frontier point of following condition simultaneously:
a)1<?N(?p1)<7
b)S(?p1)?=?1
c)p2?*?p4!*?p6=?0
d)p4?*?p6?*p8=?0
Wherein N (p1) is the number of the non-zero adjoint point of p1, and S (p1) is with p2, p3.., the number of times that the value of these points from 0 to 1 changed when p9 was order, when all frontier points are all checked complete after, with all marks point remove;
2) same step 1) is only with front condition c) P4 * p8=0; Condition d) change condition p2* p6*p8=0 into, equally when all frontier points are all checked complete after, with all marks point remove, more than the operation of two steps consist of an iteration, until point does not satisfy flag condition again, then algorithm termination;
(6) Character segmentation
Image after the refinement is carried out cutting, so that each subgraph only contains single character, then candidate's cut-off after at first definite refinement in the image adopts the dynamic programming algorithm searching from an optimal path of origin-to-destination, and the criterion of optimization is the probability sum of character recognition in all subgraphs;
(7) identification
To the character after cutting apart, adopt identification module to carry out character recognition, identification module adopts support vector machine as sorter;
(8) sorter
Select some character pictures as training sample, carry out vectorization by feature extraction, adopt support vector machine to make up sorter, the feature of wherein extracting comprises image:
Thick meshed feature: picture is divided into the 4*4 grid, adds up and deceive the number percent that pixel accounts for whole sub-grid in each grid, obtain 16 dimensional features;
Hand over and cut feature: at horizontal and vertical set direction 10 pixels, 20 pixels, three lines of 30 pixels, the intersection point number of statistics picture prospect and these lines obtains 6 dimensional features;
Framework characteristic: the quantity of end points, triradius, four crunodes in the statistics picture foreground point obtains 3 dimensional features;
Projection properties: picture is divided into 4 zones, and these 4 zones are comprised of 12 limits, and the projection number of statistics picture foreground point on these limits obtains 12 dimensional features;
Position, first foreground point: along level, vertical, eight the direction ecto-entads that tilt, add up the coordinate of position, first foreground point, obtain 8 dimensional features;
Peripheral characteristic: picture according to ranks 4 five equilibriums, successively along the scanning of four direction ecto-entad, is recorded the area of the non-character part that forms when every row runs into the foreground point pixel for the first time, obtain 4 dimensional features, along four direction scanning, obtain altogether 16 dimensional features;
The Zernike moment characteristics: the repetition rate of calculating picture is the n rank Zernike square of m:
Get Z00-Z33, totally 16 dimensions are as feature.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210349375.XA CN102930277B (en) | 2012-09-19 | 2012-09-19 | A kind of character picture method for recognizing verification code based on Recognition feedback |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210349375.XA CN102930277B (en) | 2012-09-19 | 2012-09-19 | A kind of character picture method for recognizing verification code based on Recognition feedback |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102930277A true CN102930277A (en) | 2013-02-13 |
CN102930277B CN102930277B (en) | 2016-04-27 |
Family
ID=47645074
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210349375.XA Active CN102930277B (en) | 2012-09-19 | 2012-09-19 | A kind of character picture method for recognizing verification code based on Recognition feedback |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102930277B (en) |
Cited By (44)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103345703A (en) * | 2013-06-17 | 2013-10-09 | 上海方付通商务服务有限公司 | Banking transaction authentication method and system based on image authentication |
CN103400400A (en) * | 2013-08-07 | 2013-11-20 | 南京巨鲨显示科技有限公司 | Region-based image correction method |
CN104065666A (en) * | 2014-07-03 | 2014-09-24 | 北京齐尔布莱特科技有限公司 | Method and device for generating picture identification code |
CN104112130A (en) * | 2014-06-26 | 2014-10-22 | 小米科技有限责任公司 | Optical character recognition method and device |
CN104252620A (en) * | 2014-09-25 | 2014-12-31 | 同程网络科技股份有限公司 | Character-touching graph verification code recognition method |
CN104331688A (en) * | 2014-11-05 | 2015-02-04 | 中北大学 | Detonator shell dot character identifying method |
CN104408194A (en) * | 2014-12-15 | 2015-03-11 | 北京国双科技有限公司 | Acquisition method and device of web crawler request |
CN104731764A (en) * | 2015-03-20 | 2015-06-24 | 深圳市银之杰科技股份有限公司 | Anti-fake digital font construction method and anti-fake digital font construction system |
CN104778432A (en) * | 2014-01-10 | 2015-07-15 | 携程计算机技术(上海)有限公司 | Image recognition method |
CN105187443A (en) * | 2015-09-28 | 2015-12-23 | 上海斐讯数据通信技术有限公司 | System and method for testing WEB verification code |
CN105447508A (en) * | 2015-11-10 | 2016-03-30 | 上海珍岛信息技术有限公司 | Identification method and system for character image verification codes |
CN105740863A (en) * | 2014-12-08 | 2016-07-06 | 阿里巴巴集团控股有限公司 | Information processing method and device |
CN105868590A (en) * | 2015-01-19 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Method and device for processing handwriting data |
CN105894475A (en) * | 2016-04-21 | 2016-08-24 | 上海师范大学 | International phonetic symbol image character refining method |
CN106780535A (en) * | 2016-12-21 | 2017-05-31 | 潘小胜 | A kind of gray level image processing method |
CN106920266A (en) * | 2015-12-28 | 2017-07-04 | 腾讯科技(深圳)有限公司 | The Background Generation Method and device of identifying code |
CN106934814A (en) * | 2015-12-31 | 2017-07-07 | 腾讯科技(深圳)有限公司 | A kind of background information recognition methods and device based on image |
CN107038445A (en) * | 2017-02-13 | 2017-08-11 | 上海大学 | A kind of binaryzation and dividing method for Chinese character identifying code |
CN107220983A (en) * | 2017-04-13 | 2017-09-29 | 中国农业大学 | A kind of live pig detection method and system based on video |
CN107688812A (en) * | 2017-08-25 | 2018-02-13 | 重庆慧都科技有限公司 | A kind of food production date ink-jet font restorative procedure based on machine vision |
CN107730511A (en) * | 2017-09-20 | 2018-02-23 | 北京工业大学 | A kind of Tibetan language historical document line of text cutting method based on baseline estimations |
CN108038484A (en) * | 2017-12-11 | 2018-05-15 | 中国人民解放军战略支援部队信息工程大学 | Hollow identifying code method for quickly identifying |
CN108182437A (en) * | 2017-12-29 | 2018-06-19 | 北京金堤科技有限公司 | One kind clicks method for recognizing verification code, device and user terminal |
CN108491844A (en) * | 2018-02-07 | 2018-09-04 | 西安工程大学 | Water meter automatic checkout system based on image procossing and its image processing method |
CN108805126A (en) * | 2017-04-28 | 2018-11-13 | 上海斯睿德信息技术有限公司 | A kind of long interfering line minimizing technology of text image |
CN109086772A (en) * | 2018-08-16 | 2018-12-25 | 成都市映潮科技股份有限公司 | A kind of recognition methods and system distorting adhesion character picture validation code |
CN109101969A (en) * | 2018-08-23 | 2018-12-28 | 深圳市深晓科技有限公司 | A kind of image processing method and device based on character recognition |
CN109101810A (en) * | 2018-08-14 | 2018-12-28 | 电子科技大学 | A kind of text method for recognizing verification code based on OCR technique |
CN109189683A (en) * | 2018-08-28 | 2019-01-11 | 中金金融认证中心有限公司 | A kind of method and system automatically entered for identifying code in APP test |
CN109202886A (en) * | 2017-06-30 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | Based on the gesture identification method and system under fixed background |
CN109410215A (en) * | 2018-08-02 | 2019-03-01 | 北京三快在线科技有限公司 | Image processing method, device, electronic equipment and computer-readable medium |
CN109919160A (en) * | 2019-03-04 | 2019-06-21 | 深圳先进技术研究院 | Method for recognizing verification code, device, terminal and storage medium |
CN109948621A (en) * | 2019-03-20 | 2019-06-28 | 南京工业大学 | A kind of image procossing and character segmentation method based on picture validation code |
CN110020655A (en) * | 2019-04-19 | 2019-07-16 | 厦门商集网络科技有限责任公司 | A kind of character denoising method and terminal based on binaryzation |
CN110111165A (en) * | 2019-05-13 | 2019-08-09 | 重庆天蓬网络有限公司 | True from false of bills checking method, system, medium and electronic equipment |
CN110263875A (en) * | 2019-06-27 | 2019-09-20 | 重庆市筑智建信息技术有限公司 | Method and system for comparing contour similarity of members in building BIM management |
CN110667147A (en) * | 2017-05-28 | 2020-01-10 | 中国计量大学 | Character extraction method for film section image |
CN110703760A (en) * | 2019-10-30 | 2020-01-17 | 杭州叙简科技股份有限公司 | Newly-increased suspicious object detection method for security inspection robot |
CN111091371A (en) * | 2018-10-24 | 2020-05-01 | 北京意锐新创科技有限公司 | Quick payment method and device |
CN111178352A (en) * | 2019-12-13 | 2020-05-19 | 中国建设银行股份有限公司 | Method and device for identifying verification code characters |
CN111383293A (en) * | 2020-02-26 | 2020-07-07 | 北京京东叁佰陆拾度电子商务有限公司 | Image element vectorization method and device |
CN113807359A (en) * | 2020-06-17 | 2021-12-17 | 中国石油化工股份有限公司 | Intelligent identification method for communication path between wells and electronic equipment |
CN115037471A (en) * | 2022-03-21 | 2022-09-09 | 远光软件股份有限公司 | Method, device and storage medium for checking job qualification |
CN111553317B (en) * | 2020-05-14 | 2023-08-08 | 北京惠朗时代科技有限公司 | Anti-fake code acquisition method and device, computer equipment and storage medium |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110044901B (en) * | 2019-04-30 | 2022-04-08 | 重庆康巨全弘生物科技有限公司 | Reagent card gray level analysis system |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101201939A (en) * | 2007-12-20 | 2008-06-18 | 腾讯科技(深圳)有限公司 | Method and system for generating picture identifying code |
CN101882298A (en) * | 2010-06-30 | 2010-11-10 | 中山大学 | Image checking code generating method based on invertible matrix |
CN101944177A (en) * | 2010-08-30 | 2011-01-12 | 深圳市多赢软件技术有限公司 | Method for recognizing verification code |
-
2012
- 2012-09-19 CN CN201210349375.XA patent/CN102930277B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101201939A (en) * | 2007-12-20 | 2008-06-18 | 腾讯科技(深圳)有限公司 | Method and system for generating picture identifying code |
CN101882298A (en) * | 2010-06-30 | 2010-11-10 | 中山大学 | Image checking code generating method based on invertible matrix |
CN101944177A (en) * | 2010-08-30 | 2011-01-12 | 深圳市多赢软件技术有限公司 | Method for recognizing verification code |
Cited By (64)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103345703A (en) * | 2013-06-17 | 2013-10-09 | 上海方付通商务服务有限公司 | Banking transaction authentication method and system based on image authentication |
CN103400400B (en) * | 2013-08-07 | 2017-03-01 | 南京巨鲨显示科技有限公司 | A kind of method for correcting image based on region |
CN103400400A (en) * | 2013-08-07 | 2013-11-20 | 南京巨鲨显示科技有限公司 | Region-based image correction method |
CN104778432B (en) * | 2014-01-10 | 2018-10-12 | 上海携程商务有限公司 | Image-recognizing method |
CN104778432A (en) * | 2014-01-10 | 2015-07-15 | 携程计算机技术(上海)有限公司 | Image recognition method |
CN104112130A (en) * | 2014-06-26 | 2014-10-22 | 小米科技有限责任公司 | Optical character recognition method and device |
CN104112130B (en) * | 2014-06-26 | 2017-08-01 | 小米科技有限责任公司 | optical character recognition method and device |
CN104065666A (en) * | 2014-07-03 | 2014-09-24 | 北京齐尔布莱特科技有限公司 | Method and device for generating picture identification code |
CN104065666B (en) * | 2014-07-03 | 2017-08-01 | 北京齐尔布莱特科技有限公司 | A kind of method and device for generating picture validation code |
CN104252620A (en) * | 2014-09-25 | 2014-12-31 | 同程网络科技股份有限公司 | Character-touching graph verification code recognition method |
CN104252620B (en) * | 2014-09-25 | 2017-06-06 | 同程网络科技股份有限公司 | The graphical verification code recognition methods of Characters Stuck |
CN104331688A (en) * | 2014-11-05 | 2015-02-04 | 中北大学 | Detonator shell dot character identifying method |
CN105740863A (en) * | 2014-12-08 | 2016-07-06 | 阿里巴巴集团控股有限公司 | Information processing method and device |
CN104408194A (en) * | 2014-12-15 | 2015-03-11 | 北京国双科技有限公司 | Acquisition method and device of web crawler request |
CN104408194B (en) * | 2014-12-15 | 2017-11-21 | 北京国双科技有限公司 | The acquisition methods and device of web crawlers request |
CN105868590A (en) * | 2015-01-19 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Method and device for processing handwriting data |
CN104731764A (en) * | 2015-03-20 | 2015-06-24 | 深圳市银之杰科技股份有限公司 | Anti-fake digital font construction method and anti-fake digital font construction system |
CN105187443A (en) * | 2015-09-28 | 2015-12-23 | 上海斐讯数据通信技术有限公司 | System and method for testing WEB verification code |
CN105187443B (en) * | 2015-09-28 | 2018-03-06 | 上海斐讯数据通信技术有限公司 | A kind of system and method for test WEB identifying codes |
CN105447508A (en) * | 2015-11-10 | 2016-03-30 | 上海珍岛信息技术有限公司 | Identification method and system for character image verification codes |
CN106920266B (en) * | 2015-12-28 | 2019-11-05 | 腾讯科技(深圳)有限公司 | The Background Generation Method and device of identifying code |
CN106920266A (en) * | 2015-12-28 | 2017-07-04 | 腾讯科技(深圳)有限公司 | The Background Generation Method and device of identifying code |
CN106934814A (en) * | 2015-12-31 | 2017-07-07 | 腾讯科技(深圳)有限公司 | A kind of background information recognition methods and device based on image |
CN106934814B (en) * | 2015-12-31 | 2020-08-14 | 腾讯科技(深圳)有限公司 | Background information identification method and device based on image |
CN105894475A (en) * | 2016-04-21 | 2016-08-24 | 上海师范大学 | International phonetic symbol image character refining method |
CN106780535A (en) * | 2016-12-21 | 2017-05-31 | 潘小胜 | A kind of gray level image processing method |
CN107038445A (en) * | 2017-02-13 | 2017-08-11 | 上海大学 | A kind of binaryzation and dividing method for Chinese character identifying code |
CN107220983B (en) * | 2017-04-13 | 2019-09-24 | 中国农业大学 | A kind of live pig detection method and system based on video |
CN107220983A (en) * | 2017-04-13 | 2017-09-29 | 中国农业大学 | A kind of live pig detection method and system based on video |
CN108805126B (en) * | 2017-04-28 | 2021-09-10 | 上海斯睿德信息技术有限公司 | Method for removing long interference lines of text image |
CN108805126A (en) * | 2017-04-28 | 2018-11-13 | 上海斯睿德信息技术有限公司 | A kind of long interfering line minimizing technology of text image |
CN110667147B (en) * | 2017-05-28 | 2021-10-22 | 中国计量大学 | Character extraction method for film section image |
CN110667147A (en) * | 2017-05-28 | 2020-01-10 | 中国计量大学 | Character extraction method for film section image |
CN109202886A (en) * | 2017-06-30 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | Based on the gesture identification method and system under fixed background |
CN107688812B (en) * | 2017-08-25 | 2020-04-21 | 重庆慧都科技有限公司 | Food production date ink-jet font repairing method based on machine vision |
CN107688812A (en) * | 2017-08-25 | 2018-02-13 | 重庆慧都科技有限公司 | A kind of food production date ink-jet font restorative procedure based on machine vision |
CN107730511B (en) * | 2017-09-20 | 2020-10-27 | 北京工业大学 | Tibetan historical literature text line segmentation method based on baseline estimation |
CN107730511A (en) * | 2017-09-20 | 2018-02-23 | 北京工业大学 | A kind of Tibetan language historical document line of text cutting method based on baseline estimations |
CN108038484B (en) * | 2017-12-11 | 2020-05-05 | 中国人民解放军战略支援部队信息工程大学 | Method for quickly identifying hollow verification code |
CN108038484A (en) * | 2017-12-11 | 2018-05-15 | 中国人民解放军战略支援部队信息工程大学 | Hollow identifying code method for quickly identifying |
CN108182437A (en) * | 2017-12-29 | 2018-06-19 | 北京金堤科技有限公司 | One kind clicks method for recognizing verification code, device and user terminal |
CN108182437B (en) * | 2017-12-29 | 2020-07-03 | 北京金堤科技有限公司 | Click verification code identification method and device and user terminal |
CN108491844A (en) * | 2018-02-07 | 2018-09-04 | 西安工程大学 | Water meter automatic checkout system based on image procossing and its image processing method |
CN109410215A (en) * | 2018-08-02 | 2019-03-01 | 北京三快在线科技有限公司 | Image processing method, device, electronic equipment and computer-readable medium |
CN109101810B (en) * | 2018-08-14 | 2021-07-06 | 电子科技大学 | Character verification code recognition method based on OCR technology |
CN109101810A (en) * | 2018-08-14 | 2018-12-28 | 电子科技大学 | A kind of text method for recognizing verification code based on OCR technique |
CN109086772A (en) * | 2018-08-16 | 2018-12-25 | 成都市映潮科技股份有限公司 | A kind of recognition methods and system distorting adhesion character picture validation code |
CN109101969A (en) * | 2018-08-23 | 2018-12-28 | 深圳市深晓科技有限公司 | A kind of image processing method and device based on character recognition |
CN109189683A (en) * | 2018-08-28 | 2019-01-11 | 中金金融认证中心有限公司 | A kind of method and system automatically entered for identifying code in APP test |
CN111091371A (en) * | 2018-10-24 | 2020-05-01 | 北京意锐新创科技有限公司 | Quick payment method and device |
CN109919160A (en) * | 2019-03-04 | 2019-06-21 | 深圳先进技术研究院 | Method for recognizing verification code, device, terminal and storage medium |
CN109919160B (en) * | 2019-03-04 | 2021-03-23 | 深圳先进技术研究院 | Verification code identification method, device, terminal and storage medium |
CN109948621A (en) * | 2019-03-20 | 2019-06-28 | 南京工业大学 | A kind of image procossing and character segmentation method based on picture validation code |
CN110020655A (en) * | 2019-04-19 | 2019-07-16 | 厦门商集网络科技有限责任公司 | A kind of character denoising method and terminal based on binaryzation |
CN110111165A (en) * | 2019-05-13 | 2019-08-09 | 重庆天蓬网络有限公司 | True from false of bills checking method, system, medium and electronic equipment |
CN110263875A (en) * | 2019-06-27 | 2019-09-20 | 重庆市筑智建信息技术有限公司 | Method and system for comparing contour similarity of members in building BIM management |
CN110703760A (en) * | 2019-10-30 | 2020-01-17 | 杭州叙简科技股份有限公司 | Newly-increased suspicious object detection method for security inspection robot |
CN110703760B (en) * | 2019-10-30 | 2023-06-02 | 杭州叙简科技股份有限公司 | Newly-added suspicious object detection method for security inspection robot |
CN111178352A (en) * | 2019-12-13 | 2020-05-19 | 中国建设银行股份有限公司 | Method and device for identifying verification code characters |
CN111383293A (en) * | 2020-02-26 | 2020-07-07 | 北京京东叁佰陆拾度电子商务有限公司 | Image element vectorization method and device |
CN111553317B (en) * | 2020-05-14 | 2023-08-08 | 北京惠朗时代科技有限公司 | Anti-fake code acquisition method and device, computer equipment and storage medium |
CN113807359A (en) * | 2020-06-17 | 2021-12-17 | 中国石油化工股份有限公司 | Intelligent identification method for communication path between wells and electronic equipment |
CN113807359B (en) * | 2020-06-17 | 2024-05-10 | 中国石油化工股份有限公司 | Intelligent identification method for inter-well communication path and electronic equipment |
CN115037471A (en) * | 2022-03-21 | 2022-09-09 | 远光软件股份有限公司 | Method, device and storage medium for checking job qualification |
Also Published As
Publication number | Publication date |
---|---|
CN102930277B (en) | 2016-04-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102930277B (en) | A kind of character picture method for recognizing verification code based on Recognition feedback | |
Zhang et al. | Interpreting adversarially trained convolutional neural networks | |
CN110533084B (en) | Multi-scale target detection method based on self-attention mechanism | |
CN110032998B (en) | Method, system, device and storage medium for detecting characters of natural scene picture | |
CN103942550B (en) | A kind of scene text recognition methods based on sparse coding feature | |
CN110598690B (en) | End-to-end optical character detection and recognition method and system | |
CN102163284B (en) | Chinese environment-oriented complex scene text positioning method | |
CN104751142B (en) | A kind of natural scene Method for text detection based on stroke feature | |
CN103353938B (en) | A kind of cell membrane dividing method based on hierarchical level feature | |
CN104821011A (en) | Method of generating 3D house type model by 2D house type model based on camera shooting | |
CN111738055B (en) | Multi-category text detection system and bill form detection method based on same | |
CN105701488A (en) | Identity card identification method | |
CN108399424B (en) | Point cloud classification method, intelligent terminal and storage medium | |
CN104156730B (en) | A kind of antinoise Research of Chinese Feature Extraction method based on skeleton | |
CN105279506A (en) | Manchu script central axis positioning method | |
CN105447508A (en) | Identification method and system for character image verification codes | |
CN110598686A (en) | Invoice identification method, system, electronic equipment and medium | |
CN110969620A (en) | Method and device for detecting magnetic shoe ripple defects | |
CN102136074B (en) | Man-machine interface (MMI) based wood image texture analyzing and identifying method | |
CN105404885A (en) | Two-dimensional character graphic verification code complex background noise interference removal method | |
CN112819837A (en) | Semantic segmentation method based on multi-source heterogeneous remote sensing image | |
CN104484679B (en) | Non- standard rifle shooting warhead mark image automatic identifying method | |
CN109271882B (en) | Method for extracting color-distinguished handwritten Chinese characters | |
CN106780535A (en) | A kind of gray level image processing method | |
Moustafa et al. | Hieroglyphs language translator using deep learning techniques (Scriba) |
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 | ||
CP03 | Change of name, title or address | ||
CP03 | Change of name, title or address |
Address after: 200437 room 547, new building 291, wunshui East Road, Hongkou District, Shanghai. Patentee after: Information technology of the island of Shanghai (Shanghai) Limited by Share Ltd Address before: 200434 1876, room 60, Lane 465, Liangcheng Road, Hongkou District, Shanghai. Patentee before: Shanghai Trueland Information Technology Co., Ltd. |