CN103699876B - Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images - Google Patents
Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images Download PDFInfo
- Publication number
- CN103699876B CN103699876B CN201310613538.5A CN201310613538A CN103699876B CN 103699876 B CN103699876 B CN 103699876B CN 201310613538 A CN201310613538 A CN 201310613538A CN 103699876 B CN103699876 B CN 103699876B
- Authority
- CN
- China
- Prior art keywords
- character
- image
- license number
- feature
- segmentation
- 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.)
- Active
Links
Landscapes
- Image Analysis (AREA)
- Character Input (AREA)
Abstract
The invention provides a method for identifying a vehicle number based on linear array CCD (Charge Coupled Device) images. The method comprises the following steps: acquiring a real-time image of the vehicle number; acquiring the approximate location of the vehicle number from screenshot of the real-time image; determining the accurate location of the vehicle number; dividing the vehicle number in the image into single characters; extracting the features of the divided single characters; comparing the extracted features with the features in a character library to obtain a result. The method provided by the invention can effectively improve vehicle number acquisition accuracy through image capture, and can greatly improve the identification speed by correspondingly reducing the area of image for identification in the technical scheme.
Description
Technical field
The invention is related to image processing field, particularly a kind of method based on linear array CCD image identification license number and
Device
Background technology
In organization of rail transport, it is the work wasted time and energy to read, check locomotive identification, artificial to read what is recorded
Way usually occurs artificial unnecessary mistake again.With bullet train continuing to develop in China, some website vehicles
Turnover is frequent, the various actions of calculating vehicle, and is arranged and organized as a mixed and disorderly complicated job according to it.Believe as railway
An important ring in breathization --- car number automatic identification be just particularly important Digital Image Processing then because its processing accuracy is high,
Process content is abundant, can carry out the nonlinear operation and good flexibility of complexity.Have become the head of car number automatic identification
Choosing.But how to be still a problem for computer disposal by the corresponding character that is identified as of image font size accurate stable.Linear array
CCD simple structures, cost is relatively low.A line TV signal can simultaneously be stored because the number of its single photosensitive unit can do
Obtain a lot, on the premise of equal certainty of measurement, it is larger that its measurement range can be done, and due to line array CCD real-time Transmission
Light-to-current inversion signal and self-scanning speed is fast, frequency response is high, can realize dynamic measurement, and can be worked under low-light (level), institute
Many fields such as product size measurement and classification, noncontact dimension measurement, bar code are widely used in line array CCD.
The content of the invention
The invention problem to be solved is the respective digital that license number how is fast and accurately obtained by picture catching
Information is used for system.
In order to solve the above technical problems, the technical scheme that the invention is used is:It is a kind of to be recognized based on linear array CCD image
The method of license number, described method comprises the following steps:
Obtain the realtime graphic of license number;
The Position Approximate of license number is obtained from the sectional drawing of realtime graphic;
It is determined that accurate license number position;
License number in picture is divided into single character;
Single character after to segmentation carries out feature extraction;
The feature of extraction is compared with the feature in character repertoire, result is obtained.
Further, the described Position Approximate that license number is obtained from sectional drawing comprises the following steps:
Image in the realtime graphic of license number positioned at license number position height is intercepted;
Gaussian filtering is carried out to truncated picture;
Sobel operator operations are carried out to the image after gaussian filtering process;
The central point in brightness highest region in image after lookup treatment;
Region Breadth Maximum and maximum height setting rectangle according to license number;
The rectangular centre point for setting is overlapped with brightness highest zone central point, the big position approximate with rectangular area as license number
Put.
Further, the accurate license number position of described determination comprises the following steps:
To obtaining the sectional drawing after Position Approximate operated based on Canny operators, obtain edge-detected image;
Character pitch distance and two pre-set threshold values of character height are set(DX, DY);
Carry out Y-direction histogram analysis to edge-detected image, lookup meet the abscissa of character pitch predetermined threshold value-
X’;
Carry out X-direction histogram analysis to edge-detected image, lookup meet the ordinate of character height predetermined threshold value-
Y’;
It is determined that accurate license number position.
Further, it is described license number in picture is divided into single character to comprise the following steps:
Image after spotting is carried out into binary conversion treatment;
Suitable inter-character space predetermined threshold value is set;
The histogram analysis of Y-direction are carried out to the image after binary conversion treatment, lookup meets inter-character space predetermined threshold value
Coordinate;
Character is divided according to inter-character space coordinate is met.
Further, it is described to segmentation after single character carry out feature extraction and comprise the following steps:
Corner Detection is carried out to the character after segmentation;
Extract the characteristic vector of the character after each segmentation.
Further, the described feature by extraction is compared with the feature in character repertoire, and obtaining result includes following step
Suddenly:
Character feature in the character feature of extraction and character repertoire is entered into row distance to calculate;
The minimum character of selected characteristic vector distance is matching result.
Further, described characteristic character storehouse is set up and is comprised the following steps:
The image pattern that license number will be included is divided into single character;
Single character picture is carried out into gaussian filtering;
The image that gaussian filtering will be carried out carries out binary conversion treatment;
Interpolating transform is carried out to the image after above-mentioned treatment, makes the width of image high consistent with Prototype drawing image width high standard
The characteristic vector of character is extracted by corner detection operator;
Repeat to carry out image pattern of the multiple comprising license number aforesaid operations, generation character feature storehouse.
Further, the characteristic vector of described character includes profile corner location, angle point distance, angle point angle.
Further, described gaussian filtering is 2-d gaussian filterses
Further, described Sobel operators are
Further, the described Corner Detection that carried out to the character after segmentation uses harris algorithms, Harris operators:
IxIt is image in X-direction difference result, IyIt is image difference result in the Y direction
Angle point amount is calculated:
According to another aspect of the present invention, a kind of device that license number is recognized based on linear array CCD image is additionally provided, including:
License number image acquiring device, for obtaining the realtime graphic comprising license number image;
License number about location drawing picture acquisition device, for from realtime graphic obtain license number about position and sectional drawing obtain
Take;
License number accurate location image acquiring device, for obtaining license number accurate location from license number about position sectional drawing and cutting
Figure is obtained;
Character segmentation device, for being single character combination by Character segmentation in the image comprising license number accurate location;
Character feature extraction element, for extracting the characteristic vector in single character picture;
Character feature comparison match device, for character feature in the character feature of extraction vector and character repertoire to be compared
Relatively match;And
Character repertoire, the character feature of comparison match is can be used for for storing.
Further, license number about location drawing picture acquisition device includes in the device of described image recognition license number:
Height image capture device, for being intercepted positioned at license number position height image in the realtime graphic by license number;
Gaussian filtering device, for carrying out gaussian filtering operation to interception license number position height image;
Sobel operator devices, for carrying out Sobel operator operations to the image after gaussian filtering process;
Device is searched in brightness, the central point in brightness highest region in the image after being processed for lookup;
Rectangle setting device, for setting rectangle according to the region Breadth Maximum and maximum height of license number;
Coincidence arrangement, for the rectangular centre point for setting to be overlapped with brightness highest zone central point, be with rectangular area
The Position Approximate of license number.
Further, license number accurate location image acquiring device includes in the device of described image recognition license number:
Edge detecting device, for carrying out the edge inspection based on Canny operators to obtaining the sectional drawing after Position Approximate treatment
Survey;
Threshold values sets device, for setting character pitch height and two threshold values of width;
Device is searched in Y-direction analysis, and for carrying out Y-direction histogram analysis to edge-detected image, lookup meets character
The coordinate of interval threshold;
Device is searched in X-direction analysis, and for carrying out X-direction histogram analysis to edge-detected image, lookup meets character
The coordinate of height threshold.
Further, described Character segmentation device includes:
Binary conversion treatment device, for the image after spotting to be carried out into binary conversion treatment;
Inter-character space threshode devices, for setting suitable inter-character space threshold value;
Y-direction is analyzed and coordinate searches device, and the histogram for carrying out Y-direction to the image after binary conversion treatment is looked into
Look for the coordinate for meeting inter-character space;
Character divide device, for according to meet inter-character space coordinate divide character.
Further, described character feature extraction element includes:
Corner Detection device, for carrying out Corner Detection to the character after segmentation;
Feature-vector extraction means, the characteristic vector for extracting the character after each segmentation.
Further, described character feature comparison match device includes:
Apart from computing device, calculated for character feature in the character feature of extraction and character repertoire to be entered into row distance;
As a result coalignment, is matching result for the minimum character of selected characteristic vector distance.
Further, the device of described image recognition license number also sets up device including character repertoire, and described character repertoire is set up
Device includes:
Character segmentation device, the image pattern for will include license number is divided into single character;
Gaussian filtering device, for single character picture to be carried out into gaussian filtering;
Binaryzation device, the image for will carry out gaussian filtering carries out binary conversion treatment;
Normalized device, for carrying out Interpolating transform to the image after above-mentioned treatment, makes the width of image high and template image
High standard wide is consistent
Feature-vector extraction means, the characteristic vector for extracting character by corner detection operator;
Character feature storehouse, for storing multiple character feature vectors.
The invention has the advantages and positive effects that:Using the technical program, can effectively improve by figure
The accuracy rate of license number is obtained as catching, due to also accordingly reducing the image area being identified in the technical program, can be with
Increase substantially recognition speed.
Brief description of the drawings
Fig. 1 is the image pattern schematic diagram provided in one embodiment in the present invention
Fig. 2 is the motor-car license number image schematic diagram gathered in one embodiment in the present invention
Fig. 3 is the schematic flow sheet for setting up character feature storehouse in the present invention in one embodiment
Fig. 4 be in the present invention in one embodiment to the schematic flow sheet of image recognition character
Specific embodiment
In order to the more deep understanding of the invention, a specific embodiment be set forth below, and with reference to accompanying drawing, to this hair
Bright creation is described in further detail.
In the present embodiment, it is acquired by image pattern first, and identifying system is trained, is used to set up
Corresponding character feature storehouse.
As shown in figure 1, Fig. 1 is an example of the image pattern of offer in this example, in fact, being to identification to reach
System is trained, and is used to set up the purpose in corresponding character feature storehouse, employs many item of image samples for using.The image of collection
Sample can come from photo or other manner.
Sample license number is divided into single character first, can be split license number by manual type or other manner
Be single character, should as far as possible according to the single character font complete display after segmentation, and each character picture size phase
Together, facilitating the later stage accurately to provide corresponding character feature.
Single character picture for having split carries out gaussian filtering operation, general using 2-d gaussian filterses operation, can
Using following formula:
2-d gaussian filterses operation is carried out to image, the main purpose of the step is to eliminate Gaussian noise, for right
Image pattern noise abatement, facilitates follow-up operation.
In general, image pattern is probably gray level image, it is also possible to be coloured image, for colour, it should to the greatest extent
Gray scale may be translated into, facilitates the later stage to process, binaryzation, binaryzation can be carried out to it by the gray value of pixel
Formula it is as follows:
P (x, y) is pixel in image,
For the picture as sample, it is very important to meet correct size, if slightly deformation will give
The treatment in later stage brings larger error, for the picture for not meeting size, it is possible to use interpolation method is expanded picture,
Make up to standard size.Interpolation method is the pixel according to original image, is calculated according to certain computing mode, is produced
New pixel, and the neighbouring gap of original pixel is inserted it into, so as to realize increased pixel total amount and increase picture
The purpose of plain density.
The characteristic vector V of character is extracted by corner detection operatori, characteristic vector includes:Profile corner location, angle point away from
From angle point angle Vi=(Xi,Yi,Li,θi), and store all Corner Feature vectors of the character;
Harris operators:
IxIt is image in X-direction difference result, IyIt is image difference result in the Y direction
Angle point amount is calculated:
By the characteristic vector V of each character outlineiClassification based training is carried out, final character feature storehouse is generated.Character feature
Storehouse can constantly increase new image and be practised, and the scope for being allowed to character feature is maintained at reasonable, accurate level.
Fig. 2 is the motor-car license number image gathered in this example, and red block is to look for out license number exact image position, for
For the image for being caught, if for whole image carries out lookup contrast, technical accuracy and operand all can be into
It is big problem.
For the image for capturing, its corresponding accurate location with font size image will be can determine first, because dynamic
The height of car is fixed, and camera head position and height are also fixed, and in general, font size is in the picture for collecting
The position of appearance is also relatively fixed, can by default height by this portion intercepts of picture out.
Gaussian filtering, Sobel operator operations are carried out to the above-mentioned image for intercepting out:
2-d gaussian filterses formula:
Sobel operators:
Image after above-mentioned treatment, license number zone luminance value is high, sets the rectangle of W*H, average bright in detection image
Degree highest region, determines license number substantially picture position.(W is license number region Breadth Maximum, and H is license number region maximum height)It is logical
Cross this step operation, it may be determined that font size Position Approximate in the picture, this Position Approximate should be greater than actual size, it is still desirable under
Single stepping can just access accurate position.
License number image is carried out based on Canny rim detections, the histogram analysis by Y-direction marginal information, X-direction
The histogram analysis of marginal information carry out being accurately positioned for license number image.
The rim detection based on Canny operators is carried out to image first;
Character pitch distance and two pre-set threshold values of character height are set(DX, DY);
This preset value is our the actual conditions settings in history image.(what is used in this patent is default
Value:DX=30, DY=85) Y-direction histogram analysis are carried out to edge-detected image, lookup meets license number region and chinese character area
Domain is spaced the coordinate-X ' of predetermined threshold value DX;
Carry out X-direction histogram analysis to edge-detected image, lookup meet the coordinate of license number character height predetermined threshold value-
Y’;
It is determined that accurate license number position-[X ', Y '].
, it is necessary to each font size is split from picture after aforesaid operations are completed, segmentation can be using such as lower section
Method:
Symbol image binaryzation;
P (x, y) is pixel in image,
The predetermined threshold value D of character pitch is set, and preset value can be according to the interval setting of character in real image.At this
In the preset value that uses:D=10
Binary image carries out the histogram analysis of Y-direction, and lookup meets the coordinate of inter-character space, character is divided
Cut.The character after segmentation is finally carried out into feature extraction, its extracting mode is identical with the above-mentioned method for setting up character repertoire, herein not
Repeat, character feature in the character feature of extraction and character repertoire is entered into row distance calculates, choose minimal characteristic vector distance and be
Final recognition result.
Min(Di)={||V-Vi||}
The embodiment to the invention has been described in detail above, but the content is only the preferable of the invention
Embodiment, it is impossible to be considered as limiting practical range of the invention.All impartial changes made according to the invention scope with
Improve etc., all should still belong within this patent covering scope.
Claims (10)
1. it is a kind of based on linear array CCD image recognize license number method, it is characterised in that described method comprises the following steps:
Obtain the realtime graphic of license number;
The Position Approximate of license number is obtained from the sectional drawing of realtime graphic by default height;
It is determined that accurate license number position;
License number in picture is divided into single character;
Single character after to segmentation carries out feature extraction;
The feature of extraction is compared with the feature in character repertoire, result is obtained;The wherein described sectional drawing from realtime graphic
The middle Position Approximate for obtaining license number comprises the following steps:
Image in the realtime graphic of license number positioned at license number position height is intercepted;
Gaussian filtering is carried out to truncated picture;
Sobel operator operations are carried out to the image after gaussian filtering process;
The central point in brightness highest region in image after lookup treatment;
Region Breadth Maximum and maximum height setting rectangle according to license number;
The rectangular centre point for setting is overlapped with brightness highest zone central point, the Position Approximate with rectangular area as license number;
The accurate license number position of described determination comprises the following steps:
Operated based on Canny operators to obtaining the sectional drawing after Position Approximate treatment, obtain edge-detected image;
Character pitch distance and two threshold values of character height are set;
Y-direction histogram analysis are carried out to edge-detected image, lookup meets the abscissa X of character pitch predetermined threshold value;
X-direction histogram analysis are carried out to edge-detected image, lookup meets the ordinate Y of character height predetermined threshold value;
It is determined that accurate license number position;
It is described license number in picture is divided into single character to comprise the following steps:
Image after spotting is carried out into binary conversion treatment;
Suitable inter-character space threshold value is set;
The histogram analysis of Y-direction are carried out to the image after binary conversion treatment, lookup meets the coordinate of inter-character space predetermined threshold value;
Character is divided according to inter-character space coordinate is met;
It is described to segmentation after single character carry out feature extraction and comprise the following steps:
Corner Detection is carried out to the character after segmentation;
Extract the characteristic vector of the character after each segmentation;
The described feature by extraction is compared with the feature in character repertoire, is obtained result and is comprised the following steps:
Character feature in the character feature of extraction and character repertoire is entered into row distance to calculate;
The minimum character of selected characteristic vector distance is matching result.
2. it is according to claim 1 based on linear array CCD image recognize license number method, it is characterised in that described character repertoire
Foundation comprises the following steps:
The image pattern that license number will be included is divided into single character;
Single character picture is carried out into gaussian filtering;
The image that gaussian filtering will be carried out carries out binary conversion treatment;
Interpolating transform is carried out to the image after above-mentioned treatment, makes the width of image high consistent with Prototype drawing image width high standard;
The characteristic vector of character is extracted by corner detection operator;
Repeat to carry out image pattern of the multiple comprising license number aforesaid operations, generation character feature storehouse.
3. it is according to claim 1 based on linear array CCD image recognize license number method, it is characterised in that:Described character
Characteristic vector include profile corner location, angle point distance, angle point angle.
4. it is according to claim 2 based on linear array CCD image recognize license number method, it is characterised in that:Described Gauss
It is filtered into 2-d gaussian filterses
5. it is according to claim 2 based on linear array CCD image recognize license number method, it is characterised in that:Described Sobel
Operator is
。
6. the method that license number is recognized based on linear array CCD image according to claim 5 or claim 1, its feature exists
In it is described to segmentation after character carry out Corner Detection using harris algorithms, harris operators:
IxIt is image in X-direction difference result, IyFor image, difference result angle point amount is calculated in the Y direction:
7. it is a kind of based on linear array CCD image recognize license number device, it is characterised in that including:
License number image acquiring device, for obtaining the realtime graphic comprising license number image;
License number about location drawing picture acquisition device, the about position for obtaining license number from realtime graphic by default height
And sectional drawing is obtained;
License number accurate location image acquiring device, for obtaining license number accurate location from license number about position sectional drawing and sectional drawing is obtained
Take;
Character segmentation device, for being single character combination by Character segmentation in the image comprising license number accurate location;
Character feature extraction element, for extracting the characteristic vector in single character picture;
A character feature comparison match device, for character feature in the character feature of extraction vector and character repertoire to be compared
Match somebody with somebody;And
Character repertoire, the character feature of comparison match is can be used for for storing;Wherein
License number about location drawing picture acquisition device includes in the device of described image recognition license number:
Height image capture device, for being intercepted positioned at license number position height image in the realtime graphic by license number;
Gaussian filtering device, for carrying out gaussian filtering operation to interception license number position height image;
Sobel operator devices, for carrying out Sobel operator operations to the image after gaussian filtering process;
Device is searched in brightness, the central point in brightness highest region in the image after being processed for lookup;
Rectangle setting device, for setting rectangle according to the region Breadth Maximum and maximum height of license number;
Coincidence arrangement, for the rectangular centre point for setting to be overlapped with brightness highest zone central point, with rectangular area as license number
Position Approximate;
License number accurate location image acquiring device includes in the device of described image recognition license number:
Edge detecting device, for carrying out rim detection to obtaining the sectional drawing after Position Approximate treatment;
Threshold value sets device, for setting character pitch height and two threshold values of width;
Device is searched in Y-direction analysis, and for carrying out Y-direction histogram analysis to edge-detected image, lookup meets character pitch
The coordinate of threshold value;
Device is searched in X-direction analysis, and for carrying out X-direction histogram analysis to edge-detected image, lookup meets character height
The coordinate of threshold value;
Described Character segmentation device includes:
Binary conversion treatment device, for the image after spotting to be carried out into binary conversion treatment;Inter-character space threshode devices,
For setting inter-character space threshold value;
Y-direction is analyzed and coordinate searches device, and the histogram for carrying out Y-direction to the image after binary conversion treatment searches symbol
Close the coordinate of inter-character space;
Character divide device, for according to meet inter-character space coordinate divide character.
8. the device that license number is recognized based on linear array CCD image according to claim 7, it is characterised in that described character is special
Levying extraction element includes:
Corner Detection device, for carrying out Corner Detection to the character after segmentation;
Feature-vector extraction means, the characteristic vector for extracting the character after each segmentation.
9. the device that license number is recognized based on linear array CCD image according to claim 7, it is characterised in that described character is special
Levying comparison match device includes:Apart from computing device, for character feature in the character feature of extraction and character repertoire to be entered into line-spacing
From calculating;
As a result coalignment, is matching result for the minimum character of selected characteristic vector distance.
10. it is according to claim 7 based on linear array CCD image recognize license number device, it is characterised in that described image
Recognize that the device of license number also sets up device including character repertoire, described character repertoire sets up device to be included:
Character segmentation device, the image pattern for will include license number is divided into single character;
Gaussian filtering device, for single character picture to be carried out into gaussian filtering;
Binaryzation device, the image for will carry out gaussian filtering carries out binary conversion treatment;
Normalized device, for carrying out Interpolating transform to the image after above-mentioned treatment, makes the width of image high high with Prototype drawing image width
Standard is consistent;
Feature-vector extraction means, the characteristic vector for extracting character by corner detection operator;
Character feature storehouse, for storing multiple character feature vectors.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310613538.5A CN103699876B (en) | 2013-11-26 | 2013-11-26 | Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310613538.5A CN103699876B (en) | 2013-11-26 | 2013-11-26 | Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103699876A CN103699876A (en) | 2014-04-02 |
CN103699876B true CN103699876B (en) | 2017-05-17 |
Family
ID=50361399
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310613538.5A Active CN103699876B (en) | 2013-11-26 | 2013-11-26 | Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103699876B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107967479B (en) * | 2016-10-19 | 2021-11-12 | 深圳怡化电脑股份有限公司 | Character recognition method and system with stained bill |
CN107203768B (en) * | 2017-06-12 | 2021-04-13 | 歌尔光学科技有限公司 | LED display screen digital automatic identification method and system |
CN109492641A (en) * | 2018-09-18 | 2019-03-19 | 上海延华智能科技(集团)股份有限公司 | Energy bills input method and system, storage medium, server based on image recognition |
CN110095296A (en) * | 2019-03-22 | 2019-08-06 | 南宁中车轨道交通装备有限公司 | A kind of control system and method for testing erosion control for the dynamic debugging of city rail vehicle |
CN111782552B (en) * | 2020-08-07 | 2021-05-18 | 广州极点三维信息科技有限公司 | Automatic testing method and device based on region division |
CN113327426B (en) * | 2021-05-26 | 2022-09-09 | 国能朔黄铁路发展有限责任公司 | Vehicle type code identification method and device and vehicle number identification method and device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101872416A (en) * | 2010-05-06 | 2010-10-27 | 复旦大学 | Vehicle license plate recognition method and system of road image |
CN102096821A (en) * | 2010-04-15 | 2011-06-15 | 西安理工大学 | Number plate identification method under strong interference environment on basis of complex network theory |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2004084121A1 (en) * | 2003-03-17 | 2004-09-30 | Fujitsu Limited | Car identifying method and device |
CN102043959B (en) * | 2010-12-28 | 2012-09-26 | 青岛海信网络科技股份有限公司 | License plate character segmentation method |
CN102693423B (en) * | 2012-05-15 | 2015-09-02 | 公安部第三研究所 | One pinpoint method of car plate under intense light conditions |
CN102722704B (en) * | 2012-06-12 | 2014-04-30 | 厦门宸天电子科技有限公司 | Method and system for recognizing vehicle license plate by integrating video dynamic tracking |
-
2013
- 2013-11-26 CN CN201310613538.5A patent/CN103699876B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102096821A (en) * | 2010-04-15 | 2011-06-15 | 西安理工大学 | Number plate identification method under strong interference environment on basis of complex network theory |
CN101872416A (en) * | 2010-05-06 | 2010-10-27 | 复旦大学 | Vehicle license plate recognition method and system of road image |
Non-Patent Citations (1)
Title |
---|
"复杂背景下的车辆牌照定位算法研究及实现";罗键;《中国优秀硕士学位论文全文数据库 信息科技辑》;20120115(第1期);正文第9页第1.4节、第二章第11-17页、第25-44页第3.2-3.3节 * |
Also Published As
Publication number | Publication date |
---|---|
CN103699876A (en) | 2014-04-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103699876B (en) | Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images | |
Pan et al. | A robust system to detect and localize texts in natural scene images | |
CN105046196B (en) | Front truck information of vehicles structuring output method based on concatenated convolutional neutral net | |
CN104751142B (en) | A kind of natural scene Method for text detection based on stroke feature | |
CN109426814B (en) | Method, system and equipment for positioning and identifying specific plate of invoice picture | |
CN111382704A (en) | Vehicle line-pressing violation judgment method and device based on deep learning and storage medium | |
CN108985170A (en) | Transmission line of electricity hanger recognition methods based on Three image difference and deep learning | |
CN106485274A (en) | A kind of object classification method based on target property figure | |
CN108537787B (en) | Quality judgment method for face image | |
CN106815583B (en) | Method for positioning license plate of vehicle at night based on combination of MSER and SWT | |
CN105760842A (en) | Station caption identification method based on combination of edge and texture features | |
Soomro et al. | Vehicle number recognition system for automatic toll tax collection | |
CN104700092A (en) | Small-character number identification method based on template and feature matching | |
CN104102909B (en) | Vehicle characteristics positioning and matching process based on lenticular information | |
CN110598566A (en) | Image processing method, device, terminal and computer readable storage medium | |
CN110751619A (en) | Insulator defect detection method | |
Garz et al. | A binarization-free clustering approach to segment curved text lines in historical manuscripts | |
CN105678737A (en) | Digital image corner point detection method based on Radon transform | |
CN108961262B (en) | Bar code positioning method in complex scene | |
CN111476804A (en) | Method, device and equipment for efficiently segmenting carrier roller image and storage medium | |
CN110689003A (en) | Low-illumination imaging license plate recognition method and system, computer equipment and storage medium | |
CN108921006B (en) | Method for establishing handwritten signature image authenticity identification model and authenticity identification method | |
Giri | Text information extraction and analysis from images using digital image processing techniques | |
Gui et al. | A fast caption detection method for low quality video images | |
CN113642570A (en) | Method for recognizing license plate of mine car in dark environment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |