CN103226696A - License plate recognition system and license plate recognition method - Google Patents
License plate recognition system and license plate recognition method Download PDFInfo
- Publication number
- CN103226696A CN103226696A CN2013101177220A CN201310117722A CN103226696A CN 103226696 A CN103226696 A CN 103226696A CN 2013101177220 A CN2013101177220 A CN 2013101177220A CN 201310117722 A CN201310117722 A CN 201310117722A CN 103226696 A CN103226696 A CN 103226696A
- Authority
- CN
- China
- Prior art keywords
- character
- rectangular area
- characters
- car plate
- license plate
- 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)
- Image Analysis (AREA)
Abstract
The invention provides a license plate recognition method which comprises the steps that a frequency spectrum threshold of an image containing a license plate after filtering is set; a rectangular area with a frequency spectrum exceeding the threshold and in the same color as the license plate is selected; whether a shape of the rectangular area conforms to the shape of the license plate is judged; if so, the characteristics of characters in the rectangular area are extracted, and whether the characteristics of the characters in the rectangular area are consistent is judged; if the characteristics are consistent, the rectangular area is subjected to graying, binaryzation, frame removal and lutetium nail removal; each character in the rectangular area is segmented; whether the number of the segmented characters meets the number of characters of the license plate is judged; if so, whether shapes of the segmented characters are the same, and whether the characters are linearly arrayed are judged; if so, sizes of the characters are normalized; the extracted character statistical characteristics are classified; and a license plate recognition result is output. With the adoption of the method, the license plate with a random resolution ratio can be recognized, and the method is high in recognition rate, and high in environmental adaptability. The invention further provides a license plate recognition system.
Description
Technical field
The present invention relates to computer vision and area of pattern recognition, particularly a kind of recognition system of car plate and method.
Background technology
The robotization of traffic administration, the intellectuality of traffic system are the development trends of 21 century world's road traffic; In intelligent transportation system, automatic license plate identification system is a very important developing direction.
Demand increase along with the full-automatic supervisory system of vehicle, market presses for the reasonable Vehicle License Plate Recognition System of performance, particularly at the car plate identification of arbitrary resolution, makes in camera higher gears or camera nearer apart from car plate, be resolution when big, can accurately discern car plate; Simultaneously, at camera than low grade or camera apart from car plate when far away, promptly resolution hour still can accurately be discerned car plate; In addition because the environment difference, the camera model is also inequality, how to develop a kind of can widely used Vehicle License Plate Recognition System, also become main direction of studying in the industry, therefore, the system and method that propose a kind ofly to have versatility, can both discern car plate under arbitrary resolution is necessary really.
Summary of the invention
At the deficiencies in the prior art, the invention provides a kind of recognition system and method for car plate, make it can be applicable to various environment, be particularly useful for the identification of car plate under the arbitrary resolution.
For realizing above purpose, the present invention is achieved by the following technical programs:
A kind of recognition methods of car plate may further comprise the steps:
S1, obtain the image that contains car plate and it is carried out filtering;
S2, setting frequency spectrum threshold value select described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate, if, execution in step S3 then, otherwise continue execution in step S2;
Character local feature in S3, the described rectangular area of extraction judges whether described character local feature is consistent; If consistent, execution in step S4 then, otherwise return execution in step S2;
S4, described rectangular area is carried out gray processing, binaryzation, trimming frame and removed gold-plating nail, and the character zone in this rectangular area cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2;
S5, judge whether each character shape after cutting apart identical, each intercharacter linear array whether, if, execution in step S6 then, otherwise return execution in step S2;
S6, the size of described each character is carried out normalization, extract the statistical nature of character and, the output license plate recognition result its classification.
Preferably, described step S2 comprises that further whether the width of judging described rectangular area is twice at least of its height, if, execution in step S3 then, otherwise continue execution in step S2.
Preferably, described step S4 further comprises:
S41, gray processing, binaryzation are carried out in described rectangular area, obtain binary image;
S42, search for described binary image line by line, whether the black and white change frequency of judging every row greater than 12 times, if then this row is a character zone, if not, then this row is carried out the trimming frame and remove the gold-plating nail;
S43, described character zone is cut apart, judged whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2.
Preferably, described step S43 further comprises: the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
Preferably, described step S6 further comprises the size of described each character carried out normalization, extracts the statistical nature of character and it sent into SVM or multilayer neural network is classified, the output license plate recognition result.
A kind of recognition system of car plate includes:
Filter unit is used to obtain the image that contains car plate and it is carried out filtering;
First recognition unit is used to set the frequency spectrum threshold value, selects described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate;
Second recognition unit is used for extracting the character local feature of described rectangular area, judges whether described character local feature is consistent;
The Character segmentation unit is used for described rectangular area is carried out gray processing, binaryzation, trimming frame and removed the gold-plating nail, and the character zone in this rectangular area is cut apart, and judges whether the number of characters after cutting apart satisfies the number of characters of car plate;
The 3rd recognition unit is used to judge whether each character shape after cutting apart is identical, each intercharacter linear array whether;
Output unit is used for the size normalization to described each character, extracts the statistical nature of character and with its classification, exports license plate recognition result.
Preferably, described first recognition unit is further used for judging whether the width of described rectangular area is the twice at least of its height.
Preferably, described Character segmentation unit is further used for the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
Preferably, described output unit is further used for the size of described each character is carried out normalization, extracts the statistical nature of character and it is sent into SVM or multilayer neural network is classified, the output license plate recognition result.
Recognition system and the method for the present invention by a kind of car plate is provided, the car plate that can quick identification goes out arbitrary resolution, all can discern for image and the little image of size that size is big, its recognition speed is fast, the discrimination height, strong to the adaptability of environment, can also be by province priority be set, improve discrimination, can be widely used in intelligent transportation system and the electronic police system.
Description of drawings
Fig. 1 is the process flow diagram of one embodiment of the invention;
Fig. 2 is the system and device figure of one embodiment of the invention.
Embodiment
Regard to the recognition system and the method for a kind of car plate proposed by the invention down, describe in detail in conjunction with the accompanying drawings and embodiments.
The invention provides a kind of recognition methods of car plate, as shown in Figure 1, may further comprise the steps:
S1, obtain the image that contains car plate and it is carried out filtering;
S2, setting frequency spectrum threshold value select described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate, if, execution in step S3 then, otherwise continue execution in step S2;
Character local feature in S3, the described rectangular area of extraction judges whether described character local feature is consistent; If consistent, execution in step S4 then, otherwise return execution in step S2;
S4, described rectangular area is carried out gray processing, binaryzation, trimming frame and removed gold-plating nail, and the character zone in this rectangular area cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2;
S5, judge whether each character shape after cutting apart identical, each intercharacter linear array whether, if, execution in step S6 then, otherwise return execution in step S2; Because character will satisfy certain length breadth ratio, each character boundary of a car plate should be identical;
S6, the size of described each character is carried out normalization, extract the statistical nature of character and, the output license plate recognition result its classification.
Preferably, described step S2 comprises that further whether the width of judging described rectangular area is twice at least of its height, if, execution in step S3 then, otherwise continue execution in step S2.
Character local feature among the described step S3 can be the stroke width of character, can obtain character stroke by the algorithm of edge extracting, calculates the width of stroke again with vertical information, has only each character stroke width unanimity, could be as license plate area.
Preferably, described step S4 further comprises:
S41, gray processing, binaryzation are carried out in described rectangular area, obtain binary image;
S42, search for described binary image line by line, whether the black and white change frequency of judging every row greater than 12 times, if then this row is a character zone, if not, then this row is carried out the trimming frame and remove the gold-plating nail;
S43, described character zone is cut apart, judged whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2.
Preferably, described step S43 further comprises: the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
After license plate binary, carry out vertical projection, promptly add up the number of each row white or black pixel point, and draw a histogram, the part of each character correspondence will have very big numerical value, where character can be judged with this characteristic,, a character can be determined again in conjunction with the width and the elevation information of character;
Because each numeral and capitalization English letter all are connected regions,, utilize the method for mark expansion to obtain connected region, and then judge the position at character place according to this characteristic;
Characters on license plate is arranged certain rule, a point is arranged in the middle of the car plate of single character and 7 above characters are arranged, and five words below general top two words of the car plate of double character with such situation template, can obtain the position of each character exactly.
Preferably, described step S6 further comprises the size of described each character carried out normalization, extracts the statistical nature of character and it sent into SVM or multilayer neural network is classified, the output license plate recognition result.
As shown in Figure 2, the present invention also provides a kind of recognition system of car plate, includes:
Filter unit is used to obtain the image that contains car plate and it is carried out filtering;
First recognition unit is used to set the frequency spectrum threshold value, selects described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate;
Second recognition unit is used for extracting the character local feature of described rectangular area, judges whether described character local feature is consistent;
The Character segmentation unit is used for described rectangular area is carried out gray processing, binaryzation, trimming frame and removed the gold-plating nail, and the character zone in this rectangular area is cut apart, and judges whether the number of characters after cutting apart satisfies the number of characters of car plate;
The 3rd recognition unit is used to judge whether each character shape after cutting apart is identical, each intercharacter linear array whether;
Output unit is used for the size normalization to described each character, extracts the statistical nature of character and with its classification, exports license plate recognition result.
Preferably, described first recognition unit is further used for judging whether the width of described rectangular area is the twice at least of its height.
Preferably, described Character segmentation unit is further used for the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
Preferably, described output unit is further used for the size of described each character is carried out normalization, extracts the statistical nature of character and it is sent into SVM or multilayer neural network is classified, the output license plate recognition result.
By setting up recognizer, RGB that contains license plate image or jpeg data are input in the function, just can obtain to contain the recognition result of the number-plate number, car plate color, car plate particular location, this system and method can be discerned the image of arbitrary resolution, its recognition speed is very fast, usually about 30ms, discrimination is more than 98%, can accurately discern blueness, black, white and yellow four kinds of car plates, can discern the whole nation car plate in province arbitrarily, more can improve discrimination by province priority is set
Recognition system and the method for the present invention by a kind of car plate is provided, the car plate that can quick identification goes out arbitrary resolution, all can discern for image and the little image of size that size is big, its recognition speed is fast, the discrimination height, strong to the adaptability of environment, can also be by province priority be set, improve discrimination, can be widely used in intelligent transportation system and the electronic police system.
Above embodiment only is used to illustrate the present invention; and be not limitation of the present invention; the those of ordinary skill in relevant technologies field; under the situation that does not break away from the spirit and scope of the present invention; can also make various variations and modification; therefore all technical schemes that are equal to also belong to category of the present invention, and scope of patent protection of the present invention should be defined by the claims.
Claims (9)
1. the recognition methods of a car plate is characterized in that, may further comprise the steps:
S1, obtain the image that contains car plate and it is carried out filtering;
S2, setting frequency spectrum threshold value select described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate, if, execution in step S3 then, otherwise continue execution in step S2;
Character local feature in S3, the described rectangular area of extraction judges whether described character local feature is consistent; If consistent, execution in step S4 then, otherwise return execution in step S2;
S4, described rectangular area is carried out gray processing, binaryzation, trimming frame and removed gold-plating nail, and the character zone in this rectangular area cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2;
S5, judge whether each character shape after cutting apart identical, each intercharacter linear array whether, if, execution in step S6 then, otherwise return execution in step S2;
S6, the size of described each character is carried out normalization, extract the statistical nature of character and, the output license plate recognition result its classification.
2. the method for claim 1 is characterized in that, described step S2 comprises that further whether the width of judging described rectangular area is twice at least of its height, if, execution in step S3 then, otherwise continue execution in step S2.
3. the method for claim 1 is characterized in that, described step S4 further comprises:
S41, gray processing, binaryzation are carried out in described rectangular area, obtain binary image;
S42, search for described binary image line by line, whether the black and white change frequency of judging every row greater than 12 times, if then this row is a character zone, if not, then this row is carried out the trimming frame and remove the gold-plating nail;
S43, described character zone is cut apart, judged whether the number of characters after cutting apart satisfies the number of characters of car plate, if, execution in step S5 then, otherwise return execution in step S2.
4. method as claimed in claim 3, it is characterized in that, described step S43 further comprises: the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
5. the method for claim 1 is characterized in that, described step S6 further comprises the size of described each character is carried out normalization, extracts the statistical nature of character and it is sent into SVM or multilayer neural network is classified, the output license plate recognition result.
6. the recognition system of a car plate is characterized in that, includes:
Filter unit is used to obtain the image that contains car plate and it is carried out filtering;
First recognition unit is used to set the frequency spectrum threshold value, selects described vision intermediate frequency spectrum to surpass this threshold value and the color rectangular area identical with the car plate color; Judge whether the shape of described rectangular area meets the shape of car plate;
Second recognition unit is used for extracting the character local feature of described rectangular area, judges whether described character local feature is consistent;
The Character segmentation unit is used for described rectangular area is carried out gray processing, binaryzation, trimming frame and removed the gold-plating nail, and the character zone in this rectangular area is cut apart, and judges whether the number of characters after cutting apart satisfies the number of characters of car plate;
The 3rd recognition unit is used to judge whether each character shape after cutting apart is identical, each intercharacter linear array whether;
Output unit is used for the size normalization to described each character, extracts the statistical nature of character and with its classification, exports license plate recognition result.
7. system as claimed in claim 6 is characterized in that, described first recognition unit is further used for judging whether the width of described rectangular area is the twice at least of its height.
8. system as claimed in claim 6, it is characterized in that, described Character segmentation unit is further used for the binary image of described character zone correspondence is carried out vertical projection, regional connectivity and situation template comparison, determine the position of each character, and described character zone cut apart, judge whether the number of characters after cutting apart satisfies the number of characters of car plate.
9. system as claimed in claim 6 is characterized in that, described output unit is further used for the size of described each character is carried out normalization, extracts the statistical nature of character and it is sent into SVM or multilayer neural network is classified, the output license plate recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310117722.0A CN103226696B (en) | 2013-04-07 | 2013-04-07 | The identification system and method for car plate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310117722.0A CN103226696B (en) | 2013-04-07 | 2013-04-07 | The identification system and method for car plate |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103226696A true CN103226696A (en) | 2013-07-31 |
| CN103226696B CN103226696B (en) | 2016-07-06 |
Family
ID=48837137
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310117722.0A Expired - Fee Related CN103226696B (en) | 2013-04-07 | 2013-04-07 | The identification system and method for car plate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103226696B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103902981A (en) * | 2014-04-02 | 2014-07-02 | 浙江师范大学 | Method and system for identifying license plate characters based on character fusion features |
| CN103971126A (en) * | 2014-05-12 | 2014-08-06 | 百度在线网络技术（北京）有限公司 | Method and device for identifying traffic signs |
| CN105005757A (en) * | 2015-03-12 | 2015-10-28 | 电子科技大学 | Method for recognizing license plate characters based on Grassmann manifold |
| WO2018090771A1 (en) * | 2016-11-16 | 2018-05-24 | 杭州海康威视数字技术股份有限公司 | Vehicle license plate recognition method and apparatus |
| CN108073926A (en) * | 2016-11-17 | 2018-05-25 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method and device |
| CN108090484A (en) * | 2016-11-23 | 2018-05-29 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method and device |
| CN108288403A (en) * | 2018-01-31 | 2018-07-17 | 中国地质大学（武汉） | Parking management system based on intelligent space lock |
| CN108805008A (en) * | 2018-04-19 | 2018-11-13 | 江苏理工学院 | A kind of community's vehicle security system based on deep learning |
| CN109978132A (en) * | 2018-12-24 | 2019-07-05 | 中国科学院深圳先进技术研究院 | A kind of neural network method and system refining vehicle identification |
| CN112950950A (en) * | 2021-01-26 | 2021-06-11 | 上海启迪睿视智能科技有限公司 | Parking auxiliary device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| US20130022234A1 (en) * | 2011-07-22 | 2013-01-24 | Honeywell International Inc. | Object tracking |
-
2013
- 2013-04-07 CN CN201310117722.0A patent/CN103226696B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
| US20130022234A1 (en) * | 2011-07-22 | 2013-01-24 | Honeywell International Inc. | Object tracking |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103902981A (en) * | 2014-04-02 | 2014-07-02 | 浙江师范大学 | Method and system for identifying license plate characters based on character fusion features |
| CN103971126A (en) * | 2014-05-12 | 2014-08-06 | 百度在线网络技术（北京）有限公司 | Method and device for identifying traffic signs |
| CN103971126B (en) * | 2014-05-12 | 2017-08-08 | 百度在线网络技术（北京）有限公司 | A kind of traffic sign recognition method and device |
| CN105005757A (en) * | 2015-03-12 | 2015-10-28 | 电子科技大学 | Method for recognizing license plate characters based on Grassmann manifold |
| CN105005757B (en) * | 2015-03-12 | 2018-04-06 | 电子科技大学 | A kind of license plate character recognition method popular based on Grassmann |
| WO2018090771A1 (en) * | 2016-11-16 | 2018-05-24 | 杭州海康威视数字技术股份有限公司 | Vehicle license plate recognition method and apparatus |
| CN108073926A (en) * | 2016-11-17 | 2018-05-25 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method and device |
| CN108090484A (en) * | 2016-11-23 | 2018-05-29 | 杭州海康威视数字技术股份有限公司 | A kind of licence plate recognition method and device |
| CN108288403A (en) * | 2018-01-31 | 2018-07-17 | 中国地质大学（武汉） | Parking management system based on intelligent space lock |
| CN108805008A (en) * | 2018-04-19 | 2018-11-13 | 江苏理工学院 | A kind of community's vehicle security system based on deep learning |
| CN109978132A (en) * | 2018-12-24 | 2019-07-05 | 中国科学院深圳先进技术研究院 | A kind of neural network method and system refining vehicle identification |
| CN112950950A (en) * | 2021-01-26 | 2021-06-11 | 上海启迪睿视智能科技有限公司 | Parking auxiliary device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103226696B (en) | 2016-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103226696A (en) | License plate recognition system and license plate recognition method | |
| CN108073928B (en) | License plate recognition method and device | |
| CN101872416B (en) | Vehicle license plate recognition method and system of road image | |
| Saleem et al. | Automatic license plate recognition using extracted features | |
| CN104951784B (en) | A kind of vehicle is unlicensed and license plate shading real-time detection method | |
| CN103136528B (en) | A kind of licence plate recognition method based on dual edge detection | |
| CN106650553A (en) | License plate recognition method and system | |
| CN108268867B (en) | License plate positioning method and device | |
| CN103413147A (en) | Vehicle license plate recognizing method and system | |
| Jusoh et al. | Application of freeman chain codes: An alternative recognition technique for Malaysian car plates | |
| CN104156704A (en) | Novel license plate identification method and system | |
| CN103207992A (en) | Character and color combined recognition method of license plates | |
| CN108229466B (en) | License plate recognition method and device | |
| CN107301418A (en) | Printed page analysis in optical character identification | |
| KR101246120B1 (en) | A system for recognizing license plate using both images taken from front and back faces of vehicle | |
| CN103390167A (en) | Multi-characteristic layered traffic sign identification method | |
| CN104200207A (en) | License plate recognition method based on Hidden Markov models | |
| CN101727581B (en) | Plate number tilt correcting method based on character pre-cut | |
| CN109657664A (en) | A kind of recognition methods, device and the electronic equipment of license plate type | |
| Shah et al. | OCR-based chassis-number recognition using artificial neural networks | |
| Baten et al. | Bangla license plate reader for metropolitan cities of Bangladesh using template matching | |
| Ingole et al. | Characters feature based Indian vehicle license plate detection and recognition | |
| CN104463134A (en) | License plate detection method and system | |
| Mammeri et al. | MSER-based text detection and communication algorithm for autonomous vehicles | |
| KR101794632B1 (en) | The logo pattern recognizing method in number plate of electric vehicle |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160706 Termination date: 20190407 |