CN104778432A - Image recognition method - Google Patents
Image recognition method Download PDFInfo
- Publication number
- CN104778432A CN104778432A CN201410011080.0A CN201410011080A CN104778432A CN 104778432 A CN104778432 A CN 104778432A CN 201410011080 A CN201410011080 A CN 201410011080A CN 104778432 A CN104778432 A CN 104778432A
- Authority
- CN
- China
- Prior art keywords
- pixel
- color
- image
- demarcation
- location
- 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
Abstract
The invention discloses an image recognition method. The method comprises steps: S1, gray processing and binary processing are carried out on the image so as to acquire a processing image formed by a first color and a second color, and rectangles are recognized; S2, the mutually-connected rectangles form a rectangle area; S3, if the number of pixels of the first color in a surrounding area, except the pixels, in a first pixel area is larger than the number of pixels of the second color, the pixels are the first color, and if the number of pixels in the surrounding area is smaller than the number of pixels of the second color, the pixels are the second color; S4, the rectangle area is set to be the first color; S5, the processing image is divided to acquire single character images; and S6, the characteristic value of each single character image is calculated, and according to the characteristic value, a character with the nearest characteristic value is acquired in a character sample bank to serve as the character corresponding to the single character image. According to the image recognition method of the invention, the character recognition accuracy rate is high.
Description
Technical field
The present invention relates to a kind of image-recognizing method, particularly a kind ofly identify that light and shade shade that application level and vertical stripes interweave is to represent the image-recognizing method of the picture of character.
Background technology
Image recognition is by a series of process such as the information stored and photo current information compares, calculating, realizes re-recognizing image.Image recognition is a key areas of artificial intelligence.Image recognition technology can provide a lot of convenience for user, and such as carrying out identification for car plate can rapid screening car plate; For another example identification is carried out to the identifying code of number of site and can facilitate the operations such as user places an order, purchase, more can provide for user the convenient service substituting and perform some operation.But existing recognition methods is comparatively single, the accuracy rate of identification is lower.At present, there is light and shade shade that a kind of comparatively conventional application level and vertical stripes interweave to represent character.Very low to the accuracy rate of this picture recognition in the middle of prior art, it is the requirement that user serves that the accuracy rate of identification can not reach website use identification code.
Summary of the invention
The technical problem to be solved in the present invention is that the method in order to overcome recognition image in prior art is comparatively single, the light and shade shade that application level and vertical stripes interweave represents the low-down defect of the recognition accuracy of the picture of character, provides the image-recognizing method that a kind of recognition accuracy is higher.
The present invention solves above-mentioned technical matters by following technical proposals: a kind of image-recognizing method, and its feature is, described image-recognizing method comprises:
S
1, described image carried out to gray proces and binary conversion treatment to obtain the process image be made up of the first color and the second color, in identifying processing image, all sizes be made up of the pixel of the second color are the rectangle of a*b, wherein a is the integer that pixel value is greater than 2, and b is the integer that pixel value is greater than 2;
S
2, travel through described process image and interconnective described rectangle formed rectangular area;
S
3, for each pixel, in the first pixel region that size is c*c centered by described pixel, if the quantity of the pixel of the first color is greater than the quantity of the pixel of the second color in the peripheral region of the first pixel region except described pixel, described pixel is the first color, if the quantity of the first color is less than the quantity of the pixel of the second color in described peripheral region, described pixel is the second color, and wherein c is the odd number that pixel value is greater than 1;
S
4, rectangular area is set to the first color;
S
5, divide described process image to obtain single character picture;
S
6, calculate each single character picture eigenwert and in character sample storehouse, obtain the immediate character of eigenwert as character corresponding to single character picture according to eigenwert.
In the present invention, described image is through gray proces and binary conversion treatment, and the first color can be black or white, and the second color can be the white different from the first color or black, if the first color is white, the second color is black.Described pixel is the pixel of the rectangle only having a kind of color, and such as two pixels are transversely arranged, and composition size is the rectangle of 2*1.
Described rectangular area can be an irregular figure, and such as size is that the rectangular area that 3*3 only comprises the pixel of the second color comprises multiple sub-rectangle, and size is the rectangle of 2*2,3*2 etc.After this little rectangle of identification, described sub-rectangle can be combined into 3*3 rectangular area.
For each pixel, decide the color of pixel self according to the color around pixel, after this step process, described process image can demonstrate character picture well, character picture and background image is significantly distinguished.For example, in the first pixel region that size is 3*3 centered by the pixel of black color, if the quantity of white colour pixel is greater than the quantity of black colored pixels in 8 pixels beyond the pixel going out described black color, then the black colored pixels at the first pixel region center is set to white.
Through step S
4after, in described process image, character picture is made up of the pixel of the first color, and background image is made up of the pixel of the second color.Step S
5middle division described process image can utilize existing techniques in realizing to obtain single character picture.
Equally, to the calculating of single character image feature value and obtain the immediate character of eigenwert according to eigenwert in character sample storehouse prior art also can be utilized to complete.
Preferably, step S
3comprise:
S
31, for each pixel, judge whether the quantity of the pixel of the first color in described peripheral region is less than the quantity of the pixel of the second color, then described pixel is that then the first color performs step S if not
4if then perform step S
32;
S
32, for step S
31in pixel, if the quantity of the first color is less than the quantity of the pixel of the second color in described peripheral region, then described pixel is problem pixel, judge whether described problem pixel is the first color, if then described problem pixel to be saved in a duplicating image and then to perform step S after the problem pixel of described duplicating image is set to the second color
33, then perform step S if not
4, the pixel of wherein said duplicating image except problem pixel is corresponding consistent with the pixel of described process image;
S
33, for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, judge whether centered by the pixel processing image size is less than the quantity of the pixel of the second color as the quantity of the pixel of the first color in second pixel region of d*d and size is d*d centered by the pixel of duplicating image the 3rd pixel region, if then the pixel of described process image is the second color, the pixel then processing image is if not the first color, and wherein d is the odd number that pixel value is greater than 1.
Because the pixel of character picture by the first color forms, it is more careful to need the processes pixel of the first color, the pixel representing character picture and the pixel representing background image can be identified more accurately by above-mentioned steps, thus improve recognition accuracy further.
Preferably, step S
33comprise: for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, if the quantity of the pixel of the first color equals the quantity of the pixel of the second color in described second pixel region and described 3rd pixel region, then judge whether the center of the pixel processing image is greater than the distance of center to the center of the pixel of the second nearest color of the pixel of duplicating image to the centre distance of the pixel of the first nearest color, if the pixel of described process image is the second color, then the pixel of described process image is the first color if not.
The pixel that can be identified pixel and the expression background image representing character picture by above-mentioned steps is more accurate, improves recognition accuracy significantly.
Preferably, described binary conversion treatment is: for each pixel, and the pixel that the gray-scale value of pixel is greater than a first threshold is the first color, and the pixel that described gray-scale value is less than described first threshold is the second color.
Preferably, step S
5for:
S
5, divide process image to obtain single character picture, wherein said process image through the second color retouch limit process.
After retouching limit process, it is all second colors that described process image goes out beyond character picture, is more prone to identify for background and character picture.
Preferably, step S
4after comprise:
S
41, the noise pixel removed in described process image, then perform step S
5.
Remove noise pixel and namely remove pixel isolated in image, thus it is comparatively accurate to make next step obtain single character picture.
Preferably, step S
4after comprise:
S
41, for each pixel, with the center of described pixel on the same line and two pixels adjacent to described pixel are polishing pixel, if the color of described polishing pixel is identical, then described pixel is set to the color identical with described polishing pixel, then performs step S
5.
Isolated pixel can not only be removed by above-mentioned steps, can also more easily identify to make the character picture in image polishing image border.
Preferably, the bearing of trend of character picture is first direction, is second direction perpendicular to described bearing of trend, and perpendicular to first direction and the rectangle that size is 1*e is the first sweep trace, wherein e is second direction max pixel value, step S
5comprise:
S
51, identify that the position on the first direction that the first sweep trace that the first sweep trace of being all made up of the pixel of the second color and the pixel by the first color and the second color form is adjacent is the first demarcation of location;
S
52, identify spacing and meet and be greater than Second Threshold and two the first demarcation of location being less than the 3rd threshold value, the region that described two the first demarcation of location divide comprises the pixel of the first color, and described two the first demarcation of location are a single character picture two edges in a first direction respectively;
S
53, for described in each two first demarcation of location divide region, perpendicular to second direction and the rectangle that size is f*1 is the second sweep trace, wherein f is the pixel value of the spacing of described two the first demarcation of location, identify the second sweep trace of being all made up of the pixel of the second color and the position comprised in the adjacent second direction of second sweep trace of pixel of the first color is the second demarcation of location, maximum two the second demarcation of location of mutual spacing are two edges in the second direction of a single character picture respectively.
Single character picture can be accurately marked off by above-mentioned steps.
Preferably, step S
32in c equal step S
33in d equal 3.
Preferably, step S
6comprise:
S
61, to by two the first demarcation of location and two second demarcation of location divide single character picture carry out standardization;
S
62, discrete cosine transform is carried out to single character picture after, obtain two-dimensional matrix;
S
63, hash is carried out to described two-dimensional matrix after, obtain the eigenwert of single character picture.
All can be realized by existing means the two-dimensional matrix hash of the standardization of single character picture, discrete cosine transform and single character picture, repeat no more herein.
On the basis meeting this area general knowledge, above-mentioned each optimum condition, can combination in any, obtains the preferred embodiments of the invention.
Positive progressive effect of the present invention is: image-recognizing method provided by the invention can identify the character picture of prior art None-identified, and the accuracy rate of identification character is high, can reach more than 95 percent.Utilizing image-recognizing method of the present invention to carry out identification to identification code picture just can replace user to realize many operations, as air ticket, hotel predetermined, thus provide convenient for user.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the image-recognizing method of the embodiment of the present invention.
Fig. 2 is the subregion of the image of the image-recognizing method of the embodiment of the present invention.
Fig. 3 is the image that the image-recognizing method of the embodiment of the present invention carries out identifying.
Embodiment
Mode below by embodiment further illustrates the present invention, but does not therefore limit the present invention among described scope of embodiments.
Embodiment
The present embodiment provides a kind of image-recognizing method, and for identifying that light and shade shade that application level as shown in Figure 3 and vertical stripes interweave is to represent the image of character, see Fig. 1, described image-recognizing method comprises:
Step 100, described image carried out to gray proces and binary conversion treatment to obtain the process image be made up of white and black, in identifying processing image, all sizes be made up of the pixel of black are the rectangle of a*b, wherein a is the integer that pixel value is greater than 2, and b is the integer that pixel value is greater than 2.
In above-mentioned steps, white is the preferred of the first color, and black is the preferred of the second color.Described binary conversion treatment is: for each pixel, and the gray-scale value of pixel is greater than the pixel of a first threshold for white, and the pixel that described gray-scale value is less than described first threshold is black.
Step 101, travel through described process image and interconnective described rectangle is formed rectangular area.
Described rectangular area can be an irregular figure, the subregion of image of the present invention shown in Figure 2, region is made up of the pixel 11 of white and the pixel 21 of black, the pixel of white and the pixel of black all sized by be the rectangle of 1*1, comprise multiple sub-rectangle in the rectangular area that size is 6*4, after identifying this little rectangle, described sub-rectangle can be combined into the rectangular area with oblique line in Fig. 2.
Step 102, for each pixel, in the first pixel region that size is 3*3 centered by described pixel, if the quantity of the pixel of white is greater than the quantity of the pixel of black in the peripheral region of the first pixel region except described pixel, described pixel is white, if the quantity of white is less than the quantity of the pixel of black in described peripheral region, described pixel is black.
For each pixel, decide the color of pixel self according to the color around pixel, after this step process, described process image can demonstrate character picture well, character picture and background image is significantly distinguished.In the first pixel region that size is 3*3 centered by the pixel of black, if the quantity of white pixel is greater than the quantity of black picture element in 8 pixels beyond the pixel going out described black, then the black picture element at the first pixel region center is set to white.
In order to improve recognition accuracy further, step 102 be preferably:
Step 1021, for each pixel, to judge in described peripheral region whether the quantity of the pixel of white is less than the quantity of the pixel of black, and then described pixel is that then white perform step 103 if not, if then perform step 1022.
Step 1022, for the pixel in step 1021, if the quantity of white is less than the quantity of the pixel of black in described peripheral region, then described pixel is problem pixel, judge whether described problem pixel is white, if then described problem pixel to be saved in a duplicating image and then to perform step 1023 after the problem pixel of described duplicating image is set to black, then perform step 103 if not, the pixel of wherein said duplicating image except problem pixel is corresponding consistent with the pixel of described process image.
Step 1023, for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, judge whether centered by the pixel processing image size is less than the quantity of the pixel of black as the quantity of the pixel of white in second pixel region of 3*3 and size is 3*3 centered by the pixel of duplicating image the 3rd pixel region, if then the pixel of described process image is black, then process the pixel of image if not for white, then perform step 103.
Wherein step 1023 comprises: for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, if the quantity of the pixel of white equals the quantity of the pixel of black in described second pixel region and described 3rd pixel region, then judge whether the center of the pixel processing image is greater than the distance of center to the center of the pixel of nearest black of the pixel of duplicating image to the centre distance of the pixel of nearest white, if the pixel of described process image is black, then the pixel of described process image is white if not.
Step 103, rectangular area is set to white.
Step 104, the noise pixel removed in described process image.
Step 105, for each pixel, with the center of described pixel on the same line and two pixels adjacent to described pixel are polishing pixel, if the color of described polishing pixel is identical, then described pixel is set to the color identical with described polishing pixel.
Position on the first direction that the first sweep trace that the first sweep trace that step 106, identification are all made up of the pixel of black and the pixel by white and black form is adjacent is the first demarcation of location.
Step 107, identify spacing and meet and be greater than Second Threshold and two the first demarcation of location being less than the 3rd threshold value, the region that described two the first demarcation of location divide comprises the pixel of white, and described two the first demarcation of location are a single character picture two edges in a first direction respectively.
Step 108, the region that two the first demarcation of location described in each are divided, perpendicular to second direction and the rectangle that size is e*1 is the second sweep trace, identify the second sweep trace of being all made up of the pixel of the second color and the position comprised in the adjacent second direction of second sweep trace of pixel of the first color is the second demarcation of location, maximum two the second demarcation of location of mutual spacing are two edges in the second direction of a single character picture respectively.
Wherein, the bearing of trend of character picture is first direction, is second direction perpendicular to described bearing of trend, and perpendicular to first direction and the rectangle that size is 1*f is the first sweep trace, and described process image retouches limit process through black.After retouching limit process, it is all black that described process image goes out beyond character picture, is more prone to identify for background and character picture.
Wherein e is second direction max pixel value, and f is the pixel value of the spacing of described two the first demarcation of location.
Step 109, to by two the first demarcation of location and two second demarcation of location divide single character picture carry out standardization.
Step 110, discrete cosine transform is carried out to single character picture after, obtain two-dimensional matrix.
Step 111, hash is carried out to described two-dimensional matrix after, obtain the eigenwert of single character picture.
Then, in character sample storehouse, the immediate character of eigenwert is obtained as character corresponding to single character picture according to eigenwert.
The image-recognizing method of the present embodiment can identify the character picture of prior art None-identified, and the accuracy rate of identification character is high, and accuracy rate can reach more than 95 percent.
Although the foregoing describe the specific embodiment of the present invention, it will be understood by those of skill in the art that these only illustrate, protection scope of the present invention is defined by the appended claims.Those skilled in the art, under the prerequisite not deviating from principle of the present invention and essence, can make various changes or modifications to these embodiments, but these change and amendment all falls into protection scope of the present invention.
Claims (10)
1. an image-recognizing method, is characterized in that, described image-recognizing method comprises:
S
1, described image carried out to gray proces and binary conversion treatment to obtain the process image be made up of the first color and the second color, in identifying processing image, all sizes be made up of the pixel of the second color are the rectangle of a*b, wherein a is the integer that pixel value is greater than 2, and b is the integer that pixel value is greater than 2;
S
2, travel through described process image and interconnective described rectangle formed rectangular area;
S
3, for each pixel, in the first pixel region that size is c*c centered by described pixel, if the quantity of the pixel of the first color is greater than the quantity of the pixel of the second color in the peripheral region of the first pixel region except described pixel, described pixel is the first color, if the quantity of the first color is less than the quantity of the pixel of the second color in described peripheral region, described pixel is the second color, and wherein c is the odd number that pixel value is greater than 1;
S
4, rectangular area is set to the first color;
S
5, divide described process image to obtain single character picture;
S
6, calculate each single character picture eigenwert and in character sample storehouse, obtain the immediate character of eigenwert as character corresponding to single character picture according to eigenwert.
2. image-recognizing method as claimed in claim 1, is characterized in that, step S
3comprise:
S
31, for each pixel, judge whether the quantity of the pixel of the first color in described peripheral region is less than the quantity of the pixel of the second color, then described pixel is that then the first color performs step S if not
4if then perform step S
32;
S
32, for step S
31in pixel, if the quantity of the first color is less than the quantity of the pixel of the second color in described peripheral region, then described pixel is problem pixel, judge whether described problem pixel is the first color, if then described problem pixel to be saved in a duplicating image and then to perform step S after the problem pixel of described duplicating image is set to the second color
33, then perform step S if not
4, the pixel of wherein said duplicating image except problem pixel is corresponding consistent with the pixel of described process image;
S
33, for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, judge whether centered by the pixel processing image size is less than the quantity of the pixel of the second color as the quantity of the pixel of the first color in second pixel region of d*d and size is d*d centered by the pixel of duplicating image the 3rd pixel region, if then the pixel of described process image is the second color, the pixel then processing image is if not the first color, and wherein d is the odd number that pixel value is greater than 1.
3. image-recognizing method as claimed in claim 2, is characterized in that, step S
33comprise: for the pixel of each process image and the pixel of the duplicating image corresponding with the pixel of described process image, if the quantity of the pixel of the first color equals the quantity of the pixel of the second color in described second pixel region and described 3rd pixel region, then judge whether the center of the pixel processing image is greater than the distance of center to the center of the pixel of the second nearest color of the pixel of duplicating image to the centre distance of the pixel of the first nearest color, if the pixel of described process image is the second color, then the pixel of described process image is the first color if not.
4. image-recognizing method as claimed in claim 1, it is characterized in that, described binary conversion treatment is: for each pixel, and the pixel that the gray-scale value of pixel is greater than a first threshold is the first color, and the pixel that described gray-scale value is less than described first threshold is the second color.
5. image-recognizing method as claimed in claim 1, is characterized in that, step S
5for:
S
5, divide process image to obtain single character picture, wherein said process image through the second color retouch limit process.
6. image-recognizing method as claimed in claim 5, is characterized in that, step S
4after comprise:
S
41, the noise pixel removed in described process image, then perform step S
5.
7. image-recognizing method as claimed in claim 5, is characterized in that, step S
4after comprise:
S
41, for each pixel, with the center of described pixel on the same line and two pixels adjacent to described pixel are polishing pixel, if the color of described polishing pixel is identical, then described pixel is set to the color identical with described polishing pixel, then performs step S
5.
8. image-recognizing method as claimed in claim 7, it is characterized in that, the bearing of trend of character picture is first direction, be second direction perpendicular to described bearing of trend, perpendicular to first direction and the rectangle that size is 1*e is the first sweep trace, wherein e is second direction max pixel value, step S
5comprise:
S
51, identify that the position on the first direction that the first sweep trace that the first sweep trace of being all made up of the pixel of the second color and the pixel by the first color and the second color form is adjacent is the first demarcation of location;
S
52, identify spacing and meet and be greater than Second Threshold and two the first demarcation of location being less than the 3rd threshold value, the region that described two the first demarcation of location divide comprises the pixel of the first color, and described two the first demarcation of location are a single character picture two edges in a first direction respectively;
S
53, for described in each two first demarcation of location divide region, perpendicular to second direction and the rectangle that size is f*1 is the second sweep trace, wherein f is the pixel value of the spacing of described two the first demarcation of location, identify the second sweep trace of being all made up of the pixel of the second color and the position comprised in the adjacent second direction of second sweep trace of pixel of the first color is the second demarcation of location, maximum two the second demarcation of location of mutual spacing are two edges in the second direction of a single character picture respectively.
9. image-recognizing method as claimed in claim 2, is characterized in that, step S
32in c equal step S
33in d equal 3.
10. image-recognizing method as claimed in claim 8, is characterized in that, step S
6comprise:
S
61, to by two the first demarcation of location and two second demarcation of location divide single character picture carry out standardization;
S
62, discrete cosine transform is carried out to single character picture after, obtain two-dimensional matrix;
S
63, hash is carried out to described two-dimensional matrix after, obtain the eigenwert of single character picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410011080.0A CN104778432B (en) | 2014-01-10 | 2014-01-10 | Image-recognizing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410011080.0A CN104778432B (en) | 2014-01-10 | 2014-01-10 | Image-recognizing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104778432A true CN104778432A (en) | 2015-07-15 |
| CN104778432B CN104778432B (en) | 2018-10-12 |
Family
ID=53619886
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410011080.0A Active CN104778432B (en) | 2014-01-10 | 2014-01-10 | Image-recognizing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104778432B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022142A (en) * | 2016-05-04 | 2016-10-12 | 泰康人寿保险股份有限公司 | Image privacy information processing method and device |
| CN106940804A (en) * | 2017-02-23 | 2017-07-11 | 杭州仟金顶卓筑信息科技有限公司 | Architectural engineering material management system form data method for automatically inputting |
| CN113311830A (en) * | 2016-06-03 | 2021-08-27 | 苏州宝时得电动工具有限公司 | Automatic walking equipment and target area identification method |
| CN114494474A (en) * | 2022-01-17 | 2022-05-13 | 广东石油化工学院 | Generation method, verification method and device of universal verification code of human and legal equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20090123523A (en) * | 2008-05-28 | 2009-12-02 | 고려대학교 산학협력단 | System and method for recognizing optical characters |
| US7949155B2 (en) * | 2005-03-08 | 2011-05-24 | Fuji Xerox Co., Ltd. | Image processing apparatus, image processing method, and storage medium |
| CN102739620A (en) * | 2011-04-14 | 2012-10-17 | 鸿富锦精密工业（深圳）有限公司 | Verification system and verification method |
| CN102930277A (en) * | 2012-09-19 | 2013-02-13 | 上海珍岛信息技术有限公司 | Character picture verification code identifying method based on identification feedback |
-
2014
- 2014-01-10 CN CN201410011080.0A patent/CN104778432B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7949155B2 (en) * | 2005-03-08 | 2011-05-24 | Fuji Xerox Co., Ltd. | Image processing apparatus, image processing method, and storage medium |
| KR20090123523A (en) * | 2008-05-28 | 2009-12-02 | 고려대학교 산학협력단 | System and method for recognizing optical characters |
| CN102739620A (en) * | 2011-04-14 | 2012-10-17 | 鸿富锦精密工业（深圳）有限公司 | Verification system and verification method |
| CN102930277A (en) * | 2012-09-19 | 2013-02-13 | 上海珍岛信息技术有限公司 | Character picture verification code identifying method based on identification feedback |
Non-Patent Citations (1)
| Title |
|---|
| 许宏涛: ""自动化图像内容识别方法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022142A (en) * | 2016-05-04 | 2016-10-12 | 泰康人寿保险股份有限公司 | Image privacy information processing method and device |
| CN106022142B (en) * | 2016-05-04 | 2019-12-10 | 泰康保险集团股份有限公司 | Image privacy information processing method and device |
| CN113311830A (en) * | 2016-06-03 | 2021-08-27 | 苏州宝时得电动工具有限公司 | Automatic walking equipment and target area identification method |
| CN106940804A (en) * | 2017-02-23 | 2017-07-11 | 杭州仟金顶卓筑信息科技有限公司 | Architectural engineering material management system form data method for automatically inputting |
| CN114494474A (en) * | 2022-01-17 | 2022-05-13 | 广东石油化工学院 | Generation method, verification method and device of universal verification code of human and legal equipment |
| CN114494474B (en) * | 2022-01-17 | 2022-09-20 | 广东石油化工学院 | Generation method, verification method and device of universal verification code of human and legal equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104778432B (en) | 2018-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Arévalo et al. | Shadow detection in colour high‐resolution satellite images | |
| CN106097444A (en) | High-precision map generates method and apparatus | |
| US9811746B2 (en) | Method and system for detecting traffic lights | |
| CN102930268B (en) | A kind of for polluting and the accurate positioning method of DataMatrix code in various visual angles situation | |
| KR101182173B1 (en) | Method and system for recognizing vehicle plate | |
| CN105488492B (en) | A kind of color image preprocess method, roads recognition method and relevant apparatus | |
| CN106203433A (en) | In a kind of vehicle monitoring image, car plate position automatically extracts and the method for perspective correction | |
| CN102968760A (en) | Image dehazing method and image dehazing system | |
| CN104778432A (en) | Image recognition method | |
| CN103093203A (en) | Human body re-recognition method and human body re-recognition system | |
| JP6890427B2 (en) | Car number recognition device | |
| US20160379031A1 (en) | High capacity 2d color barcode design and processing method for camera based applications | |
| CN108205671A (en) | Image processing method and device | |
| KR101549495B1 (en) | An apparatus for extracting characters and the method thereof | |
| CN111382658B (en) | Road traffic sign detection method in natural environment based on image gray gradient consistency | |
| CN104021527A (en) | Rain and snow removal method in image | |
| US8000535B2 (en) | Methods and systems for refining text segmentation results | |
| CN103279755A (en) | Vehicle bottom shadow characteristics-based rapid license plate positioning method | |
| JP2019528543A (en) | Digital object unique identifier (DOI) recognition method and apparatus | |
| JP2017016635A (en) | High capacity 2d color barcode design and processing method for camera based applications | |
| CN103544716A (en) | Method and device for classifying colors of pixels of image | |
| CN106485663A (en) | A kind of lane line image enchancing method and system | |
| US20140205200A1 (en) | Image binarization using dynamic sub-image division | |
| CN102831419B (en) | Method for detecting and blurring plate number in street view image rapidly | |
| CN104091166B (en) | Dynamic threshold method for lane line edge extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160205 Address after: 200335 Shanghai city Changning District Admiralty Road No. 968 Building No. 16 10 floor Applicant after: SHANGHAI XIECHENG BUSINESS CO., LTD. Address before: 200335 Shanghai City, Changning District Fuquan Road No. 99, Ctrip network technology building Applicant before: Ctrip computer technology (Shanghai) Co., Ltd. |
|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |