CN101901479A - Method for optimizing license plate image with character color dropout - Google Patents
Method for optimizing license plate image with character color dropout Download PDFInfo
- Publication number
- CN101901479A CN101901479A CN 201010241650 CN201010241650A CN101901479A CN 101901479 A CN101901479 A CN 101901479A CN 201010241650 CN201010241650 CN 201010241650 CN 201010241650 A CN201010241650 A CN 201010241650A CN 101901479 A CN101901479 A CN 101901479A
- Authority
- CN
- China
- Prior art keywords
- license plate
- plate image
- value
- image
- pixel point
- 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
Landscapes
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a method for optimizing a license plate image with character color dropout, in particular to a method which is used for improving the stroke definition of a character color in the license plate image in an automatic blurred image recognition process of a motor vehicle, lays a basis for increasing license plate character information recognition accuracy and is suitable for performing intellectualized and informationalized processing and analysis on an image recorded by road traffic technology monitoring equipment. According to a license plate image optimization algorithm for the character color dropout, grey image reconstruction is performed on an original 24-bit true color license plate image, and then the character stroke definition is improved in horizontal and vertical directions respectively, so that a new license plate image definition enhancement pattern is formed, and the stroke profiles of license plate characters in the pattern are improved.
Description
Technical field
The present invention relates to a kind of license plate image disposal route, the method for optimizing license plate image that specifically a kind of character color comes off.
Background technology
Along with the sharp sharp increase length of vehicle possess amount and the quick propelling of road construction mileage, road traffic technology monitoring technique has obtained tremendous development, forming the identification of collection characteristics of image, traffic offence evidence obtaining, video frequency graphic monitoring, bayonet socket, to raid technology such as deploy to ensure effective monitoring and control of illegal activities be the automobile traffic monitoring technique system in transit of one, for motor vehicle conduct monitoring at all levels, scientific management, information-based service are laid a good foundation, and automobile traffic monitoring technique system core in transit is an automotive number plate image automatic identification technology.Current, China's automotive number plate is produced and is provided by various places public safety traffic management department tissue.Because the difference of various places car plate manufacturing process, and characters on license plate suffers environmental pollutions such as muddy water, light, friction easily in the actual travel process, comprise artificial broken ring etc., automotive number plate unavoidably produces phenomenons such as the characters on license plate color comes off in use for some time, and this discerns for the automotive number plate image automatically and has brought many difficulties.Therefore, automotive number plate is the I.D. that motor vehicle is set out on a journey and travelled, the license plate image identification level automatically that raising characters on license plate color comes off is the innate need of the economy that in-depth road traffic technology watch-dog is used, the inevitable requirement of the comprehensive management and control of motor vehicle characteristic information especially.
Summary of the invention
The objective of the invention is in order to strengthen the characters on license plate stroke profile in the license plate image, promote the image definition that the characters on license plate color comes off, improve the recognition accuracy of characters on license plate information, the method for optimizing license plate image that provides a kind of character color to come off, be used for the automatic identifying of automotive number plate blurred picture character color in the license plate image is carried out the lifting of stroke sharpness, lay the foundation for improving characters on license plate information Recognition accuracy rate, be applicable to that the image to road traffic technology watch-dog record carries out intellectuality, the information-based processing and analysis.
According to technical scheme provided by the invention, the method for optimizing license plate image that described character color comes off is as follows: at first 24 pending each pixels of the true coloured picture of car plate are extracted new brightness value Y by R, G, three colours of B according to formula Y=0.299R+0.587G+0.114B, and with the Y value as new license plate image pixel data, thereby form the license plate image grey restructuring graph of 8 colors; Then, described license plate image grey restructuring graph is carried out the stroke sharpness promotes, form the license plate image sharpness and strengthen figure, be specially:
(1) set up the image pixel point array V[W * H of described license plate image grey restructuring graph], wherein W is the picture traverse value, H is the picture altitude value, pixel number is according to from V[0] beginning according to from left to right, sequential storage from falling down;
(2) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to horizontal direction], V[I * W+J+1], wherein the value of I is 0 to H-1, the value of J is 0 to W-1;
(3) judge the reckling MIN1 of two obtained pixel point values, MIN1 is composed to V[J];
(4) value of J increases by 1, judges whether J＜W-1 sets up, and being then changes step 2, otherwise enters step 5;
(5) value of I increases by 1, judges whether I＜H sets up, and being then changes step 2, otherwise enters step 6;
(6) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to vertical direction], V[(I+1) * W+J];
(7) judge the reckling MIN2 of two obtained pixel point values, MIN2 is composed to V[I * W+J];
(8) value of J increases by 1, judges whether J＜W sets up, and being then changes step 6, otherwise enters step 9;
(9) value of I increases by 1, judges whether I＜H-1 sets up, and being then changes step 6, otherwise enters step 10;
The image pixel point array V[W * H that (10) will newly form] replace the pixel point value of relevant position in the former license plate image grey restructuring graph, form the license plate image sharpness and strengthen figure.
Advantage of the present invention is: the method for optimizing license plate image that character color of the present invention comes off, can not cause the automatic recognition accuracy of license plate image character information to reduce because of the characters on license plate color comes off, strengthened the profile of characters on license plate stroke, promoted the sharpness of license plate image, for correct identification characters on license plate information lays the foundation.
Embodiment
Because patent application document is to the requirement of accompanying drawing, this instructions does not provide license plate image in the processing procedure as accompanying drawing, can provide these images to represent the effect that each step handles according to auditor's requirement.
Character on some license plate image can be subjected to extraneous or artificial broken ring in varying degrees, makes characters on license plate stroke color that serious the change be taken place, and this has brought difficulty for characters on license plate information appliance computer automatic program identification.But the method for optimizing license plate image that comes off according to the character color of patent of the present invention, original 24 very color license plate images are carried out grey chromatic graph reconstruct, carrying out the stroke sharpness from level, vertical direction respectively then promotes, form new license plate image sharpness and strengthen figure, characters on license plate stroke profile is enhanced among the figure.
Pending image of the present invention is 24 the true coloured picture of car plate.Stipulate (GA36-2009) that according to " People's Republic of China's automotive number plate " full-sized car number plate color is yellow end surplus, and the characters on license plate in the pending image is subjected to brokenly ring to a certain extent, makes characters on license plate stroke color that serious change take place.
According to the principle of three primary colours, the single pixel value in 24 true coloured pictures is got red, green, blue three chromatic value R, G, B different proportion according to red (R), green (G), blue (B) three kinds of colours storages, forms new grey reconstructed image.Specifically extract new brightness value Y by the brightness-formula Y=0.299R+0.587G+0.114B of color system, and according to the pixel order with the Y value as new license plate image data, thereby form the license plate image grey restructuring graph of 8 colors.Yet in the described license plate image grey restructuring graph, the gray-scale value of some ruined stroke still has bigger difference in the license plate image.If carry out car plate identification, then can't accurately discern three character informations on the car plate left side according to described license plate image grey restructuring graph.
According to pixel point value distribution characteristics in the described license plate image grey restructuring graph, according to level, vertical direction license plate image is carried out the stroke sharpness respectively and promote, can form license plate image sharpness enhancing figure.Characters on license plate stroke profile is enhanced among the figure, for fast, accurately, automatically identify the license plate image character information that character color comes off and lay the foundation.Concrete steps are:
(1) set up the image pixel point array V[W * H of described license plate image grey restructuring graph], wherein W is the picture traverse value, H is the picture altitude value, pixel number is according to from V[0] beginning according to from left to right, sequential storage from falling down;
(2) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to horizontal direction], V[I * W+J+1], wherein the value of I is 0 to H-1, the value of J is 0 to W-1;
(3) judge the reckling MIN1 of two obtained pixel point values, MIN1 is composed to V[J];
(4) advance a pixel (even the value of J increases by 1) to horizontal direction, whether the determined level direction finishes, and judging promptly that J＜W-1 sets up then changes step 2, and being false enters step 5;
(5) advance delegation (value that is I increases by 1) to vertical direction, judge whether vertical direction finishes, judge that promptly I＜H establishment then changes step 2, being false enters step 6;
(6) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to vertical direction], V[(I+1) * W+J];
(7) judge the reckling MIN2 of two obtained pixel point values, MIN2 is composed to V[I * W+J];
(8) advance a pixel (value that is J increases by 1) to horizontal direction, whether the determined level direction finishes, and judging promptly that J＜W sets up then changes step 6, and being false enters step 9;
(9) advance delegation (value that is I increases by 1) to vertical direction, judge whether vertical direction finishes, judge that promptly I＜H-1 establishment then changes step 6, being false enters step 10;
(10) with formed image pixel point array V[W * H] replace the pixel point value of former license plate image grey restructuring graph relevant position, form the license plate image sharpness and strengthen figure.
The license plate image sharpness enhancing figure that obtains thus can be discerned by more existing image recognition softwares, such as according to following principle:
Whole pixel point values and 255 among the described license plate image sharpness enhancing figure are obtained new pixel point value as difference, and the color of finishing the characters on license plate stroke transfers white to from black.
Opsition dependent is cut apart characters on license plate then, and a character is one; To cut apart good character picture again and be converted to black and white two chromatic graphs respectively, can accurately discern the alphabet information in the car plate.
The invention solves the low problem of the automatic recognition accuracy of license plate image character information that comes off and cause because of the characters on license plate color, strengthened the profile of characters on license plate stroke, promoted the sharpness of characters on license plate image, for correct identification license plate image information is laid a good foundation.
Claims (1)
1. method for optimizing license plate image that character color comes off, it is characterized in that: at first 24 pending each pixels of the true coloured picture of car plate are extracted new brightness value Y by R, G, three colours of B according to formula Y=0.299R+0.587G+0.114B, and with the Y value as new license plate image pixel data, thereby form the license plate image grey restructuring graph of 8 colors; Then, described license plate image grey restructuring graph is carried out the stroke sharpness promotes, form the license plate image sharpness and strengthen figure, be specially:
(1) set up the image pixel point array V[W * H of described license plate image grey restructuring graph], wherein W is the picture traverse value, H is the picture altitude value, pixel number is according to from V[0] beginning according to from left to right, sequential storage from falling down;
(2) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to horizontal direction], V[I * W+J+1], wherein the value of I is 0 to H-1, the value of J is 0 to W-1;
(3) judge the reckling MIN1 of two obtained pixel point values, MIN1 is composed to V[J];
(4) value of J increases by 1, judges whether J＜W-1 sets up, and being then changes step 2, otherwise enters step 5;
(5) value of I increases by 1, judges whether I＜H sets up, and being then changes step 2, otherwise enters step 6;
(6) get two pixel point values, i.e. V[I * W+J in the described license plate image grey restructuring graph continuously according to vertical direction], V[(I+1) * W+J];
(7) judge the reckling MIN2 of two obtained pixel point values, MIN2 is composed to V[I * W+J];
(8) value of J increases by 1, judges whether J＜W sets up, and being then changes step 6, otherwise enters step 9;
(9) value of I increases by 1, judges whether I＜H-1 sets up, and being then changes step 6, otherwise enters step 10;
The image pixel point array V[W * H that (10) will newly form] replace the pixel point value of relevant position in the former license plate image grey restructuring graph, form the license plate image sharpness and strengthen figure.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010241650 CN101901479A (en) | 2010-07-26 | 2010-07-26 | Method for optimizing license plate image with character color dropout |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010241650 CN101901479A (en) | 2010-07-26 | 2010-07-26 | Method for optimizing license plate image with character color dropout |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101901479A true CN101901479A (en) | 2010-12-01 |
Family
ID=43226990
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010241650 Pending CN101901479A (en) | 2010-07-26 | 2010-07-26 | Method for optimizing license plate image with character color dropout |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101901479A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503707A (en) * | 2016-10-21 | 2017-03-15 | 浙江宇视科技有限公司 | Licence plate recognition method and device under the conditions of a kind of infrared light filling |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101064011A (en) * | 2006-04-26 | 2007-10-31 | 电子科技大学 | Vehicle registration extract method in complicated background based on wavelet transforming |
| CN101567041A (en) * | 2009-05-25 | 2009-10-28 | 公安部交通管理科学研究所 | Method for recognizing characters of number plate images of motor vehicles based on trimetric projection |
-
2010
- 2010-07-26 CN CN 201010241650 patent/CN101901479A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101064011A (en) * | 2006-04-26 | 2007-10-31 | 电子科技大学 | Vehicle registration extract method in complicated background based on wavelet transforming |
| CN101567041A (en) * | 2009-05-25 | 2009-10-28 | 公安部交通管理科学研究所 | Method for recognizing characters of number plate images of motor vehicles based on trimetric projection |
Non-Patent Citations (6)
| Title |
|---|
| 《15th European Signal Processing Conference》 20070907 Vahid Abolghasemi et al Local Enhancement of Car Image for License Plate Detection 全文 1 , 2 * |
| 《JOURNAL OF COMPUTERS》 20090831 Kaushik Deb et al Vehicle License Plate Detection Method Based on Sliding Concentric Windows and Histogram 全文 1 第4卷, 第8期 2 * |
| 《Pattern Recognition》 20051231 Shigueo Nomura et al A Novel Adaptive Morphological Approach for Degraded Character Image Segmentation 全文 1 第38卷, 2 * |
| 《计算机工程与科学》 20081231 张宝强 等 字符分割前车牌图像的预处理 全文 1 第30卷, 第4期 2 * |
| 《计算机应用研究》 20041231 李文举 等 一种新的面向字符分割的车牌图像预处理方法 全文 1 , 第7期 2 * |
| 《计算机辅助设计与图形学学报》 20040531 李文举 等 质量退化的车牌字符分割方法 全文 1 第16卷, 第5期 2 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106503707A (en) * | 2016-10-21 | 2017-03-15 | 浙江宇视科技有限公司 | Licence plate recognition method and device under the conditions of a kind of infrared light filling |
| CN106503707B (en) * | 2016-10-21 | 2019-08-02 | 浙江宇视科技有限公司 | Licence plate recognition method and device under a kind of infrared lighting condition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106373426B (en) | Parking stall based on computer vision and violation road occupation for parking monitoring method | |
| CN102375982B (en) | Multi-character characteristic fused license plate positioning method | |
| CN104766086B (en) | The monitoring and managing method and system of a kind of way mark | |
| US20170262713A1 (en) | Method and Device of Lane Departure Warning and Automobile Driving Assistance System | |
| CN104392212A (en) | Method for detecting road information and identifying forward vehicles based on vision | |
| DE112013001858T5 (en) | Multiple-hint object recognition and analysis | |
| CN107240079A (en) | A kind of road surface crack detection method based on image procossing | |
| CN101567041A (en) | Method for recognizing characters of number plate images of motor vehicles based on trimetric projection | |
| CN102737221A (en) | Method and apparatus for vehicle color identification | |
| Yusof et al. | Crack detection and classification in asphalt pavement images using deep convolution neural network | |
| CN106023623A (en) | Recognition and early warning method of vehicle-borne traffic signal and symbol based on machine vision | |
| CN106951898B (en) | Vehicle candidate area recommendation method and system and electronic equipment | |
| CN104700072A (en) | Lane line historical frame recognition method | |
| CN105335708A (en) | System for identifying road signs and guideboards during process of vehicle moving | |
| CN102880859A (en) | Method for recognizing number plate | |
| CN104778833A (en) | Traffic light recognition method | |
| CN107644538A (en) | The recognition methods of traffic lights and device | |
| CN102306276A (en) | Method for identifying color of vehicle body in video vehicle image based on block clustering | |
| CN103186905A (en) | Color detector for vehicle | |
| CN111144301A (en) | Road pavement defect quick early warning device based on degree of depth learning | |
| CN107145824B (en) | Lane line segmentation method and system based on significance analysis and vehicle-mounted terminal | |
| CN102364560B (en) | Traffic sign convenient for electronic identification and method for identifying traffic sign | |
| CN101901479A (en) | Method for optimizing license plate image with character color dropout | |
| CN106599846B (en) | A kind of recognition methods for the traffic mark board being easy to Computer Vision Recognition | |
| KR20220138894A (en) | Prediction and recognition method of road marking information and road maintenance method |
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 |
Open date: 20101201 |