CN101882375A - Vehicle license plate filtering method and system - Google Patents
Vehicle license plate filtering method and system Download PDFInfo
- Publication number
- CN101882375A CN101882375A CN2009100505289A CN200910050528A CN101882375A CN 101882375 A CN101882375 A CN 101882375A CN 2009100505289 A CN2009100505289 A CN 2009100505289A CN 200910050528 A CN200910050528 A CN 200910050528A CN 101882375 A CN101882375 A CN 101882375A
- Authority
- CN
- China
- Prior art keywords
- license plate
- candidate
- picture
- plate area
- neural network
- 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.)
- Pending
Links
Images
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to vehicle license plate filtering method and system. The method comprises the following steps of: (1) license plate locating: carrying out feature locating on the input picture and searching for candidate license plate regions; (2) feature extraction of candidate license plates: extracting the feature values of the candidate license plate regions and judging whether the deviations of the extracted feature values from the normal value are too large, if so, filtering out the candidate license plate regions, and if not, entering step (3); and (3) artificial neural network: forming a vector quantity by an artificial neural network module by using the extracted feature value and inputting the vector quantity to the neural network, outputting a probable value by the neural network, if an output probable value in the picture is greater than the candidate license plate region of a certain threshold, judging that the candidate license plate region has a license plate and storing the picture, and if no output probable value in the picture is greater than the candidate license plate region of a certain threshold, filtering out the picture. The invention has the advantages of high speed and low internal memory occupation.
Description
Technical field
The present invention relates to the vehicle license plate filtering technical field, particularly a kind of vehicle license plate filtering method and system.
Background technology
In the real-time traffic Video Detection because the influence of various complex environment factors in the restriction of technical development at present and the practical application, cause capture and have a large amount of useless sheets that does not have vehicle license in the picture.Because the quantity of useless sheet is bigger, makes the validity of data reduce greatly, the while has also been caused unnecessary interference to user's use.
In order to filter the useless sheet in the traffic video detection fast, expeditiously, the licence plate filtering technique arises at the historic moment.The licence plate filtering technique is in the technical development of license plate identification, has points of resemblance with the license plate identification technology, and the characteristics of oneself are also arranged.Compare with the license plate identification technology, the licence plate filtering technique at aspect such as licence plate location, the identification of licence plate color, licence plate feature extraction and license plate identification technology type seemingly, but lack technology such as licence plate classification of type, Character segmentation, character recognition.Simultaneously, licence plate filters does not need to locate very accurately licence plate.
Existing licence plate filtering technique is to adopt license plate identification to realize.Though this class methods accuracy rate height has taken a large amount of system resource, to the performance requirement height of system.Under the certain condition of system resource, use license plate identification to realize vehicle license plate filtering, can influence the overall performance of system.
Summary of the invention
The objective of the invention is to, a kind of vehicle license plate filtering method and system are provided, in the time of with occupying system resources as few as possible, can filter vehicle license fast and efficiently again.
The present invention adopts following technical scheme:
A kind of vehicle license plate filtering method may further comprise the steps:
1) licence plate positioning step is used for the picture of input is carried out feature location, seeks candidate's license plate area;
2) candidate's licence plate characteristic extraction step is used for candidate's license plate area eigenwert is extracted, and whether deviation is excessive to judge the eigenwert extracted and standard value, if, then filter this and select license plate area, if not, then enter step 3);
3) artificial neural network step, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, neural network is output as a probable value, if there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, and preserve this picture; If do not have the candidate license plate area of the probable value of output in the picture, then filter this picture greater than some threshold values.
Further, described step 2) in, the feature extraction of candidate's licence plate comprises the extraction to color characteristic, textural characteristics, gray distribution features.
Further, described step 1) specifically may further comprise the steps:
Picture to input carries out feature location, seeks candidate's license plate area, if there is no candidate's license plate area, and this picture of direct filtration then, and the picture of next input handled, if there is candidate's license plate area, then enter step 2).
Further, described step 2) specifically may further comprise the steps:
21) eigenwert of candidate's license plate area of Ti Quing, whether deviation is excessive for eigenwert that judgement is extracted and standard value, if enter step 22); If not, change step 23);
22) judge that this candidate's license plate area is pseudo-licence plate, judge whether also to exist candidate's license plate area of not judging,, then change step 21) if having; If do not have, then change step 1);
24) judge that might there be licence plate in this candidate's license plate area, then enters step 3);
Further, described step 3) specifically may further comprise the steps:
The artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, and neural network is output as a probable value; If there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, this picture is effective, preserves this picture, changes step 1); Otherwise, judge that then this candidate's license plate area is pseudo-licence plate, and judge whether to also have candidate's license plate area of not judging, if having, then change step 2), if do not have, then filter this picture, change step 1).
The present invention also provides a kind of vehicle license plate filtering system, comprising:
The licence plate locating module is used for the picture of input is carried out feature location, seeks candidate's license plate area;
Candidate's licence plate characteristic extracting module is used for candidate's license plate area eigenwert is extracted, and whether deviation is excessive with standard value to judge the eigenwert of extracting, if, then filter this and select license plate area, if not, the eigenwert of extraction is sent into the artificial neural network module;
The artificial neural network module, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, neural network is output as a probable value, if there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, preserve this picture,, then filter this picture if do not have candidate's license plate area of the probable value of output in the picture greater than some threshold values.
Vehicle license plate filtering method provided by the invention and system, hand by image being carried out Flame Image Process such as feature location, color judgement, feature extraction and in conjunction with artificial neural network publishes picture whether contain vehicle license in the picture with the highest probabilistic determination in the shortest time.The scale and the quantity of its neural network are little more than license plate identification, and it has the following advantages:
1, speed is very fast.The chances are 30% of the existing license plate identification method consuming time of handling under identical environment that identical picture licence plate filters.
2, the EMS memory occupation amount is low.Under identical environment, handle identical picture licence plate and filter shared internal memory the chances are 35% of existing license plate identification method.
Further specify the present invention below in conjunction with drawings and Examples.
Description of drawings
Fig. 1 is a vehicle license plate filtering method embodiment process flow diagram of the present invention.
Embodiment
A kind of vehicle license plate filtering system comprises:
The licence plate locating module is used for the picture of input is carried out feature location, seeks candidate's license plate area;
Candidate's licence plate characteristic extracting module is used for candidate's license plate area eigenwert is extracted, and whether deviation is excessive with standard value to judge the eigenwert of extracting, if, then filter this and select license plate area, if not, the eigenwert of extraction is sent into the artificial neural network module;
The artificial neural network module, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, neural network is output as a probable value, if there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, preserve this picture,, then filter this picture if do not have candidate's license plate area of the probable value of output in the picture greater than some threshold values.
As shown in Figure 1, a kind of vehicle license plate filtering method may further comprise the steps:
1) licence plate positioning step is used for the picture of input is carried out feature location, seeks candidate's license plate area as much as possible, if there is no candidate's license plate area, this picture of direct filtration then, and the picture of next input handled, if there is candidate's license plate area, then enter step 2);
2) candidate's licence plate characteristic extraction step is chosen candidate's license plate area, and features such as its color characteristic, textural characteristics, intensity profile are extracted and handled.Because features such as the color characteristic of standard deck photograph, textural characteristics, intensity profile have clear regularity, it specifically may further comprise the steps:
21) eigenwert of candidate's license plate area of Ti Quing, whether deviation is excessive for eigenwert that judgement is extracted and standard value, if enter step 22); If not, change step 23);
22) judge that this candidate's license plate area is pseudo-licence plate, judge whether also to exist candidate's license plate area of not judging,, then change step 21) if having; If do not have, then change step 1);
24) judge that might there be licence plate in this candidate's license plate area, then enters step 3);
3) artificial neural network step, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, and neural network is output as a probable value, as the probable value between 0~1; If this probable value of output judges then that greater than some threshold values there is licence plate in this candidate's license plate area, this picture is effective, preserves this picture, changes step 1); Otherwise, judge that then this candidate's license plate area is pseudo-licence plate, and judge whether to also have candidate's license plate area of not judging, if having, then change step 2), if do not have, then filter this picture, change step 1).
Above-described embodiment only is used to illustrate technological thought of the present invention and characteristics, its purpose makes those skilled in the art can understand content of the present invention and is implementing according to this, when can not only limiting claim of the present invention with present embodiment, no matter adopt which kind of existing artificial neural network technology, be all equal variation or modifications of doing according to disclosed spirit, still drop in the claim of the present invention.
Claims (6)
1. vehicle license plate filtering method is characterized in that may further comprise the steps:
1) licence plate positioning step is used for the picture of input is carried out feature location, seeks candidate's license plate area;
2) candidate's licence plate characteristic extraction step is used for candidate's license plate area eigenwert is extracted, and whether deviation is excessive to judge the eigenwert extracted and standard value, if, then filter this and select license plate area, if not, then enter step 3);
3) artificial neural network step, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, probable value of neural network output, if there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, and preserve this picture; If do not have the candidate license plate area of the probable value of output in the picture, then filter this picture greater than some threshold values.
2. vehicle license plate filtering method according to claim 1 is characterized in that:
Described step 2) in, the feature extraction of candidate's licence plate comprises the extraction to color characteristic, textural characteristics, gray distribution features.
3. vehicle license plate filtering method according to claim 1 and 2 is characterized in that: described step 1) specifically may further comprise the steps:
Picture to input carries out feature location, seeks candidate's license plate area, if there is no candidate's license plate area, and this picture of direct filtration then, and the picture of next input handled, if there is candidate's license plate area, then enter step 2);
4. vehicle license plate filtering method according to claim 3 is characterized in that: described step 2) specifically may further comprise the steps:
21) eigenwert of candidate's license plate area of Ti Quing, whether deviation is excessive for eigenwert that judgement is extracted and standard value, if enter step 22); If not, change step 23);
22) judge that this candidate's license plate area is pseudo-licence plate, judge whether also to exist candidate's license plate area of not judging,, then change step 21) if having; If do not have, then change step 1);
24) judge that might there be licence plate in this candidate's license plate area, then enters step 3).
5. vehicle license plate filtering method according to claim 4 is characterized in that: described step 3) specifically may further comprise the steps:
The artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, and neural network is output as a probable value; If this probable value of output judges then that greater than some threshold values there is licence plate in this candidate's license plate area, this picture is effective, preserves this picture, changes step 1); Otherwise, judge that then this candidate's license plate area is pseudo-licence plate, and judge whether to also have candidate's license plate area of not judging, if having, then change step 2), if do not have, then filter this picture, change step 1).
6. vehicle license plate filtering system is characterized in that comprising:
The licence plate locating module is used for the picture of input is carried out feature location, seeks candidate's license plate area;
Candidate's licence plate characteristic extracting module is used for candidate's license plate area eigenwert is extracted, and whether deviation is excessive with standard value to judge the eigenwert of extracting, if, then filter this and select license plate area, if not, the eigenwert of extraction is sent into the artificial neural network module;
The artificial neural network module, the artificial neural network module forms the eigenwert of extracting a vector and inputs to neural network, neural network is output as a probable value, if there be the candidate license plate area of the probable value of output in the picture greater than some threshold values, judge that then there is licence plate in this candidate's license plate area, preserve this picture,, then filter this picture if do not have candidate's license plate area of the probable value of output in the picture greater than some threshold values.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009100505289A CN101882375A (en) | 2009-05-04 | 2009-05-04 | Vehicle license plate filtering method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009100505289A CN101882375A (en) | 2009-05-04 | 2009-05-04 | Vehicle license plate filtering method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101882375A true CN101882375A (en) | 2010-11-10 |
Family
ID=43054382
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2009100505289A Pending CN101882375A (en) | 2009-05-04 | 2009-05-04 | Vehicle license plate filtering method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101882375A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103390156A (en) * | 2012-11-05 | 2013-11-13 | 深圳市捷顺科技实业股份有限公司 | License plate recognition method and device |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0862132A2 (en) * | 1997-03-01 | 1998-09-02 | Institute of Systems Science | Robust identification code recognition system |
US6281928B1 (en) * | 1998-05-13 | 2001-08-28 | Chuo Hatsujo Kabushiki Kaisha | Positional detector device for a vehicular license plate |
US6373962B1 (en) * | 1998-04-24 | 2002-04-16 | Ume Tech Yugen Kaisha | License plate information reader device for motor vehicles |
CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
CN101408933A (en) * | 2008-05-21 | 2009-04-15 | 浙江师范大学 | Method for recognizing license plate character based on wide gridding characteristic extraction and BP neural network |
CN101408942A (en) * | 2008-04-17 | 2009-04-15 | 浙江师范大学 | Method for locating license plate under a complicated background |
-
2009
- 2009-05-04 CN CN2009100505289A patent/CN101882375A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0862132A2 (en) * | 1997-03-01 | 1998-09-02 | Institute of Systems Science | Robust identification code recognition system |
US6373962B1 (en) * | 1998-04-24 | 2002-04-16 | Ume Tech Yugen Kaisha | License plate information reader device for motor vehicles |
US6281928B1 (en) * | 1998-05-13 | 2001-08-28 | Chuo Hatsujo Kabushiki Kaisha | Positional detector device for a vehicular license plate |
CN101408942A (en) * | 2008-04-17 | 2009-04-15 | 浙江师范大学 | Method for locating license plate under a complicated background |
CN101408933A (en) * | 2008-05-21 | 2009-04-15 | 浙江师范大学 | Method for recognizing license plate character based on wide gridding characteristic extraction and BP neural network |
CN101398894A (en) * | 2008-06-17 | 2009-04-01 | 浙江师范大学 | Automobile license plate automatic recognition method and implementing device thereof |
Non-Patent Citations (2)
Title |
---|
刘晓芳,程丹松,刘家锋,管宁: "采用改进HSI模型的车牌区域检测和定位方法", 《哈尔滨工业大学学报》 * |
郭捷,施鹏飞: "基于颜色和纹理分析的车牌定位方法", 《中国图象图形学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103390156A (en) * | 2012-11-05 | 2013-11-13 | 深圳市捷顺科技实业股份有限公司 | License plate recognition method and device |
CN103390156B (en) * | 2012-11-05 | 2016-09-21 | 深圳市捷顺科技实业股份有限公司 | A kind of licence plate recognition method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI497422B (en) | A system and method for recognizing license plate image | |
US9256783B2 (en) | Systems and methods for tax data capture and use | |
WO2020107866A1 (en) | Text region obtaining method and apparatus, storage medium and terminal device | |
CN104616021B (en) | Traffic sign image processing method and device | |
CN110598800A (en) | Garbage classification and identification method based on artificial intelligence | |
CN108052931B (en) | License plate recognition result fusion method and device | |
CN112487848B (en) | Character recognition method and terminal equipment | |
KR101246120B1 (en) | A system for recognizing license plate using both images taken from front and back faces of vehicle | |
CN109934077B (en) | Image identification method and electronic equipment | |
WO2020258714A1 (en) | Rider re-identification method, apparatus and device | |
CN104951553B (en) | A kind of accurate content of data processing is collected and data mining platform and its implementation | |
CN103226696A (en) | License plate recognition system and license plate recognition method | |
CN106228106B (en) | A kind of improved real-time vehicle detection filter method and system | |
CN106921804A (en) | Method, device and the terminal device of schedule are created in the terminal | |
CN103164697A (en) | Processing time and recognition precision self-adaption plate number recognition method | |
CN109447117A (en) | The double-deck licence plate recognition method, device, computer equipment and storage medium | |
CN110135514A (en) | A kind of workpiece classification method, device, equipment and medium | |
CN107133574B (en) | Vehicle feature identification method and device | |
CN107657251A (en) | Determine the device and method of identity document display surface, image-recognizing method | |
CN112597995B (en) | License plate detection model training method, device, equipment and medium | |
CN111401438B (en) | Image sorting method, device and system | |
WO2017069741A1 (en) | Digitized document classification | |
JP6089698B2 (en) | Information processing apparatus and method | |
CN101882375A (en) | Vehicle license plate filtering method and system | |
CN111178398B (en) | Method, system, storage medium and device for detecting tampering of identity card image information |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20101110 |