CN107274668A - A kind of congestion in road modeling method based on vehicle detection - Google Patents
A kind of congestion in road modeling method based on vehicle detection Download PDFInfo
- Publication number
- CN107274668A CN107274668A CN201610207495.4A CN201610207495A CN107274668A CN 107274668 A CN107274668 A CN 107274668A CN 201610207495 A CN201610207495 A CN 201610207495A CN 107274668 A CN107274668 A CN 107274668A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- congestion
- road
- detection
- rci
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0133—Traffic data processing for classifying traffic situation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- 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
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of congestion in road modeling method based on vehicle detection, by the analysis to data, judge that jam situation is：RCI ＞ 1, then road is very unimpeded；RCI ＞ 0.7, the coast is clear；RCI ＞ 0.3, slight congestion；RCI ＜ 0.3, severe congestion.The present invention uses machine learning algorithm by vehicle testing techniques, verification and measurement ratio is high, is filtered and is corrected after detection so that detection image is more stablized, it is suitable for doing follow-up tracking calculating, the congestion model proposed according to car speed and traffic density is capable of the jam situation of actual response road.
Description
Technical field
The present invention relates to a kind of modeling method, more particularly to a kind of congestion in road modeling side based on vehicle detection
Method.
Background technology
Intelligent transportation system is by advanced information technology, data communication transmission technology, Electronic transducer technology, control
Technology and computer technology processed etc. are effectively integrated the one kind for applying to whole ground traffic control system and setting up
In a wide range of, it is comprehensive play a role, in real time, accurately and efficiently composite communications transport is managed.
Machine learning algorithm is in area of pattern recognition extensive use, it is necessary first to carries out feature extraction, then makes
The feature of extraction is classified with grader, Haar-like features can describe the marginal information of image, car
Marginal information is enriched, and is adapted for use with Haar-like features and is described, and Adaboost graders are a kind of
Cascade of strong classifiers, is made up of multiple Weak Classifiers, can be good at recognizing target by repeatedly differentiating.
With expanding economy and the continuous expansion of city size, congestion in road is as being total to that each big city faces
Same problem, fast and accurately finds the traffic congestion in road, for formulating rationally effective traffic
Congestion dispersal strategy is significant, and every country is also studied traffic congestion model, current
Decision method is the relation by comparing some traffic parameter and a threshold value mostly, when parameter is in a certain
When individual threshold region, we just think that traffic now is in a certain state.
The content of the invention
The purpose of the present invention is that and provides a kind of road based on vehicle detection in order to solve the above problems and gather around
Stifled modeling method.
The present invention is achieved through the following technical solutions above-mentioned purpose：
The present invention comprises the following steps
A, vehicle detection；
B, vehicle discriminating；
C, vehicle correction；
D, vehicle tracking；
E, calculating car speed；
F, calculating traffic density；
G, proposition vehicle congestion model；
H, judge congestion in road situation.
Currently preferred, according to step a, the vehicle detection is obtained by machine learning method, is carried first
The Haar-like features in image are taken, and then are classified using Adaboost graders, are extracted in image
Vehicle.
It is currently preferred, according to step b, a part of flase drop in information of vehicles is found by vehicle testing techniques
Region, vehicle discriminating is carried out by vehicle size, marginal point filtering, by smoothing algorithm and rim detection,
The marginal information of detection zone is detected, if marginal information is very few, then it is assumed that be road surface, the detection zone is filtered out.
It is currently preferred, according to step c, vehicle edge is positioned by vehicle edge detection technique,
Carry out vehicle correction.
It is currently preferred, according to step d, by Kalman filtering algorithm to vehicle tracking.
It is currently preferred, according to step e, the speed of vehicle each car is calculated according to vehicle tracking result, always
The car speed of body is the average value of the speed of all cars.
It is currently preferred, according to step f, the traffic density VD refer to the vehicle number that is travelled on road with
The ratio of path area, the influence of vehicle volume is not considered, it is assumed that common n passage, and link length is L, altogether
M car is travelled, then traffic density is：
VD=m/ (n*L)；
It is currently preferred, according to step g, a kind of car is proposed according to the traffic density and the car speed
Congestion model, RCI represents congestion in road index, according to everyday common sense, and traffic density is smaller, car speed
Bigger, then road is more unimpeded, and the vehicle congestion model of foundation is：
RCI=VS/ (p*VD), wherein p are parameter；
Currently preferred, by the analysis to historical traffic data, jam situation is：RCI ＞ 1, then road
It is very unimpeded；RCI ＞ 0.7, the coast is clear；RCI ＞ 0.3, slight congestion；RCI ＜ 0.3, severe congestion.
The beneficial effects of the present invention are：
The present invention uses machine learning algorithm by vehicle testing techniques, and verification and measurement ratio is high, is carried out after detection
Filter and correction so that detection image is more stablized, are suitable for doing follow-up tracking calculating, according to car speed
The congestion model proposed with traffic density is capable of the jam situation of actual response road.
Brief description of the drawings
Fig. 1 is a kind of flow chart of congestion in road modeling method based on vehicle detection of the present invention.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings：
As shown in Figure 1：The present invention comprises the following steps
A, vehicle detection, the vehicle detection are obtained by machine learning method, are extracted first in image
Haar-like features, and then classified using Adaboost graders, extract the vehicle in image；
B, vehicle discriminating, find a part of flase drop region in information of vehicles by vehicle testing techniques, pass through car
Size, marginal point filtering carry out vehicle discriminating, pass through smoothing algorithm and rim detection, detect detection zone
Marginal information, if marginal information is very few, then it is assumed that be road surface, filter out the detection zone；
C, vehicle correction, are positioned to vehicle edge by vehicle edge detection technique, carry out vehicle correction；
D, vehicle tracking, by Kalman filtering algorithm to vehicle tracking；
E, calculating car speed, the speed of vehicle each car, overall vehicle are calculated according to vehicle tracking result
Speed is the average value of the speed of all cars；
F, calculating traffic density, the traffic density VD refer to the vehicle number travelled on road and path area
Ratio, the influence of vehicle volume is not considered, it is assumed that common n passage, link length is L, and concurrence sails m
Car, then traffic density be：
VD=m/ (n*L)；
G, proposition vehicle congestion model, propose that a kind of vehicle is gathered around according to the traffic density and the car speed
Stifled model, RCI represents congestion in road index, according to everyday common sense, and traffic density is smaller, and car speed is bigger,
Then road is more unimpeded, and the vehicle congestion model of foundation is：
RCI=VS/ (p*VD), wherein p are parameter；
H, judge congestion in road situation, by the analysis to historical traffic data, jam situation is：RCI ＞ 1,
Then road is very unimpeded；RCI ＞ 0.7, the coast is clear；RCI ＞ 0.3, slight congestion；RCI ＜ 0.3, severe is gathered around
It is stifled.
The operation principle of the present invention is as follows：
The vehicle in video is detected using auto model, in the training stage of auto model：Arrangement is substantial amounts of just
Sample vehicle image and negative sample non-vehicle image, extract the Haar-like features of sample, with reference to Adaboost
Grader is trained, and obtains two class discrimination models.In detection-phase, same feature and grader are used
And the model trained carries out the detection of vehicle, the vehicle subgraph in video is extracted.To the subgraph of extraction
As being filtered, the subgraph of extraction is entered according to priori, including vehicle size, vehicle edge information
Row filtering.Because the testing result of Adaboost graders and Haar-like features can cause the subgraph that detects
As there is deviation, vehicle edge is accurately positioned according to marginal information, the correction of testing result is carried out.Use karr
Graceful filtering algorithm is tracked to vehicle.According to tracking result, the speed of each car is calculated, it is assumed that shared m
Car, then average vehicle speed for all cars speed and divided by the obtained average values of m.Calculate traffic density.
Set up vehicle congestion model company：RCI=VS/ (p*VD), passes through the analysis to historical traffic data, p=1000
Proper, RCI is bigger, represents that road is more unimpeded, and RCI is smaller to represent that road gets over congestion.By to data
Analysis, judge that jam situation is：RCI ＞ 1, then road is very unimpeded；RCI ＞ 0.7, the coast is clear；RCI ＞ 0.3,
Slight congestion；RCI ＜ 0.3, severe congestion.
Those skilled in the art do not depart from the essence and spirit of the present invention, can have various deformation scheme to realize this
Invention, the foregoing is only preferable feasible embodiment of the invention, not thereby limits to the power of the present invention
Sharp scope, the equivalent structure change that all utilization description of the invention and accompanying drawing content are made, is both contained in this hair
Within bright interest field.
Claims (9)
1. a kind of congestion in road modeling method based on vehicle detection, it is characterised in that：Comprise the following steps
A, vehicle detection；
B, vehicle discriminating；
C, vehicle correction；
D, vehicle tracking；
E, calculating car speed；
F, calculating traffic density；
G, proposition vehicle congestion model；
H, judge congestion in road situation.
2. a kind of congestion in road modeling method based on vehicle detection according to claim 1, it is characterised in that：
According to step a, the vehicle detection is obtained by machine learning method, and the Haar-like in image is extracted first
Feature, and then classified using Adaboost graders, extract the vehicle in image.
3. a kind of congestion in road modeling method based on vehicle detection according to claim 1, it is characterised in that：
According to step b, a part of flase drop region in information of vehicles is found by vehicle testing techniques, by vehicle size,
Marginal point filtering carries out vehicle discriminating, passes through smoothing algorithm and rim detection, the edge letter of detection detection zone
Breath, if marginal information is very few, then it is assumed that be road surface, filters out the detection zone.
4. a kind of congestion in road modeling method based on vehicle detection according to claim 3, it is characterised in that：
According to step c, vehicle edge is positioned by vehicle edge detection technique, vehicle correction is carried out.
5. a kind of congestion in road modeling method based on vehicle detection according to claim 4, it is characterised in that：
According to step d, by Kalman filtering algorithm to vehicle tracking.
6. a kind of congestion in road modeling method based on vehicle detection according to claim 5, it is characterised in that：
According to step e, the speed of vehicle each car is calculated according to vehicle tracking result, overall car speed is all
The average value of the speed of car.
7. a kind of congestion in road modeling method based on vehicle detection according to claim 6, it is characterised in that：
According to step f, the traffic density VD refers to the ratio of the vehicle number travelled on road and path area, no
Consider the influence of vehicle volume, it is assumed that common n passage, link length is L, and concurrence sails m car, then car
Density is：
VD=m/ (n*L).
8. a kind of congestion in road modeling method based on vehicle detection according to claim 7, it is characterised in that：
According to step g, a kind of vehicle congestion model, RCI tables are proposed according to the traffic density and the car speed
Show congestion in road index, according to everyday common sense, traffic density is smaller, and car speed is bigger, then road is more smooth
Logical, the vehicle congestion model of foundation is：
RCI=VS/ (p*VD), wherein p are parameter.
9. a kind of congestion in road modeling method based on vehicle detection according to claim 8, it is characterised in that：
According to step h, the excessively analysis to historical traffic data, jam situation is：RCI ＞ 1, then road is very unimpeded；
RCI ＞ 0.7, the coast is clear；RCI ＞ 0.3, slight congestion；RCI ＜ 0.3, severe congestion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610207495.4A CN107274668A (en) | 2016-04-06 | 2016-04-06 | A kind of congestion in road modeling method based on vehicle detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610207495.4A CN107274668A (en) | 2016-04-06 | 2016-04-06 | A kind of congestion in road modeling method based on vehicle detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107274668A true CN107274668A (en) | 2017-10-20 |
Family
ID=60052239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610207495.4A Pending CN107274668A (en) | 2016-04-06 | 2016-04-06 | A kind of congestion in road modeling method based on vehicle detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107274668A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108805096A (en) * | 2018-06-19 | 2018-11-13 | 芜湖岭上信息科技有限公司 | A kind of melody control method and system based on road image |
| CN109147331A (en) * | 2018-10-11 | 2019-01-04 | 青岛大学 | A kind of congestion in road condition detection method based on computer vision |
| CN110335465A (en) * | 2019-07-10 | 2019-10-15 | 北京维联众诚科技有限公司 | Traffic jam detection method and system in monitor video based on AI deep learning |
| CN110570654A (en) * | 2019-09-16 | 2019-12-13 | 河南工业大学 | road section traffic jam dynamic detection method based on immunity |
| CN110688922A (en) * | 2019-09-18 | 2020-01-14 | 苏州奥易克斯汽车电子有限公司 | Deep learning-based traffic jam detection system and detection method |
| CN110782659A (en) * | 2019-09-09 | 2020-02-11 | 腾讯科技（深圳）有限公司 | Road condition determining method, road condition determining device, server and storage medium |
| CN113096397A (en) * | 2021-03-31 | 2021-07-09 | 武汉大学 | Traffic jam analysis method based on millimeter wave radar and video detection |
| CN113313950A (en) * | 2021-07-28 | 2021-08-27 | 长沙海信智能系统研究院有限公司 | Method, device and equipment for detecting vehicle congestion and computer storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11203594A (en) * | 1998-01-16 | 1999-07-30 | Toshiba Corp | Route guiding device |
| CN102521983A (en) * | 2011-12-23 | 2012-06-27 | 北京易华录信息技术股份有限公司 | Vehicle violation detection system based on high definition video technology and method thereof |
| CN103455820A (en) * | 2013-07-09 | 2013-12-18 | 河海大学 | Method and system for detecting and tracking vehicle based on machine vision technology |
| CN104091440A (en) * | 2014-07-02 | 2014-10-08 | 东南大学 | Urban traffic congestion cost estimation method based on road grades and vehicle types |
| CN104778834A (en) * | 2015-01-23 | 2015-07-15 | 哈尔滨工业大学 | Urban road traffic jam judging method based on vehicle GPS data |
-
2016
- 2016-04-06 CN CN201610207495.4A patent/CN107274668A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11203594A (en) * | 1998-01-16 | 1999-07-30 | Toshiba Corp | Route guiding device |
| CN102521983A (en) * | 2011-12-23 | 2012-06-27 | 北京易华录信息技术股份有限公司 | Vehicle violation detection system based on high definition video technology and method thereof |
| CN103455820A (en) * | 2013-07-09 | 2013-12-18 | 河海大学 | Method and system for detecting and tracking vehicle based on machine vision technology |
| CN104091440A (en) * | 2014-07-02 | 2014-10-08 | 东南大学 | Urban traffic congestion cost estimation method based on road grades and vehicle types |
| CN104778834A (en) * | 2015-01-23 | 2015-07-15 | 哈尔滨工业大学 | Urban road traffic jam judging method based on vehicle GPS data |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108805096A (en) * | 2018-06-19 | 2018-11-13 | 芜湖岭上信息科技有限公司 | A kind of melody control method and system based on road image |
| CN109147331A (en) * | 2018-10-11 | 2019-01-04 | 青岛大学 | A kind of congestion in road condition detection method based on computer vision |
| CN109147331B (en) * | 2018-10-11 | 2021-07-27 | 青岛大学 | Road congestion state detection method based on computer vision |
| CN110335465A (en) * | 2019-07-10 | 2019-10-15 | 北京维联众诚科技有限公司 | Traffic jam detection method and system in monitor video based on AI deep learning |
| CN110782659A (en) * | 2019-09-09 | 2020-02-11 | 腾讯科技（深圳）有限公司 | Road condition determining method, road condition determining device, server and storage medium |
| CN110570654A (en) * | 2019-09-16 | 2019-12-13 | 河南工业大学 | road section traffic jam dynamic detection method based on immunity |
| CN110688922A (en) * | 2019-09-18 | 2020-01-14 | 苏州奥易克斯汽车电子有限公司 | Deep learning-based traffic jam detection system and detection method |
| CN113096397A (en) * | 2021-03-31 | 2021-07-09 | 武汉大学 | Traffic jam analysis method based on millimeter wave radar and video detection |
| CN113096397B (en) * | 2021-03-31 | 2022-04-12 | 武汉大学 | Traffic jam analysis method based on millimeter wave radar and video detection |
| CN113313950A (en) * | 2021-07-28 | 2021-08-27 | 长沙海信智能系统研究院有限公司 | Method, device and equipment for detecting vehicle congestion and computer storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107274668A (en) | A kind of congestion in road modeling method based on vehicle detection | |
| CN108830188B (en) | Vehicle detection method based on deep learning | |
| CN108921083B (en) | Illegal mobile vendor identification method based on deep learning target detection | |
| CN108090429B (en) | Vehicle type recognition method for graded front face bayonet | |
| CN102568200B (en) | Method for judging vehicle driving states in real time | |
| CN102646199B (en) | Motorcycle type identifying method in complex scene | |
| CN102147866B (en) | Target identification method based on training Adaboost and support vector machine | |
| CN103198493B (en) | A kind ofly to merge and the method for tracking target of on-line study based on multiple features self-adaptation | |
| CN107492251A (en) | It is a kind of to be identified and driving condition supervision method based on the driver identity of machine learning and deep learning | |
| Møgelmose et al. | Traffic sign detection for us roads: Remaining challenges and a case for tracking | |
| CN107133974A (en) | The vehicle type classification method that Gaussian Background modeling is combined with Recognition with Recurrent Neural Network | |
| CN106652445A (en) | Road traffic accident judging method and device | |
| CN103366190A (en) | Method for identifying traffic sign | |
| CN105320966A (en) | Vehicle driving state recognition method and apparatus | |
| CN104318258A (en) | Time domain fuzzy and kalman filter-based lane detection method | |
| CN105488453A (en) | Detection identification method of no-seat-belt-fastening behavior of driver based on image processing | |
| CN102254428B (en) | Traffic jam detection method based on video processing | |
| CN103810503A (en) | Depth study based method for detecting salient regions in natural image | |
| CN110226170A (en) | A kind of traffic sign recognition method in rain and snow weather | |
| CN101329734A (en) | License plate character recognition method based on K-L transform and LS-SVM | |
| CN109050535A (en) | A kind of quick landform industry and mining city method based on vehicle attitude | |
| CN103886589A (en) | Goal-oriented automatic high-precision edge extraction method | |
| CN104809433A (en) | Zebra stripe detection method based on maximum stable region and random sampling | |
| CN104951764A (en) | Identification method for behaviors of high-speed vehicle based on secondary spectrum clustering and HMM (Hidden Markov Model)-RF (Random Forest) hybrid model | |
| CN106919902A (en) | A kind of vehicle identification and trajectory track method based on CNN |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171020 |
|
| WD01 | Invention patent application deemed withdrawn after publication |