CN106056086B - Vehicle brand type identifier method based on Fast Learning frame - Google Patents
Vehicle brand type identifier method based on Fast Learning frame Download PDFInfo
- Publication number
- CN106056086B CN106056086B CN201610383537.XA CN201610383537A CN106056086B CN 106056086 B CN106056086 B CN 106056086B CN 201610383537 A CN201610383537 A CN 201610383537A CN 106056086 B CN106056086 B CN 106056086B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- feature
- license plate
- vehicle face
- classifier
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The vehicle brand type identifier method based on Fast Learning frame that the invention discloses a kind of, comprising the following steps: build spark platform, obtain monitoring video flow；License plate is detected from the video image of acquisition, according to license plate position acquisition vehicle face image, face identification of driving a vehicle of going forward side by side；It according to the width and height of license plate, extends up and down, obtains vehicle face image；Multiple dimensioned vehicle face image is constructed, is there is the piecemeal of overlapping, partial binary and gradient orientation histogram feature are extracted to each fritter, and be combined into final vehicle face feature；Training multi-stage cascade classifier, the first order is by the way of multi-categorizer ballot, and when most classifiers are thrown to same target, otherwise just receiving should be as a result, enters next stage；The second level uses Ensemble classifier system, and by Feature Mapping, the multiple sub-classifiers of training merge each classifier result, obtain final classification results.Accurately, vehicle brand and concrete model can be quickly and stably identified in the data of magnanimity.
Description
Technical field
The present invention relates to image steganalysis and intelligent transportation field, more particularly to a kind of based on Fast Learning frame
Vehicle brand type identifier method.
Background technique
With the sustainable development of national economy, motor vehicle has become communications and transportation indispensable in people's daily life
Tool.But the following traffic problems also become increasingly conspicuous.Countries in the world all step up its investment to the management of traffic system, gradually
Form control of traffic and road research field.
Vehicle cab recognition is main task and key technology in intelligent transportation system, it has a wide range of applications, such as highway
Automatic charging, parking lot management, inspection theft vehicle, fake-licensed car management etc..The difficult point of vehicle cab recognition is the general of identifying system
All over adaptability, a good vehicle identification system is required to adapt to various scene changes, Changes in weather and vehicle itself
Abrasion etc..
Vehicle identification is concentrated mainly on vehicle cab recognition and vehicle-logo recognition both direction at present.Vehicle cab recognition is to distinguish vehicle kind
Based on style number, e.g., car, truck, bus etc., method is the most extensive with the identification based on induction coil, but should
Method installation is inconvenient, maintenance cost is high and discrimination is low.Vehicle-logo recognition can identify vehicle brand, but different brand vehicles
Mark changes greatly, and logo very little, so that logo detection is relatively difficult.
With the rapid growth of traffic information stream, data volume has reached very high scale.Traditional computing system face
To such mass data, meet demand has been difficult it.
Chinese patent literature CN 105574543 discloses a kind of vehicle brand type identifier method based on deep learning,
Including step 1, training data establishes SVM license plate discrimination model；Step 2, vehicle cab recognition mould is established based on deep approach of learning training
Type；Step 3, to the background of image in video carry out modeling obtain moving target, pursuit movement Target Acquisition movement objective orbit,
Obtain the picture comprising license plate；Step 4, the picture is handled by image processing techniques, obtains several comprising license plate
Segment is differentiated with the SVM license plate discrimination model and retains the segment comprising license plate；Step 5, according to the position of license plate, respectively to
Upper and lower, left and right four direction extends setting regions, obtains the band of position of headstock；Step 6, it is based on the vehicle cab recognition model
Vehicle is identified according to the band of position of the headstock.Although this kind of method uses one using autonomous learning feature
A classifier, classifying quality are poor.And when facing mass data, processing speed is slow, and accuracy rate is relatively low.
Summary of the invention
In view of the above technical problems, object of the present invention is to: a kind of vehicle brand model based on Fast Learning frame is provided
Recognition methods accurately, can quickly and stably identify vehicle brand and concrete model in the data of magnanimity.
The technical scheme is that
A kind of vehicle brand type identifier method based on Fast Learning frame, which comprises the following steps:
S01: building spark platform, obtains monitoring video flow；
S02: detecting license plate from the video image of acquisition, according to license plate position acquisition vehicle face image, face knowledge of driving a vehicle of going forward side by side
Not；
S03: according to the width and height of license plate, extending up and down, obtains vehicle face image；
S04: constructing multiple dimensioned vehicle face image, there is the piecemeal of overlapping, extracts partial binary (LBP) to each fritter
With gradient orientation histogram (HOG) feature, and it is combined into final vehicle face feature；
S05: training multi-stage cascade classifier, the first order using multi-categorizer vote by the way of, when most classifiers throw to
When same target, otherwise just receiving should be as a result, enters next stage；The second level uses Ensemble classifier system, passes through Feature Mapping, instruction
Practice multiple sub-classifiers, merges each classifier result, obtain final classification results.
Preferably, the step S01 includes:
S11: video flowing is received by Spark Streaming；
Video flowing: being divided into discrete RDD data set according to time interval by S12, this RDD data set is converted into multiple
Sub- RDD data set, the corresponding picture of every sub- RDD data set, handles every picture respectively.
Preferably, in the step S03, left and right extends 1.5 times of license plate width, downwards the license plate height of one times of extension,
Extend up 3.5 times of license plate height.
It preferably, further include calculating the tilt angle of license plate, and rotate image at this angle to license plate after step S03
Lower edge is in horizontal direction.
It preferably, further include vehicle face characteristic point being detected, according to minimum range and location consistency principle after step S03
Matching characteristic point adjusts vehicle face image according to characteristic point position.
Preferably, the adjustment vehicle face image the following steps are included:
S31: calculating matching characteristic point and the straight slope absolute value of characteristic point institute shape to be matched, if less than 5 degree, plus
Enter candidate matches point；
S32: the feature for calculating all candidate feature points describes operator, and fast robust (surf) characteristic point uses four haar
Wavelet character, calculates all features and describes the feature of operator and characteristic point to be matched and describe Euclidean distance between operator, if most
Small distance is less than threshold value d, then this feature point and to be matched successful match；
S33: the midpoint abscissa x of all match points is calculatedi, i=1,2 ... n, n are coupling number, if μ and σ are respectively it
Mean value and standard deviation filter out and all meet μ -3 σ < xiThe point of+3 σ of < μ, and calculate their median average c again, according to c with
The right boundary of the distance adjustment vehicle face of license plate central point.
Preferably, the step S04 the following steps are included:
S41: normalizing to fixed size for vehicle face, is divided into block, and every piece is separated into multiple units, according to a unit to the right
To the mode sliding shoe of next unit, the feature of each block is calculated；
S42: using two-dimensional Gaussian kernel G (x, y, σ) and vehicle face image I (x, y) convolution algorithm, different scale space is generated
Vehicle face image L (x, y, σ), scale space form indicate are as follows:
L (x, y, σ)=G (x, y, σ) * I (x, y)
Wherein,
S43:LBP feature uses 8 neighborhoods, and the mode that radius is 1 calculates, and counts the histogram of each unit, histogram bin
Number is 8, and each piece of all unit LBP features are connected, and forms each piece of feature.
S44: calculating the gradient magnitude of pixel and direction in each unit, and 360 degree of directions are divided into multiple directions block, width
Value counts the gradient orientation histogram of each unit as weight, and the feature of each piece of all units is connected, and is formed every
The feature of a block, and connect to form the final vehicle face feature of the block with the feature in step S43.
Compared with prior art, the invention has the advantages that
1, the present invention accurately, can quickly and stably identify vehicle brand and concrete model in the data of magnanimity.Identification
Accuracy is high, speed is fast, applicability is wide.Constructing multiple dimensioned vehicle face image can be by expression that feature is simple and efficient in different scale
On, it is conducive to image recognition；Each piecemeal and two features are used and divided using the elasticity distribution formula data set (RDD) in spark
Cloth mode calculates, and which greatly enhances computational efficiencies.
2, it obtains to make to classify as a result, comprehensively utilize the respective advantage of Various Classifiers on Regional using multi-stage cascade classifier
More robust, available more powerful classifying quality.Each classifier, which differentiates, realizes distributed computing using RDD, substantially increases
Computational efficiency.
Detailed description of the invention
The invention will be further described with reference to the accompanying drawings and embodiments:
Fig. 1 is that the present invention is based on the flow charts of the vehicle brand type identifier method of Fast Learning frame.
Fig. 2 is that the present invention is based on the vehicle cab recognition flow charts of the vehicle brand type identifier method of Fast Learning frame；
Fig. 3 is that the present invention is based on the classifiers of the vehicle brand type identifier method of Fast Learning frame to differentiate flow chart.
Specific embodiment
In order to make the objectives, technical solutions and advantages of the present invention clearer, With reference to embodiment and join
According to attached drawing, the present invention is described in more detail.It should be understood that these descriptions are merely illustrative, and it is not intended to limit this hair
Bright range.In addition, in the following description, descriptions of well-known structures and technologies are omitted, to avoid this is unnecessarily obscured
The concept of invention.
Embodiment:
A kind of vehicle brand type identifier method based on Fast Learning frame, comprising the following steps:
(1) according to monitoring demand, efficiency of algorithm and machine performance, spark platform is built in selection, obtains monitoring video flow.
Spark is a parallel data processing frame, it can allow big data to be combined together with real time data application, can be used
Spark Streaming handles real time data.
(2) license plate is detected from the video image of acquisition.
(3) according to license plate position acquisition vehicle face image, face identification of driving a vehicle of going forward side by side, specific steps are as follows:
(3-1) extends up and down according to the width and height of license plate, obtains vehicle face image.
(3-2) calculates the tilt angle of license plate, and carrousel face is to level.Vehicle face image can be prevented because monitored in this way
The influence of equipment setting angle and steering direction etc. tilts, and generates large effect to subsequent vehicle cab recognition.
(3-3) detects vehicle face characteristic point, according to minimum range and location consistency principle matching characteristic point.According to characteristic point
Position adjusts vehicle face image.In this way, the positioning of car plate detection right boundary will appear deviation, the vehicle face position detected is caused also
There is large error, the boundary of vehicle face is readjusted using characteristic point, vehicle face can be made to be aligned.
(3-4) constructs multiple dimensioned vehicle face image, is there is the piecemeal of overlapping, extracts partial binary to each fritter
(LBP) and gradient orientation histogram (HOG) feature, and it is combined into final vehicle face feature.In spark, each piecemeal is produced
Respective RDD realizes distributed computing.LBP mainly expresses image texture, and HOG mainly expresses resemblance, and the two combines can be with
The feature of expression vehicle face well；There is the piecemeal of overlapping that can express local feature and embody global property；It is multiple dimensioned can
The expression that feature is simple and efficient is conducive to image recognition on different scale；Utilize the elasticity distribution formula data set in spark
(RDD) each piecemeal and two features are calculated using distributed way, which greatly enhances computational efficiencies.
(3-5) train two-stage cascade classifier, the first order using multi-categorizer vote by the way of, when most classifiers throw to
When same target, otherwise just receiving should be as a result, enters next stage；The second level uses Ensemble classifier system, passes through Feature Mapping, instruction
Practice multiple sub-classifiers, merges each classifier result, obtain final classification results.In spark, producible pair of each classifier
The RDD answered realizes distributed computing.The beneficial effect done so is: single classifier is limited to classifier itself, is difficult to reach
To outstanding as a result, the comprehensive utilization respective advantage of Various Classifiers on Regional, can make more robust of classifying；Classifier is cascaded just as spy
Sign cascade is the same, available more powerful classifying quality；With the feature calculation in (3-4), each classifier differentiates real using RDD
Existing distributed computing, substantially increases computational efficiency.
As shown in Figure 1, step (1) specifically includes the following steps:
(1) video flowing is received by Spark Streaming.
(2) video flowing is divided into RDD data set discrete one by one according to time interval, due to there is multi-channel video, often
A RDD may include many image datas, this RDD is converted into multiple sub- RDD, the corresponding picture of every sub- RDD.
(3) vehicle cab recognition processing is carried out to every picture.
As shown in Fig. 2, vehicle face identification specifically includes the following steps:
(1) prepare training sample and test sample, a total of 1032 kinds of vehicles in this example.
(2) by texture and color combine in the way of, positioning licence plate.
(3) Radon transform is utilized, the tilt angle of license plate is calculated, and rotation image to license plate lower edge is in water at this angle
Square to.
(4) according to license plate position, left and right extends 1.5 times of license plate width, downwards the license plate height of one times of extension, expands upwards
The license plate height of 3.5 times of exhibition obtains vehicle face image according to this.
(5) rapid robust feature point (surf) is calculated, finds symmetric points therein, and vehicle face side is adjusted according to characteristic point
Boundary.Specifically:
(5-1) calculation amount matching characteristic point and the straight slope absolute value of characteristic point institute shape to be matched, if less than 5 degree,
Candidate matches point is then added.
The feature that (5-2) calculates all candidate feature points describes operator, and surf characteristic point uses four haar wavelet characters,
It is respectively as follows: the sum of the sum of horizontal direction value, horizontal direction absolute value, the sum of vertical direction, the sum of vertical direction absolute value.It calculates
The feature that all features describe operator and characteristic point to be matched describes the Euclidean distance between operator, if minimum range is less than threshold value
D, then this feature point and to be matched successful match.
(5-3) calculates the midpoint abscissa x of all match pointsi, i=1,2 ... n, n are coupling number, if μ and σ are respectively it
Mean value and standard deviation.It filters out and all meets μ -3 σ < xiThe point of+3 σ of < μ, and calculate their median average c again, according to c with
The right boundary of the distance adjustment vehicle face of license plate central point.
(6) feature extraction of vehicle face, specific steps include:
Vehicle face is normalized to 128 × 64 sizes by (6-1), is divided into 4 × 2 pieces, and every piece is separated into 2 × 2 units.According to
A right unit calculates the feature of each block to the mode sliding shoe of next unit.
(6-2) generates different scale space using two-dimensional Gaussian kernel G (x, y, σ) and vehicle face image I (x, y) convolution algorithm
Vehicle face image L (x, y, σ), scale space form indicate are as follows:
L (x, y, σ)=G (x, y, σ) * I (x, y)
Wherein, * is convolution algorithm,
This example selects the Gaussian kernel of 5 × 5 sizes, σ ∈ { 5,10 }.
(6-3) LBP feature uses 8 neighborhoods, and the mode that radius is 1 calculates, and counts the histogram of each unit, histogram
Bin number is 8.Each piece of all unit LBP features are connected, each piece of feature is formed.
(6-4) calculates the gradient magnitude of pixel and direction in each unit, and 360 degree of directions are divided into 8 direction blocks, width
Value counts the gradient orientation histogram of each unit as weight.The feature of each piece of all units is connected, is formed every
The feature of a block, and connect to form the final feature of the block with the feature in (6-2).Total characteristic dimension is 7 × 3 × 4 × (8+8)
× 3=4032.
(6-5) in spark, the corresponding RDD of each scale vehicle face, the sub- RDD of each piece of regeneration, each sub- RDD can
Distributed computing is carried out, calculating speed is greatly improved.
(6-6) extracts feature due to there is the sliding shoe of overlapping, feature redundancy inevitably occurs, finally uses principal component analysis
Method (PCA) reduces intrinsic dimensionality.Final dimension is 1200 in this example.
As shown in figure 3, this example is classified vehicle face using the cascade mode of two-level classifier, naturally it is also possible to using three-level or
More.
The first order is voted with three kinds of classifiers, and three kinds of classifiers are respectively support vector machines, artificial neural network and random
Forest.When any two classifier is identified as same type, otherwise just receiving should be as a result, enters next stage；The second level uses
Rotate forest Ensemble classifier system.By Feature Mapping, the feature set of multiple rotation transformations is generated, utilizes the training of these feature sets
Multiple sub-classifiers, sub-classifier uses support vector machines in this example, finally integrates each sub-classifier score, obtains final point
Class result.In spark, each classifier will generate corresponding RDD, by distributed computing, quickly obtain result.
It should be understood that above-mentioned specific embodiment of the invention is used only for exemplary illustration or explains of the invention
Principle, but not to limit the present invention.Therefore, that is done without departing from the spirit and scope of the present invention is any
Modification, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.In addition, appended claims purport of the present invention
Cover the whole variations fallen into attached claim scope and boundary or this range and the equivalent form on boundary and
Modification.
Claims (5)
1. a kind of vehicle brand type identifier method based on Fast Learning frame, which comprises the following steps:
S01: building spark platform, obtains monitoring video flow；
S02: detecting license plate from the video image of acquisition, according to license plate position acquisition vehicle face image, face identification of driving a vehicle of going forward side by side；
S03: according to the width and height of license plate, extending up and down, obtains vehicle face image；Vehicle face characteristic point is detected, according to most
Small distance and location consistency principle matching characteristic point adjust vehicle face image according to characteristic point position, comprising:
S31: matching characteristic point and the straight slope absolute value of characteristic point institute shape to be matched are calculated, if time is added less than 5 degree
Select match point；
S32: the feature for calculating all candidate feature points describes operator, and fast robust surf characteristic point uses four small bauds of haar
Sign, calculates all features and describes the feature of operator and characteristic point to be matched and describe Euclidean distance between operator, if minimum range
Less than threshold value d, then this feature point and to be matched successful match；
S33: the midpoint abscissa x of all match points is calculatedi, i=1,2 ... n, n are coupling number, if μ and σ be respectively its mean value and
Standard deviation filters out and all meets μ -3 σ < xiThe point of+3 σ of < μ, and their median average c is calculated again, according to c and license plate
The right boundary of the distance adjustment vehicle face of central point；
S04: constructing multiple dimensioned vehicle face image, there is the piecemeal of overlapping, extracts partial binary LBP and gradient to each fritter
Direction histogram HOG feature, and it is combined into final vehicle face feature；
S05: training multi-stage cascade classifier, the first order is by the way of multi-categorizer ballot, when most classifiers are thrown to same
When target, receive as a result, otherwise entering next stage；The second level uses Ensemble classifier system, passes through Feature Mapping, the multiple sons of training
Classifier merges each classifier result, obtains final classification results.
2. the vehicle brand type identifier method according to claim 1 based on Fast Learning frame, which is characterized in that institute
Stating step S01 includes:
S11: video flowing is received by Spark Streaming；
Video flowing: being divided into discrete RDD data set according to time interval by S12, this RDD data set is converted into multiple sons
RDD data set, the corresponding picture of every sub- RDD data set, handles every picture respectively.
3. the vehicle brand type identifier method according to claim 1 based on Fast Learning frame, which is characterized in that institute
It states in step S03, left and right extends 1.5 times of license plate width, and the license plate height of one times of extension, extends up 3.5 times of vehicle downwards
Board height.
4. the vehicle brand type identifier method according to claim 1 based on Fast Learning frame, which is characterized in that step
Further include after rapid S03 calculate the tilt angle of license plate, and at this angle rotation image to license plate lower edge in horizontal direction.
5. the vehicle brand type identifier method according to claim 1 based on Fast Learning frame, which is characterized in that institute
State step S04 the following steps are included:
S41: normalizing to fixed size for vehicle face, is divided into block, and every piece is separated into multiple units, calculates the feature of each block；
S42: using two-dimensional Gaussian kernel G (x, y, σ) and vehicle face image I (x, y) convolution algorithm, the vehicle face in different scale space is generated
Image L (x, y, σ), scale space form indicate are as follows:
L (x, y, σ)=G (x, y, σ) * I (x, y)
Wherein,
S43:LBP feature uses 8 neighborhoods, and the mode that radius is 1 calculates, and counts the histogram of each unit, histogram bin number
It is 8, each piece of all unit LBP features is connected, forms each piece of feature；
S44: calculating the gradient magnitude of pixel and direction in each unit, 360 degree of directions is divided into multiple directions block, amplitude is made
The gradient orientation histogram that each unit is counted for weight connects the feature of each piece of all units, forms each piece
Feature, and connect to form the final vehicle face feature of the block with the feature in step S43.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610383537.XA CN106056086B (en) | 2016-06-02 | 2016-06-02 | Vehicle brand type identifier method based on Fast Learning frame |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610383537.XA CN106056086B (en) | 2016-06-02 | 2016-06-02 | Vehicle brand type identifier method based on Fast Learning frame |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106056086A CN106056086A (en) | 2016-10-26 |
| CN106056086B true CN106056086B (en) | 2019-05-17 |
Family
ID=57173168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610383537.XA Active CN106056086B (en) | 2016-06-02 | 2016-06-02 | Vehicle brand type identifier method based on Fast Learning frame |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106056086B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022429B (en) * | 2016-11-04 | 2021-08-27 | 浙江大华技术股份有限公司 | Vehicle detection method and device |
| CN106778560B (en) * | 2016-12-01 | 2020-08-04 | 中科唯实科技（北京）有限公司 | Vehicle type recognition method based on FHOG characteristics and Linear SVM |
| CN108268866B (en) * | 2016-12-30 | 2020-07-10 | 纵目科技（上海）股份有限公司 | Vehicle detection method and system |
| CN106991420A (en) * | 2017-03-27 | 2017-07-28 | 新智认知数据服务有限公司 | A kind of detection method of license plate of the license plate area regression technique based on piecemeal |
| CN107085723A (en) * | 2017-03-27 | 2017-08-22 | 新智认知数据服务有限公司 | A kind of characters on license plate global recognition method based on deep learning model |
| CN107038442A (en) * | 2017-03-27 | 2017-08-11 | 新智认知数据服务有限公司 | A kind of car plate detection and global recognition method based on deep learning |
| CN107221070A (en) * | 2017-05-24 | 2017-09-29 | 广州市银科电子有限公司 | A kind of bill anti-counterfeit discrimination method recognized based on master pattern fluorescent characteristics |
| CN107392119A (en) * | 2017-07-04 | 2017-11-24 | 西北农林科技大学 | A kind of cattle farm monitor video method for real time filtering and system based on Spark Streaming |
| CN108171971A (en) * | 2017-12-18 | 2018-06-15 | 武汉烽火众智数字技术有限责任公司 | Vehicular real time monitoring method and system based on Spark Streaming |
| CN108154127A (en) * | 2017-12-27 | 2018-06-12 | 天津智芯视界科技有限公司 | A kind of vehicle identification method based on video and radar |
| CN108319907A (en) * | 2018-01-26 | 2018-07-24 | 腾讯科技（深圳）有限公司 | A kind of vehicle identification method, device and storage medium |
| CN110889867B (en) * | 2018-09-10 | 2022-11-04 | 浙江宇视科技有限公司 | Method and device for detecting damaged degree of car face |
| CN109948610B (en) * | 2019-03-14 | 2022-11-18 | 西南交通大学 | Deep learning-based method for classifying fine granularity of vehicles in video |
| CN112036421A (en) * | 2019-05-16 | 2020-12-04 | 搜狗（杭州）智能科技有限公司 | Image processing method and device and electronic equipment |
| CN110362713B (en) * | 2019-07-12 | 2023-06-06 | 四川长虹云数信息技术有限公司 | Video monitoring and early warning method and system based on Spark Streaming |
| CN110765861A (en) * | 2019-09-17 | 2020-02-07 | 中控智慧科技股份有限公司 | Unlicensed vehicle type identification method and device and terminal equipment |
| CN113505863B (en) * | 2021-09-08 | 2021-12-03 | 武大吉奥信息技术有限公司 | Image multistage classification method and system based on cascade mean vector comprehensive scoring |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102411710A (en) * | 2011-12-09 | 2012-04-11 | 东南大学 | Vehicle type recognition method based on vehicle face features |
| CN104820831A (en) * | 2015-05-13 | 2015-08-05 | 沈阳聚德视频技术有限公司 | Front vehicle face identification method based on AdaBoost license plate location |
| US9195908B2 (en) * | 2013-05-22 | 2015-11-24 | Xerox Corporation | Snow classifier context window reduction using class t-scores and mean differences |
| CN105206060A (en) * | 2015-10-14 | 2015-12-30 | 武汉烽火众智数字技术有限责任公司 | Vehicle type recognition device and method based on SIFT characteristics |
| CN105335702A (en) * | 2015-10-15 | 2016-02-17 | 浙江捷尚视觉科技股份有限公司 | Bayonet vehicle type recognition method based on statistical learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9460367B2 (en) * | 2014-11-26 | 2016-10-04 | Xerox Corporation | Method and system for automating an image rejection process |
-
2016
- 2016-06-02 CN CN201610383537.XA patent/CN106056086B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102411710A (en) * | 2011-12-09 | 2012-04-11 | 东南大学 | Vehicle type recognition method based on vehicle face features |
| US9195908B2 (en) * | 2013-05-22 | 2015-11-24 | Xerox Corporation | Snow classifier context window reduction using class t-scores and mean differences |
| CN104820831A (en) * | 2015-05-13 | 2015-08-05 | 沈阳聚德视频技术有限公司 | Front vehicle face identification method based on AdaBoost license plate location |
| CN105206060A (en) * | 2015-10-14 | 2015-12-30 | 武汉烽火众智数字技术有限责任公司 | Vehicle type recognition device and method based on SIFT characteristics |
| CN105335702A (en) * | 2015-10-15 | 2016-02-17 | 浙江捷尚视觉科技股份有限公司 | Bayonet vehicle type recognition method based on statistical learning |
Non-Patent Citations (1)
| Title |
|---|
| 基于分区决策树组合分类器的多目标分类方法;师永征;《电脑知识与技术》;20151231;论文第2节 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106056086A (en) | 2016-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106056086B (en) | Vehicle brand type identifier method based on Fast Learning frame | |
| CN109829403B (en) | Vehicle anti-collision early warning method and system based on deep learning | |
| US7724962B2 (en) | Context adaptive approach in vehicle detection under various visibility conditions | |
| CN100485710C (en) | Method for recognizing vehicle type by digital picture processing technology | |
| Artan et al. | Driver cell phone usage detection from HOV/HOT NIR images | |
| CN105335702B (en) | A kind of bayonet model recognizing method based on statistical learning | |
| CN103942546B (en) | Traffic marking identifying system and method are oriented in a kind of urban environment | |
| CN103258432A (en) | Traffic accident automatic identification processing method and system based on videos | |
| CN108520528B (en) | Mobile vehicle tracking method based on improved difference threshold and displacement matching model | |
| CN104143091A (en) | Single-sample face recognition method based on improved mLBP | |
| Ketcham et al. | Recognizing the Illegal Parking Patterns of Cars on the Road in Front of the Bus Stop Using the Support Vector Machine | |
| Mammeri et al. | North-American speed limit sign detection and recognition for smart cars | |
| Kamijo et al. | Pedestrian detection algorithm for on-board cameras of multi view angles | |
| Zaklouta et al. | Segmentation masks for real-time traffic sign recognition using weighted HOG-based trees | |
| Do et al. | Speed limit traffic sign detection and recognition based on support vector machines | |
| Pradeep et al. | An improved technique for night-time vehicle detection | |
| Hamdi et al. | Road signs classification by ANN for real-time implementation | |
| Ho et al. | Intelligent speed bump system with dynamic license plate recognition | |
| Emami et al. | Real time vehicle make and model recognition based on hierarchical classification | |
| KR20160067631A (en) | Method for recognizing vehicle plate | |
| CN111860113A (en) | Lane line detection method and system | |
| Artan et al. | Comparison of face detection and image classification for detecting front seat passengers in vehicles | |
| Tan et al. | Shape template based side-view car detection algorithm | |
| Syed et al. | Color edge enhancement based fuzzy segmentation of license plates | |
| CN107545216A (en) | A kind of vehicle identification method available for power-line patrolling |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201217 Address after: No.88 Zhujiang Road, Xuecheng District, Zaozhuang City, Shandong Province Patentee after: Shandong Weidi Information Technology Co., Ltd Address before: 215000 2-b803 Creative Industry Park, no.328 Xinghu street, Suzhou Industrial Park, Jiangsu Province Patentee before: JIANGSU HUATONG DATA CLOUD TECHNOLOGY Co.,Ltd. |