CN107590500A - A kind of color recognizing for vehicle id method and device based on color projection classification - Google Patents
A kind of color recognizing for vehicle id method and device based on color projection classification Download PDFInfo
- Publication number
- CN107590500A CN107590500A CN201710596985.2A CN201710596985A CN107590500A CN 107590500 A CN107590500 A CN 107590500A CN 201710596985 A CN201710596985 A CN 201710596985A CN 107590500 A CN107590500 A CN 107590500A
- Authority
- CN
- China
- Prior art keywords
- color
- car plate
- license plate
- vehicle
- grader
- 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 Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of color recognizing for vehicle id method and device based on color projection, belong to technical field of image processing.Methods described includes car plate sample collection, training sample pretreatment, construction car plate color grader and color recognizing for vehicle id.Car plate color is identified the method that the present invention is projected by color, the type of car plate is judged according to car plate color, polytype number-plate number is identified, identification especially to new energy car plate, police car car plate can provide effectively booster action, effectively improve number-plate number recognition accuracy, in addition, it also can effectively aid in the identification of body color.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of color recognizing for vehicle id based on color projection classification
Method and device.
Background technology
Car license recognition is the core component of intelligent transportation system, in vehicle toll management, monitoring vehicle breaking regulation, traffic
Played an important role in the occasions such as traffic monitoring.Although the research for license plate recognition technology is more and more ripe, modern society
Meeting vehicle is more and more, and car plate type is also constantly updating, particularly the use of new energy car plate；Along with weather,
Illumination, shooting angle, it is stained, the influence of the factor such as torsional deformation, accurate Car license recognition facing challenges in real time are still very big.
Car license recognition process includes the steps such as License Plate, correction, segmentation, character recognition.With increasing for car plate type,
The difficulty increase of Car license recognition, it is used as the technology of Car license recognition auxiliary judgment condition and few using to car plate color and type at present
See.Existing colour recognition technology is all based on greatly the recognition methods of color histogram, SVMs, neutral net.
The patent that Chinese Patent Application No. is 201510607901.1 proposes a kind of body color recognition methods and device,
This method inputs training in advance by extracting the multidimensional color histogram feature in region to be identified, by multidimensional color histogram feature
Neutral net, colour type is determined according to the output result of the neutral net.This method identification object be body color and
Non- car plate color, and required training sample is more, it is computationally intensive, realize complicated.
The patent that Chinese Patent Application No. is 201310488682.0 proposes a kind of vehicle color identification method, the patent
In propose the method that body color reference zone is determined according to car plate color, car plate color is only divided into yellow and its by this method
Its class of color two, classification species are few, it is impossible to meet demand.
The patent that Chinese Patent Application No. is 201310117698.0 proposes a kind of vehicle license character and colours combine
Recognition methods, in the identification of car plate color, character is found in license plate area first after positioning, to non-character license plate area
Each pixel is counted, the rgb value of each pixel is converted to hsv color value, and these pixels are carried out with grader
Classification, then count a variety of colors according to classification results and distinguish how many pixel, last pixel color at most, is exactly
Car plate color.This method realizes and easily influenceed by illumination variation that robustness is bad complicated.
The content of the invention
To make up the deficiencies in the prior art, the present invention provides a kind of color recognizing for vehicle id method based on color projection classification
And device, car plate color can effectively be classified, and car plate type is judged according to car plate color, it is therefore an objective to according to difference
The different characteristics of type car plate carries out targetedly License Plate Character Segmentation and Number Reorganization.The color identification method accuracy rate
Height, robustness is good, suitable for more complicated scene, can effectively aid in the accuracy rate for improving the number-plate number or body color identification.
The present invention is achieved through the following technical solutions：
A kind of color recognizing for vehicle id method based on color projection classification, is characterized in that：Comprise the following steps：
（1）Car plate sample collection
According to the species for the car plate color to be identified, license plate image is gathered；
（2）Training sample pre-processes
To the image containing car plate collected, license plate area is extracted, can also may be used using whole license plate area as a sample
License plate area is divided into some sub-regions, per the sub-regions sample independent as one；Extract all pictures of each sample
R, G, B value of vegetarian refreshments, the pixel value for belonging to same passage is connected to form one by rowN*3EigenmatrixF, whereinNFor one
The number of all pixels point of individual sample, 3 represent tri- passages of R, G, B, and the pixel value of each passage is as a row；It is same belonging to
The eigenmatrix of a kind of other all samplesFPut together construction oneM*3Eigenmatrix, obtain training set eigenmatrixF train , whereinMFor the sum of the pixel of all samples；
（3）Construct car plate color grader
Calculate sample set eigenmatrixF train Within-cluster variance, total within-cluster variance, inter _ class relationship, and obtain accordingly
Fisher criterion functions；The solution of projection vector is obtained according to Fisher criterion functions；Select average or the weighting of projection centre flat
Threshold value of the average as decision function；Construct multi classifier；
（4）Color recognizing for vehicle id
To every two field picture in video flowing according to algorithm of locating license plate of vehicle positioning licence plate, license plate area is obtained；To whole license plate area
It is identified or some sub-regions is chosen in a license plate image and is identified；All pictures in each region to be identified of extraction
R, G of vegetarian refreshments, the pixel value of channel B, the pixel value that same passage is belonged in region is connected to form one by rowN*3Feature
MatrixMat feat , wherein, N is the number of all pixels point of a sample, and 3 represent tri- passages of R, G, B, the picture of each passage
Element value is as a row；EigenmatrixMat feat With the projection vector of first graderwMultiplication obtains projection matrixValue,
CalculateValueIn be more than the grader decision function threshold valuew 0 Element number account for element sum ratio c_rate；C_
Rate and given threshold c_thr relatively after selected to enter next grader, wherein c_thr span according to voting results
For (0,1)；Colour type according to belonging to the voting results of grader judge the region, if the whole car plate chosen is as identification
Object, then the result of grader output is exactly the colour type belonging to the car plate, if being carried out to some subregions in car plate
Identification, then obtain the recognition result of whole car plate color to this several recognition result comprehensive descision.
A kind of color recognizing for vehicle id method based on color projection classification of the present invention, step（2）Middle extraction license plate area
Can also Manual interception using algorithm of locating license plate of vehicle.
Further, a kind of color recognizing for vehicle id method based on color projection classification of the invention, step（3）Middle construction
Multi classifier includes two ways：1）One grader for being directed to multiclass of design, disposably considers all samples, and solve one
The optimization problem of individual multiple objective function, disposably obtain multiple classifying faces；2）It is rightkClass treats experimental data, trains a line two-by-two
Property identification and classification device, allows each grader to be voted to adjudicate classification, needs to design altogetherk*(k-1)/2Individual grader.
Further, a kind of color recognizing for vehicle id method based on color projection classification of the invention, step（4）Middle car plate
Positioning includes car plate coarse positioning and fine positioning；Car plate coarse positioning includes：Denoising, rim detection, vertical projection, shape are carried out to image
After state processing, the candidate region of car plate is obtained according to features such as the shapes of car plate；Car plate fine positioning includes：Coarse positioning is obtained
License plate area rotated, after the processing such as gray processing, binaryzation, rim detection, Character segmentation is come out, according to what is be partitioned into
Character number further determines whether that for license plate area, next step operation is carried out if license plate area is judged as.
For a kind of device for the color recognizing for vehicle id method based on color projection classification for realizing the present invention, including video
Data acquisition equipment and video data analytical equipment, the video data acquiring equipment are video camera；The video data analysis
Equipment is the one or more in video camera, NVR, PC, server；It is sequentially connected in the video data analytical equipment provided with car
Board locating module, eigenmatrix extraction module, colour recognition module, car plate color decision-making module.
The beneficial effects of the invention are as follows：
（1）The color recognizing for vehicle id accuracy rate of the present invention is high, preferable to the robustness of illumination variation, can be applied to more complicated field
Jing Zhong.
（2）Car plate color is identified using the method for color projection by the present invention, judges car plate according to car plate color
Type, then according to the characteristics of different type car plate carry out targetedly Character segmentation and identification, to polytype car plate
Number Reorganization, the especially identification to new energy car plate, police car car plate can provide effectively booster action, effectively improve car
Trade mark code recognition accuracy, in addition, also can effectively aid in the identification of body color.
（3）The color recognizing for vehicle id method based on color projection classification of the present invention, amount of calculation is small, realizes simply, in real time
Property is good.
Brief description of the drawings
Accompanying drawing 1 is the schematic flow sheet of the present invention；
Accompanying drawing 2 is the schematic flow sheet of the construction car plate color grader of the present invention；
Accompanying drawing 3 is the schematic flow sheet of the color recognizing for vehicle id process of the present invention；
Accompanying drawing 4 is the module diagram for the device that the present invention realizes the color recognizing for vehicle id method based on color projection classification.
Embodiment
The present invention is further detailed explanation with specific embodiment below in conjunction with the accompanying drawings, to help the technology of this area
Personnel have more complete, accurate and deep understanding to inventive concept of the invention, technical scheme, and protection scope of the present invention includes
But following examples are not limited to, it is any to technical scheme on the premise of without departing from spirit and scope
The modification that details and form are made is each fallen within protection scope of the present invention.
Embodiment 1
A kind of color recognizing for vehicle id method based on color projection classification, this method utilize different colours license plate image R, G, B tri-
Eigenmatrix is obtained after the pixel value pretreatment of passage, grader is trained according to Fisher criterions, and car is realized according to grader
Board colour recognition.As shown in figure 1, the present embodiment comprises the following steps：
S11, car plate sample collection
In order to improve the accuracy rate of color recognizing for vehicle id, it is necessary to assure the adequate and diversity of training sample.Specifically, pass through
Gather enough car plate sample images so that required all car plate color species are included in car plate sample image, by
Influenceed in the license plate image of candid photograph by illumination and shooting angle, even pattern gap that the car plate of same color is taken nor
Chang great, in order to improve the accuracy rate of colour recognition, the number of samples of each color car plate very little, and will can not as far as possible ensure car
Diversity in the class of board color category.
S12, pre-processed to obtain training setF train ；
Specifically, pretreatment is followed the steps below to the license plate image collected：1）License plate area is extracted, car can be used
Board location algorithm can also Manual interception, employ algorithm of locating license plate of vehicle in the present embodiment；2）Can be whole license plate area
As a sample, license plate area can also be divided into more sub-regions and using each sub-regions sample independent as one
This, in the present embodiment, extracts 4 ~ 5 sub-regions in each car plate, and subregion width size is about the width of a character,
Highly it is about the height of car plate；3）R, G, the pixel value of channel B of all pixels point of each sample are extracted, it is same logical belonging to
The pixel value in road is connected to form one by rowN*3EigenmatrixF, as shown in formula one, in formulawThe width of image is represented,h
The height of image is represented,N=w*hTri- passages of R, G, B, each passage are represented for the number of all pixels point of a sample, 3
Pixel value as one row；4）The eigenmatrix for all samples for belonging to same colour type is put together construction oneM*3
Eigenmatrix, obtain training sample eigenmatrixF train , as shown in formula two, whereinM=k*NFor sample pixel it is total
Number,kFor the number of training sample image.
Formula one
Formula two
S13, construction car plate color grader
The present embodiment is according to Fisher criterion structural classification devices, the general principle of Fisher criterions：Find a most suitable throwing
The axis of shadow, make two class samples projected on the axle between distance as far as possible, and per a kind of sample projection it is as compact as possible, from
And classifying quality is set to be optimal.Specifically, as shown in Fig. 2 comprising the steps of：
S131, to training sample setF train , according to the within-cluster variance of the calculating sample of formula threeS i , total class is calculated according to formula four
Interior dispersionS w , inter _ class relationship is calculated according to formula fiveS b ；
Formula three
Formula four
Formula five
Wherein,m i ForiThe average of class sample.
S132, it is to ensure that the class inherited after projection is maximum, degree of polymerization highest in class, defines criterion function such as formula six；
Formula six
S133, above formula is reached maximum, then obtain the solution of projection vectorw p For：
Formula seven
It is according to last solution is obtained after the normalized of formula eightw；
Formula eight
Due to being for the pixel value that the characteristic vector of training is tri- Color Channels of R, G, B, the then solution for the projection vector obtained
Include the column vector of three elements.
S134, average either threshold value of the weighted mean as decision function for selecting two class projection centres, calculation formula
As shown in formula nine, formula ten, formula 11, Car license recognition color sorter parameter is obtained, is selected in the present embodiment according to public affairs
Formula 11 calculates the threshold value of decision function；
Formula nine
Formula ten
Formula 11
Wherein, l1And l2It is the sum of the first kind and the second class sampled pixel point respectively.
S135, construction multi classifier, construction multi classifier can use two ways：One kind is that design one is directed to
The grader of multiclass, disposably consider all samples, and solve the optimization problem of a multiple objective function, disposably obtain multiple
Classifying face；Another kind is, rightkClass treats experimental data, trains a linear discrimination classification device two-by-two, allows each grader to be thrown
Ticket adjudicates classification, needs to design altogetherk*(k-1)/2Individual grader.In the present embodiment, second of make, root are selected
Multiple binary classifiers, such as the car plate for the three kinds of different colours that to classify are constructed according to the step in S13, then needs training three two
Class grader.
S14, car plate color judge
Car plate color is identified grader according to training, as shown in figure 3, comprising the following steps that：
S141, the car plate in video image is positioned, the License Plate includes car plate coarse positioning and fine positioning.Car plate is thick
Positioning specifically includes：After carrying out denoising, rim detection, vertical projection, Morphological scale-space to image, according to spies such as the shapes of car plate
Sign obtains the candidate region of car plate；Car plate fine positioning specifically includes：The license plate area that coarse positioning obtains is rotated, gray scale
After the processing such as change, binaryzation, rim detection, Character segmentation is come out, further determined whether according to the character number being partitioned into
For license plate area, the operation of next step is carried out if license plate area is judged as.
S142, to a car plate, whole license plate area can be identified, can also be chosen in a license plate area
Some sub-regions are identified respectively, then the color class according to belonging to the recognition result comprehensive descision car plate per sub-regions
Not.In the present embodiment, chooseMSub-regions are identified, and the step of every sub-regions are identified is as follows：1）Extraction is every
R, G, the pixel value of channel B of all pixels point of individual subregion to be identified, belong to same as shown in formula 12, in sub-areas
The pixel value of one passage is connected to form one by rowN*3EigenmatrixMat feat , in formulawThe width of image is represented,hRepresent
The height of image,N=w*hTri- passages of R, G, B, the picture of each passage are represented for the number of all pixels point of a sample, 3
Element value is as a row；2）EigenmatrixMat feat According to formula 13 and the projection vector of first graderwMultiplication obtains
Projection matrixValue, calculating matrixValueIn be more than the grader decision function threshold valuew 0 The number of element account forValueIn
The ratio c_rate of element sum；3）C_rate compared with given threshold c_thr, entered according to voting results selection next
Grader, c_thr span is (0,1), takes c_thr=0.5 in the present embodiment；4）According to the voting results of grader
Judge the colour type belonging to the subregion；5）After processing all is identified to every sub-regions, statistics is in each respectively
The number of the car plate subregion of colour type, the largest number of colour types of subregion are the colour type of the car plate.
Formula 12
Formula 13
S143, car plate type, such as yellow card, new energy car plate, police car car plate etc. are judged according to car plate color.Different types of car
The characteristics of board, is different, such as the digit of the number-plate number, Chinese letters order etc., and judging can be according to these feature after car plate type
Targetedly carry out follow-up License Plate Segmentation, Number Reorganization.
Embodiment 2
The present embodiment provides a kind of dress for the color recognizing for vehicle id method based on color projection classification for being used to realize embodiment 1
Put, as shown in figure 4, the device includes：
Video data acquiring equipment, for all equipment that can gather video image, the present embodiment is taken the photograph using Internet video
Camera；
Video data processing device, can be the equipment such as video camera, NVR, PC, server, the present embodiment is using network video
Frequency video camera；
Specifically, the processing procedure in the video data processing device also includes following module：
License Plate module, for the car plate in positioning video image；
Eigenmatrix extraction module, the eigenmatrix of the overall or more sub-regions for extracting the car plate navigated to；
Colour recognition module, the color of license plate area is identified according to eigenmatrix using the grader trained；
Car plate color decision-making module, for judging the color of car plate according to the colour recognition result of car plate entirety or subregion.
It will be understood to those skilled in the art that program can be passed through by realizing all or part of step of above-described embodiment
Related hardware is instructed to complete, above-mentioned program can be stored in computer read/write memory medium, and the program is performing
When, execution the step of including above method embodiment；The storage medium includes：ROM, RAM etc. are various can be with storage program
The medium of code.
Claims (5)
- A kind of 1. color recognizing for vehicle id method based on color projection classification, it is characterised in that：Comprise the following steps：（1）Car plate sample collectionAccording to the species for the car plate color to be identified, license plate image is gathered；（2）Training sample pre-processesTo the image containing car plate collected, license plate area is extracted, can also may be used using whole license plate area as a sample License plate area is divided into some sub-regions, per the sub-regions sample independent as one；Extract all pictures of each sample R, G of vegetarian refreshments, the pixel value of channel B, the pixel value for belonging to same passage is connected to form eigenmatrix by rowF, it is same belonging to The eigenmatrix of a kind of other all samplesFPut together construction training set eigenmatrixF train ；（3）Construct car plate color graderCalculate training set eigenmatrixF train Within-cluster variance, total within-cluster variance, inter _ class relationship, and obtain accordingly Fisher criterion functions；The solution of projection vector is obtained according to Fisher criterion functions；Select average or the weighting of projection centre flat Threshold value of the average as decision function；Construct multi classifier；（4）Color recognizing for vehicle idTo every two field picture in video flowing according to algorithm of locating license plate of vehicle positioning licence plate, license plate area is obtained；To whole license plate area It is identified or some sub-regions is chosen in a license plate image and is identified；All pictures in each region to be identified of extraction R, G of vegetarian refreshments, the pixel value of channel B, the pixel value that same passage is belonged in region is connected to form an eigenmatrix by rowMat feat , eigenmatrixMat feat With the projection vector of first graderwMultiplication obtains projection matrixValue, calculate projection MatrixValueIn be more than the grader decision function threshold valuew 0 Element number account for element sum ratio c_rate；C_ Rate selects to enter next grader, wherein c_thr span compared with given threshold c_thr according to voting results For (0,1)；Colour type according to belonging to the voting results of grader judge the region, if the whole car plate chosen is as identification Object, then the result of grader output is exactly the colour type belonging to the car plate, if being carried out to some subregions in car plate Identification, then the recognition result of whole car plate color is obtained to this several recognition result comprehensive descision.
- A kind of 2. color recognizing for vehicle id method based on color projection classification according to claim 1, it is characterised in that：Step Suddenly（2）Middle extraction license plate area can also Manual interception using algorithm of locating license plate of vehicle.
- A kind of 3. color recognizing for vehicle id method based on color projection classification according to claim 1, it is characterised in that：Step Suddenly（3）Middle construction multi classifier includes two ways：1）One grader for being directed to multiclass of design, disposably considers all samples This, and the optimization problem of a multiple objective function is solved, disposably obtain multiple classifying faces；2）It is rightkClass treats experimental data, two-by-two A linear discrimination classification device is trained, allows each grader to be voted to adjudicate classification, needs to design altogetherk*(k-1)/2Individual point Class device.
- A kind of 4. color recognizing for vehicle id method based on color projection classification according to claim 1, it is characterised in that：Step Suddenly（4）Middle License Plate includes car plate coarse positioning and fine positioning；Car plate coarse positioning includes：To image carry out denoising, rim detection, After vertical projection, Morphological scale-space, the candidate region of car plate is obtained according to features such as the shapes of car plate；Car plate fine positioning includes： The license plate area that coarse positioning obtains is rotated, after the processing such as gray processing, binaryzation, rim detection, Character segmentation is come out, Character number according to being partitioned into further determines whether that for license plate area, next step behaviour is carried out if license plate area is judged as Make.
- 5. for realizing a kind of color recognizing for vehicle id method based on color projection classification any one of claim 1-4 Device, including video data acquiring equipment and video data analytical equipment, it is characterised in that：The video data acquiring equipment For video camera；The video data analytical equipment is the one or more in video camera, NVR, PC, server；The video counts Determined according to being sequentially connected in analytical equipment provided with License Plate module, eigenmatrix extraction module, colour recognition module, car plate color Plan module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710596985.2A CN107590500A (en) | 2017-07-20 | 2017-07-20 | A kind of color recognizing for vehicle id method and device based on color projection classification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710596985.2A CN107590500A (en) | 2017-07-20 | 2017-07-20 | A kind of color recognizing for vehicle id method and device based on color projection classification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107590500A true CN107590500A (en) | 2018-01-16 |
Family
ID=61042277
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710596985.2A Pending CN107590500A (en) | 2017-07-20 | 2017-07-20 | A kind of color recognizing for vehicle id method and device based on color projection classification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107590500A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537223A (en) * | 2018-04-16 | 2018-09-14 | 广东工业大学 | A kind of detection method of license plate, system and equipment and storage medium |
| CN108765967A (en) * | 2018-06-12 | 2018-11-06 | 安徽灵图壹智能科技有限公司 | One kind preventing the occupied alarm method in new energy private car parking stall and system |
| CN109304306A (en) * | 2018-09-19 | 2019-02-05 | 广东省智能制造研究所 | Production line articles sorting method, system and articles sorting system |
| CN111046891A (en) * | 2018-10-11 | 2020-04-21 | 杭州海康威视数字技术股份有限公司 | Training method of license plate recognition model, and license plate recognition method and device |
| CN111798393A (en) * | 2020-06-30 | 2020-10-20 | 深圳市慧鲤科技有限公司 | Image processing method and device, electronic device and storage medium |
| CN111815547A (en) * | 2020-06-30 | 2020-10-23 | 深圳市慧鲤科技有限公司 | Image processing method and device, electronic device and storage medium |
| CN113361534A (en) * | 2021-06-01 | 2021-09-07 | 雅宝科技(深圳)有限公司 | Full-automatic car washing method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101136141A (en) * | 2007-10-12 | 2008-03-05 | 清华大学 | Vehicle type classification method based on single frequency continuous-wave radar |
| CN103207992A (en) * | 2013-04-07 | 2013-07-17 | 布法罗机器人科技（苏州）有限公司 | Character and color combined recognition method of license plates |
| CN103870803A (en) * | 2013-10-21 | 2014-06-18 | 北京邮电大学 | Vehicle license plate recognition method and system based on coarse positioning and fine positioning fusion |
| CN105488538A (en) * | 2015-12-15 | 2016-04-13 | 云南电网有限责任公司电力科学研究院 | Method for identifying clouds and sky of sky image based on improved k-means clustering algorithm |
-
2017
- 2017-07-20 CN CN201710596985.2A patent/CN107590500A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101136141A (en) * | 2007-10-12 | 2008-03-05 | 清华大学 | Vehicle type classification method based on single frequency continuous-wave radar |
| CN103207992A (en) * | 2013-04-07 | 2013-07-17 | 布法罗机器人科技（苏州）有限公司 | Character and color combined recognition method of license plates |
| CN103870803A (en) * | 2013-10-21 | 2014-06-18 | 北京邮电大学 | Vehicle license plate recognition method and system based on coarse positioning and fine positioning fusion |
| CN105488538A (en) * | 2015-12-15 | 2016-04-13 | 云南电网有限责任公司电力科学研究院 | Method for identifying clouds and sky of sky image based on improved k-means clustering algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 刘金帅等: "基于YCbCr颜色空间和Fisher判别分析的棉花图像分割研究", 《作物学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537223A (en) * | 2018-04-16 | 2018-09-14 | 广东工业大学 | A kind of detection method of license plate, system and equipment and storage medium |
| CN108537223B (en) * | 2018-04-16 | 2022-02-15 | 广东工业大学 | License plate detection method, system and equipment and storage medium |
| CN108765967A (en) * | 2018-06-12 | 2018-11-06 | 安徽灵图壹智能科技有限公司 | One kind preventing the occupied alarm method in new energy private car parking stall and system |
| CN109304306A (en) * | 2018-09-19 | 2019-02-05 | 广东省智能制造研究所 | Production line articles sorting method, system and articles sorting system |
| CN109304306B (en) * | 2018-09-19 | 2020-08-11 | 广东省智能制造研究所 | Production line object sorting method and system and object sorting system |
| CN111046891A (en) * | 2018-10-11 | 2020-04-21 | 杭州海康威视数字技术股份有限公司 | Training method of license plate recognition model, and license plate recognition method and device |
| CN111798393A (en) * | 2020-06-30 | 2020-10-20 | 深圳市慧鲤科技有限公司 | Image processing method and device, electronic device and storage medium |
| CN111815547A (en) * | 2020-06-30 | 2020-10-23 | 深圳市慧鲤科技有限公司 | Image processing method and device, electronic device and storage medium |
| CN113361534A (en) * | 2021-06-01 | 2021-09-07 | 雅宝科技(深圳)有限公司 | Full-automatic car washing method and system |
| CN113361534B (en) * | 2021-06-01 | 2022-06-24 | 深圳市雅宝智能装备系统有限公司 | Full-automatic car washing method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107590500A (en) | A kind of color recognizing for vehicle id method and device based on color projection classification | |
| CN105373794B (en) | A kind of licence plate recognition method | |
| Ruta et al. | Real-time traffic sign recognition from video by class-specific discriminative features | |
| CN109271991B (en) | License plate detection method based on deep learning | |
| CN104951784B (en) | A kind of vehicle is unlicensed and license plate shading real-time detection method | |
| CN102043945B (en) | License plate character recognition method based on real-time vehicle tracking and binary index classification | |
| CN101923640B (en) | Method for distinguishing false iris images based on robust texture features and machine learning | |
| CN106599792B (en) | Method for detecting hand driving violation behavior | |
| CN102346847B (en) | License plate character recognizing method of support vector machine | |
| Dlagnekov | License plate detection using adaboost | |
| CN108564030A (en) | Classifier training method and apparatus towards vehicle-mounted thermal imaging pedestrian detection | |
| CN107103317A (en) | Fuzzy license plate image recognition algorithm based on image co-registration and blind deconvolution | |
| CN109918971B (en) | Method and device for detecting number of people in monitoring video | |
| CN108549864A (en) | Area-of-interest filter method towards vehicle-mounted thermal imaging pedestrian detection and device | |
| CN106529532A (en) | License plate identification system based on integral feature channels and gray projection | |
| CN104751136A (en) | Face recognition based multi-camera video event retrospective trace method | |
| CN103577838A (en) | Face recognition method and device | |
| CN102521565A (en) | Garment identification method and system for low-resolution video | |
| CN109063619A (en) | A kind of traffic lights detection method and system based on adaptive background suppression filter and combinations of directions histogram of gradients | |
| CN106650731A (en) | Robust license plate and logo recognition method | |
| CN110598560A (en) | Night monitoring and identifying method and system based on neural network enhancement | |
| CN107563413A (en) | The accurate extracting method of unmanned plane image farmland block object | |
| CN102163278B (en) | Illegal vehicle intruding detection method for bus lane | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| CN107729812A (en) | A kind of vehicle color suitable for monitoring scene knows method for distinguishing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180116 |