CN103279739B - A kind of deck detection method based on vehicle characteristics coupling - Google Patents
A kind of deck detection method based on vehicle characteristics coupling Download PDFInfo
- Publication number
- CN103279739B CN103279739B CN201310175763.5A CN201310175763A CN103279739B CN 103279739 B CN103279739 B CN 103279739B CN 201310175763 A CN201310175763 A CN 201310175763A CN 103279739 B CN103279739 B CN 103279739B
- Authority
- CN
- China
- Prior art keywords
- channel value
- belong
- pixel
- current
- bin
- 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.)
- Expired - Fee Related
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 11
- 230000008878 coupling Effects 0.000 title claims abstract description 9
- 238000010168 coupling process Methods 0.000 title claims abstract description 9
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 9
- 238000010606 normalization Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 12
- 210000000746 body region Anatomy 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
Abstract
The present invention relates to a kind of deck detection method based on vehicle characteristics coupling. The present invention is by using the priori car plate distribution rule on vehicle, and vehicle image, according to the corresponding ratio normalization of car plate, and is selected to the headstock region that car plate is corresponding, generates corresponding headstock feature with this. By the comparison of property data base, obtain the whether corresponding vehicle characteristics of car plate in matching database of current feature, the present invention, in the case of the exact position of known car plate, can utilize headstock feature, accurately judges the similarity of headstock in headstock to be judged and database. And be subject to environmental factor little, needn't be subject to car to identify the restriction of other kind.
Description
Technical field
The invention belongs to intelligent transport technology field, relate to a kind of deck detection method based on vehicle characteristics coupling.
Background technology
In industry, to fake-licensed car, identification is mainly to be judged and determined whether vehicle has deck suspicion by car mark at present. At presentIt is very high that this existing method identifies other requirement to car, and car target is of a great variety, even the car mark of a brand alsoHave various ways, and car target kind is also in continuous increase. General car mark recognizer can only be tolerated approximately 20 kinds of car targetsIdentification, detects and has brought very large difficulty to fake-licensed car.
Summary of the invention
The invention provides a kind of deck detection method based on vehicle matching characteristic, by using priori car plate at vehicleOn distribution rule, vehicle image, according to the corresponding ratio normalization of car plate, and is selected to the headstock region that car plate is corresponding, with thisGenerate corresponding headstock feature. By the comparison of property data base, obtain current feature whether in matching database car plate institute rightThe vehicle characteristics of answering, this method has overcome general car mark recognizer and has been suitable for restriction, thereby has improved the accurate of deck detectionProperty.
A deck detection method based on vehicle matching characteristic, comprising:
(1) the car plate width and the high computational headstock region R that utilize car plate to identify to obtain. Headstock region is divided into 15Sub-block。
(2) to vehicle body image be normalized, the pretreatment of sharpen edges and removal noise.
(3) calculate each sub-blockHsv color space, the color of each sub-block is carried out to statistics with histogram.
(4) calculate each sub-blockN key pointTextural characteristics.
(5) compare the characteristic sequence of current headstockHeadstock characteristic sequence with corresponding license plate number in database. If both covariances (Mahalanobis) distance is less than 0.4, this car does not have deck, otherwise is deckVehicle.
(1) described car plate width and the high computational headstock region R that utilizes car plate to identify and obtain, specifically comprises: according to carThe origin coordinates of board in the middle of imageWith car plate width, the relative position of calculating headstock.
;
;
;
;
(The origin coordinates in vehicle body region,The width of vehicle body,The height of vehicle body).
(2) described to vehicle body image be normalized, the pretreatment of sharpen edges and removal noise, specifically comprise followingStep:
1) adopt known bilinear interpolation method, by vehicle body image normalization extremely。
2) use known gaussian filtering, the license plate image after normalization is transported to the processing of row sharpen edges.
3) adopt known Gassian low-pass filter method, the license plate image after sharpen edges removed to noise placeReason.
(3) calculate each sub-blockHsv color space, the color of each sub-block is carried out to statistics with histogram method bagDraw together:
1) adopt known RGB to turn HSV algorithm and calculate the each sub-block in the R of headstock regionDo color conversion.
2) to each sub-blockCarry out statistics with histogram, statistic histogram is divided into 10 class Bin[i] [10], i[0,14], each pixel whereinAccording to following regular partition:
If a) current pixelV channel value, current pixelBelong to Bin[0];
If b) current pixelV channel value, and S channel value，Current pixelBelong to Bin[1];
If c) current pixelV channel value, and S channel value, whenPreceding pixelBelong to Bin[2];
If d) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[3];
If e) current pixelH channel value, and S, V channel value belong to (0.2,1], whenPreceding pixelBelong to Bin[4];
If f) current pixelH channel value, and S, V channel value belong to (0.2,1], whenPreceding pixelBelong to Bin[5];
If g) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[6];
If h) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[7];
If i) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[8];
If j) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[9];
All sub-blocksBin[i] [10] totally 150 dimensional features.
(4) calculate texture histogram feature and comprise, adopt the each sub-block of known sift feature calculationN key point128 dimensional features.
(5) compare the characteristic sequence of current headstockHeadstock characteristic sequence with corresponding license plate number in database. Comprise 150 dimension color characteristic Bin[i] [10] andIndividual sift Feature Combination becomes union feature. In existing property data base, search headstock feature corresponding to this license plate number, calculateWithMahalanobis distance. If both are less than 0.4 at distance, this car does not have deck, otherwise is fake-licensed car.
Beneficial effect of the present invention: in the case of the exact position of known car plate, can utilize headstock feature, accurately sentenceBreak and the similarity of headstock in headstock to be judged and database. And be subject to environmental factor little, needn't be subject to car to identify the restriction of other kind.Apply to intelligent transportation, distinguish for fake-licensed car the guarantee that provides favourable, have wide practical use.
Brief description of the drawings
Fig. 1 is method flow diagram of the present invention.
Detailed description of the invention
Below in conjunction with accompanying drawing, the present invention is further illustrated.
As shown in Figure 1, a kind of deck detection method based on vehicle characteristics coupling, comprising:
(1) the car plate width and the high computational headstock region R that utilize car plate to identify to obtain, according to car plate in the middle of imageOrigin coordinatesWith car plate width, the relative position of calculating headstock.
;
;
;
;
(The origin coordinates in vehicle body region,The width of vehicle body,The height of vehicle body)
(2) to vehicle body image be normalized, the pretreatment of sharpen edges and removal noise, comprising:
1) adopt known bilinear interpolation method, by vehicle body image normalization extremely。
2) use known gaussian filtering, the license plate image after normalization is transported to the processing of row sharpen edges.
3) adopt known Gassian low-pass filter method, the license plate image after sharpen edges removed to noise placeReason.
(3) calculate each sub-blockHsv color space, the color of each sub-block is carried out to statistics with histogram method bagDraw together:
1) adopt known RGB to turn HSV algorithm and calculate the each sub-block in the R of headstock regionDo color conversion.
2) to each sub-blockCarry out statistics with histogram, statistic histogram is divided into 10 class Bin[i] [10], i[0,14], each pixel whereinAccording to following regular partition:
If a) current pixelV channel value, current pixelBelong to Bin[0];
If b) current pixelV channel value, and S channel value，Current pixelBelong to Bin[1];
If c) current pixelV channel value, and S channel value, whenPreceding pixelBelong to Bin[2];
If d) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[3];
If e) current pixelH channel value, and S, V channel value belong to (0.2,1], whenPreceding pixelBelong to Bin[4];
If f) current pixelH channel value, and S, V channel value belong to (0.2,1], whenPreceding pixelBelong to Bin[5];
If g) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[6];
If h) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[7];
If i) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[8];
If j) current pixelH channel value, and S, V channel value belong to (0.2,1],Current pixelBelong to Bin[9];
All sub-blocksBin[i] [10] totally 150 dimensional features.
(4) calculate texture histogram feature and comprise, adopt the each sub-block of known sift feature calculationN key point128 dimensional features;
(5) compare the characteristic sequence of current headstockHeadstock characteristic sequence with corresponding license plate number in database. Comprise 150 dimension color characteristic Bin[i] [10] andIndividual sift Feature Combination becomes union feature. In existing property data base, search headstock feature corresponding to this license plate number, calculateWithMahalanobis distance. If both are less than 0.4 at distance, this car does not have deck, otherwise is fake-licensed car.
Claims (4)
1. the deck detection method based on vehicle characteristics coupling, is characterized in that the concrete steps of the method are as follows:
Car plate width and high computational headstock region R that step (1) is utilized car plate to identify to obtain, be divided into 15 by headstock regionSub-block Ai：
Step (2) to vehicle body image be normalized, the pretreatment of sharpen edges and removal noise;
Step (3) is calculated each sub-block AiHsv color space, the color of each sub-block is carried out to statistics with histogram; Specifically:
1) adopt RGB to turn HSV algorithm to the each sub-block A in the R of headstock regioniDo color conversion;
2) to each sub-block AiCarry out statistics with histogram, statistic histogram is divided into 10 class Bin[i] [10], i ∈ [0,14], whereinEach pixel PixeljAccording to following regular partition:
If a) current pixel PixeljV channel value Vj≤ 0.2, current pixel PixeljBelong to Bin[0];
If b) current pixel PixeljV channel value 0.2≤Vj< 0.8, and S channel value 0≤Vj≤ 0.2, current pixel PixeljBelong to Bin[1];
If c) current pixel PixeljV channel value 0.2≤Vj< 1, and S channel value 0.2 < Vj≤ 1, current pixel PixeljBelong toIn Bin[2];
If d) current pixel PixeljH channel value 330 < Hj≤ 22, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[3];
If e) current pixel PixeljH channel value 22 < Hj≤ 45, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[4];
If f) current pixel PixeljH channel value 45 < Hj≤ 70, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[5];
If g) current pixel PixeljH channel value 70 < Hj≤ 155, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[6];
If h) current pixel PixeljH channel value 155 < Hj≤ 186, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[7];
If i) current pixel PixeljH channel value 186 < Hj≤ 278, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[8];
If j) current pixel PixeljH channel value 278 < Hj≤ 330, and S, V channel value belong to (0.2,1], current pixelPixeljBelong to Bin[9];
All sub-block AiBin[i] [10] totally 150 dimension color characteristics;
Step (4) is calculated each sub-block AiN key point FnTextural characteristics;
Step (5) is compared the headstock characteristic sequence of corresponding license plate number in the characteristic sequence FA of current headstock and databaseTemplateFA, if both covariances distance is less than 0.4, this car does not have deck, otherwise is deck vehicle.
2. a kind of deck detection method based on vehicle characteristics coupling according to claim 1, is characterized in that: step(1) in, headstock region R is by the Relative position determination of headstock, the specifically origin coordinates (X in the middle of image according to car plateplate,Yplate) and car plate width Wplate, can calculate the relative position of headstock;
XRoi＝Xplate-2×Wplate；
YRoi＝Yplate-2×Wplate；
WRoi＝5×Wplate；
HRoi＝3×Wplate；
Wherein (XRoi,YRoi) represent the origin coordinates in vehicle body region, WRoiRepresent the width of vehicle body, HRoiRepresent the height of vehicle body.
3. a kind of deck detection method based on vehicle characteristics coupling according to claim 1, is characterized in that: step(2) specifically:
1) adopt bilinear interpolation method, by vehicle body image normalization to WRoi＝150,HRoi＝90；
2) adopt gaussian filtering, the license plate image after normalization is transported to the processing of row sharpen edges;
3) adopt Gassian low-pass filter method, to the noise processed of removing of the license plate image after sharpen edges.
4. a kind of deck detection method based on vehicle characteristics coupling according to claim 1, is characterized in that: step(4) specifically: adopt the each sub-block A of sift feature calculationiN key point Fn128 dimensional features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310175763.5A CN103279739B (en) | 2013-05-10 | 2013-05-10 | A kind of deck detection method based on vehicle characteristics coupling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310175763.5A CN103279739B (en) | 2013-05-10 | 2013-05-10 | A kind of deck detection method based on vehicle characteristics coupling |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103279739A CN103279739A (en) | 2013-09-04 |
| CN103279739B true CN103279739B (en) | 2016-05-11 |
Family
ID=49062254
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310175763.5A Expired - Fee Related CN103279739B (en) | 2013-05-10 | 2013-05-10 | A kind of deck detection method based on vehicle characteristics coupling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103279739B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778777B (en) * | 2016-11-30 | 2021-07-06 | 成都通甲优博科技有限责任公司 | Vehicle matching method and system |
| CN106650752B (en) * | 2016-12-09 | 2019-04-30 | 浙江浩腾电子科技股份有限公司 | A kind of body color recognition methods |
| CN106875693A (en) * | 2017-03-29 | 2017-06-20 | 广西信路威科技发展有限公司 | A kind of method and system of vehicle feature recognition |
| CN107680385B (en) * | 2017-10-27 | 2019-12-24 | 泰华智慧产业集团股份有限公司 | Method and system for determining fake-licensed vehicle |
| CN110991255B (en) * | 2019-11-11 | 2023-09-08 | 智慧互通科技股份有限公司 | Method for detecting fake-licensed car based on deep learning algorithm |
| CN112200765A (en) * | 2020-09-04 | 2021-01-08 | 浙江大华技术股份有限公司 | Method and device for determining false-detected key points in vehicle |
| CN117373259B (en) * | 2023-12-07 | 2024-03-01 | 四川北斗云联科技有限公司 | Expressway vehicle fee evasion behavior identification method, device, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002175549A (en) * | 2000-12-07 | 2002-06-21 | Mitsubishi Electric Corp | Method and device for toll adjustment |
| US6449555B1 (en) * | 1999-03-05 | 2002-09-10 | Kabushiki Kaisha Toshiba | Run time information arithmetic operation apparatus |
| CN101540105A (en) * | 2009-04-15 | 2009-09-23 | 四川川大智胜软件股份有限公司 | Fake-licensed car detection method based on number-plate identification and gridding supervision |
| CN102426786A (en) * | 2011-11-15 | 2012-04-25 | 无锡港湾网络科技有限公司 | Intelligent video analyzing system and method for automatically identifying fake plate vehicle |
| CN102881169A (en) * | 2012-09-26 | 2013-01-16 | 青岛海信网络科技股份有限公司 | Fake-licensed car detection method |
-
2013
- 2013-05-10 CN CN201310175763.5A patent/CN103279739B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6449555B1 (en) * | 1999-03-05 | 2002-09-10 | Kabushiki Kaisha Toshiba | Run time information arithmetic operation apparatus |
| JP2002175549A (en) * | 2000-12-07 | 2002-06-21 | Mitsubishi Electric Corp | Method and device for toll adjustment |
| CN101540105A (en) * | 2009-04-15 | 2009-09-23 | 四川川大智胜软件股份有限公司 | Fake-licensed car detection method based on number-plate identification and gridding supervision |
| CN102426786A (en) * | 2011-11-15 | 2012-04-25 | 无锡港湾网络科技有限公司 | Intelligent video analyzing system and method for automatically identifying fake plate vehicle |
| CN102881169A (en) * | 2012-09-26 | 2013-01-16 | 青岛海信网络科技股份有限公司 | Fake-licensed car detection method |
Non-Patent Citations (1)
| Title |
|---|
| 车脸图像的特征提取;姚源;《中国优秀硕士学位论文全文数据库》;20090115;论文第二章、第三章 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103279739A (en) | 2013-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103279739B (en) | A kind of deck detection method based on vehicle characteristics coupling | |
| CN109145915B (en) | Rapid distortion correction method for license plate under complex scene | |
| CN108564814B (en) | Image-based parking lot parking space detection method and device | |
| CN104751142B (en) | A kind of natural scene Method for text detection based on stroke feature | |
| CN106683076B (en) | The method of train wheel tread damage detection based on textural characteristics cluster | |
| CN107590492B (en) | A kind of vehicle-logo location and recognition methods based on convolutional neural networks | |
| CN107705301B (en) | Highway marking damage detection method based on unmanned aerial vehicle aerial highway image | |
| CN102799859A (en) | Method for identifying traffic sign | |
| CN103810474A (en) | Car plate detection method based on multiple feature and low rank matrix representation | |
| CN109886896A (en) | A kind of blue License Plate Segmentation and antidote | |
| CN108830279B (en) | Image feature extraction and matching method | |
| CN104463814B (en) | Image enhancement method based on local texture directionality | |
| CN101872416A (en) | Vehicle license plate recognition method and system of road image | |
| CN103793708A (en) | Multi-scale license plate precise locating method based on affine correction | |
| CN102708364B (en) | Cascade-classifier-based fingerprint image classification method | |
| CN105139011A (en) | Method and apparatus for identifying vehicle based on identification marker image | |
| CN104240231A (en) | Multi-source image registration based on local structure binary pattern | |
| CN106022303B (en) | Fresh water algae rough sort and method of counting based on no lens holographic imaging | |
| CN108229247A (en) | A kind of mobile vehicle detection method | |
| CN102930242B (en) | Bus type identifying method | |
| CN116958714B (en) | Automatic identification method for wafer back damage defect | |
| CN106408029B (en) | A kind of image texture classification method based on structure histogram of difference | |
| CN109165659B (en) | Vehicle color identification method based on superpixel segmentation | |
| CN103700119A (en) | Local texture description method based on local grouping comparison mode column diagram | |
| Zhang et al. | A robust license plate detection algorithm based on multi-features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent of invention or patent application | ||
| CB02 | Change of applicant information |
Address after: Hangzhou City, Zhejiang province Yuhang District 311121 West Street Wuchang No. 998 building 7 East Applicant after: Zhejiang iCare Vision Technology Co., Ltd. Address before: 310013, Zhejiang, Xihu District, Hangzhou, Tian Shan Road, No. 398, Kun building, four floor, South Block Applicant before: Zhejiang iCare Vision Technology Co., Ltd. |
|
| C53 | Correction of patent of invention or patent application | ||
| CB02 | Change of applicant information |
Address after: Hangzhou City, Zhejiang province Yuhang District 311121 West Street Wuchang No. 998 building 7 East Applicant after: ZHEJIANG ICARE VISION TECHNOLOGY CO., LTD. Address before: Hangzhou City, Zhejiang province Yuhang District 311121 West Street Wuchang No. 998 building 7 East Applicant before: Zhejiang iCare Vision Technology Co., Ltd. |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: HANGZHOU ICARE VISION TECHNOLOGY CO., LTD. TO: ZHEJIANG ICARE VISION TECHNOLOGY CO., LTD. |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Fake license plate detection method based on vehicle characteristic matching Effective date of registration: 20190821 Granted publication date: 20160511 Pledgee: Hangzhou Yuhang Small and Medium-sized Enterprise Transfer Service Co., Ltd. Pledgor: ZHEJIANG ICARE VISION TECHNOLOGY CO., LTD. Registration number: Y2019330000020 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160511 Termination date: 20210510 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |