CN104680787A - Method for detecting congestion condition of roads - Google Patents
Method for detecting congestion condition of roads Download PDFInfo
- Publication number
- CN104680787A CN104680787A CN201510058449.8A CN201510058449A CN104680787A CN 104680787 A CN104680787 A CN 104680787A CN 201510058449 A CN201510058449 A CN 201510058449A CN 104680787 A CN104680787 A CN 104680787A
- Authority
- CN
- China
- Prior art keywords
- pixel
- road
- gauss model
- background
- prospect
- 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.)
- Granted
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/758—Involving statistics of pixels or of feature values, e.g. histogram matching
Abstract
The invention relates to a method for detecting the congestion condition of roads. The method comprises the following steps: (1) processing an obtained video image of a current road to obtain a background and a foreground of the current road, and specifically, modeling for each pixel of a current frame by adopting a Gaussian mixture model; sorting vehicles and road surfaces; calculating communicating regions of the foreground, judging whether the size of each communication region is less than a preset value or not, if so, classifying the corresponding pixels into the background; for single pixel, judging whether the single pixel undergoes lighting changes or foreground changes according to change extraction features of pixels in an adjacent region; judging whether the foreground is in a non-vehicle appearing region or not, if so, enabling the corresponding pixels to be included in the background; re-training the Gaussian mixture model by adopting the pixels judged to be included in the background, and updating a classifier; (2) calculating the road occupancy and vehicle average speed on the current road according to the obtained foreground pixels. Compared with the prior art, the method for detecting congestion condition of roads has the advantages of high speed, strong robustness and the like.
Description
Technical field
The present invention relates to Vehicle Detection field, especially relate to a kind of congestion in road detection method.
Background technology
In city, along with motor vehicle has increasing rapidly of quantity, the jam situation of road is day by day serious, and especially in peak times such as upper and lower classes, jam is more general.At present, the traffic behavior parameter required for the method for discrimination of road traffic congestion is mainly obtained by various Vehicle Detection technology, and by analyzing these traffic behavior parameters, process, and then judge the jam situation of road.
Existing traffic jam detection method mainly comprises:
1) Coil Detector method: vehicle is by being embedded in the lead loop under road surface, disturb the earth magnetism magnetic flux line by coil and produce a voltage in coil, this voltage is amplified by high-gain amplifier, makes the relay work of detecting device, thus realizes testing goal.But Coil Detector method does not have directivity, surveyed area indefinite, repair or install and need suspend traffic, affect pavement life, fragile.
2) answer Coil Detector method: the same with the working method of single inductive coil detecting device, but speed parameter more accurately can be provided.Answer Coil Detector method need suspend traffic when repairing or install, and easily affect pavement life, fragile, and also more expensive than single inductive coil.
3) video detecting method: the image of video camera picked-up check point near zone, is processed image by computer program, identify thus detect vehicle.Video detecting method costly and accuracy of detection affect by weather, surveyed area surrounding brightness.
4) microwave detection method: by launched microwave signal, vehicle reflected radar microwave (Doppler effect), returns antenna and makes the detection actuating of relay thus reach testing goal.Microwave detection method can not detect vehicle that is static or low speed driving, is easily affected by the external environment, and is easily subject to the impact of people or thing when the thresholding of transmitted wave is selected incorrect, causes and knows by mistake.
In the above-mentioned methods, the advantage such as video detection technology does not damage road surface owing to having installation, can not cause the brief interruption of traffic, the information of acquisition is many, in the time in recent years, obtains development at full speed and general application.As Chinese patent CN102867415A discloses a kind of congestion in road method of discrimination based on video detection technology, the video image of the present road obtained is processed, obtains background image and the foreground target of present road; According to described foreground target, calculate average lane occupation rate and the time occupancy of described present road; According to described average lane occupation rate and time occupancy, judge the jam situation of described present road according to the road traffic congestion criterion set up.But prior art cannot process the video of low resolution/complex scene/weather condition.
Summary of the invention
Object of the present invention be exactly in order to overcome above-mentioned prior art exist defect and the congestion in road detection method that a kind of speed is fast, robust is strong is provided.
Object of the present invention can be achieved through the following technical solutions:
A kind of congestion in road detection method, comprises the following steps:
1) video image of the present road obtained is processed, obtains background and the prospect of present road, be specially:
A) to the present frame of video image, adopt each pixel of mixed Gauss model to present frame to carry out modeling, judge whether each pixel belongs to prospect, if so, then perform step b), if not, then perform step f);
B) extract texture features and the color characteristics of image, adopt the support vector machine classifier of RBF Kernel to classify to vehicle and road surface;
C) calculate the connected region of prospect, judge whether the size of each connected region is less than setting value, if so, then the pixel of correspondence is included into background, if not, then performs steps d);
D) to single pixel, extract feature according to the change of pixel in its neighborhood and judge illumination variation or prospect change;
E) judge whether prospect is in non-vehicle region, if so, then the pixel of correspondence be included into background, perform step f), if not, directly perform step f);
F) adopt the pixel re-training mixed Gauss model judging into background, and upgrade the support vector machine classifier of RBF Kernel;
2) roadway occupancy and the vehicle average velocity of present road is calculated according to the foreground pixel obtained.
In described mixed Gauss model, Gauss model number is 5.
Described step a) in, for each pixel, judge K the Gauss model belonged in mixed Gauss model according to its color value, judge whether this Gauss model is that in mixed Gauss model, energy is the strongest, if so, then respective pixel is included into background, then respective pixel is included into prospect if not.
Described step f) in, adopt the update algorithm in Online-SVM algorithm to upgrade the support vector machine classifier of RBF Kernel.
Described step f) in, the frequency of re-training mixed Gauss model be 15 minutes once.
Compared with prior art, the present invention has the following advantages:
1, reliability is high, and the present invention adopts each pixel of mixed Gauss model to present frame to carry out modeling, effectively rejects the impact of illumination variation when judging, thus obtains higher Detection accuracy;
2, RBF Kernel of the present invention support vector machine classifier and training mixed Gauss model be change upgrade, real-time is good;
3, speed is fast, meets the demand of real-time;
4, strong robustness, can use under different scene, can process complex scene change.
Accompanying drawing explanation
Fig. 1 is schematic flow sheet of the present invention.
Embodiment
Below in conjunction with the drawings and specific embodiments, the present invention is described in detail.The present embodiment is implemented premised on technical solution of the present invention, give detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.
As shown in Figure 1, the present embodiment provides a kind of congestion in road detection method, and obtain roadway occupancy and the vehicle average velocity of current time according to traffic video stream, idiographic flow is as follows:
Step S1, processes the video image of the present road obtained, obtains background and the prospect of present road, be specially:
Step S101, to the present frame of video image, adopts each pixel of mixed Gauss model to present frame to carry out modeling, judges whether each pixel belongs to prospect, if so, then perform step b), if not, then perform step f).
Time initial, average in mixed Gauss model is set to 0, variance is set to definite value, and Gauss model number is set as 5.
For each pixel, judge K the Gauss model belonged in mixed Gauss model according to its color value, judge whether this Gauss model is that in mixed Gauss model, energy is the strongest, if, then respective pixel is included into background, then respective pixel is included into prospect if not.
Suppose it is within a period of time, background (road surface) occurrence number is much larger than prospect (vehicle), and simultaneously background (road surface) color is more fixing, therefore can the maximum Gauss model of forming energy; And prospect (vehicle) etc. are owing to belonging to Different Individual, color is inconsistent, is difficult to form stable Gauss model.Now single pixel judges into the accuracy of background more than 99%.But the pixel judging into prospect also has large probability not belong to prospect (vehicle), possible because have: 1) illumination variation; 2) road debris moves; 3) raindrop under special climate moves, and therefore carries out aftertreatment.
Step S102, extracts texture features and the color characteristics of image, adopts the support vector machine classifier (support vector machine based on RBF kernel function) of RBF Kernel to classify to vehicle and road surface.Sorter uses LBP and Color two kinds of features, and main cause is: 1) LBP describes texture features, and surface of vehicle is comparatively level and smooth, and the special frame of lower frequency region is comparatively strong, and there is more noise on road surface, and the special frame of high frequency is stronger; 2) Color describes color characteristics, and for specific crossing, road surface color is fixing and vehicle color gap is large.
Step S103, calculates the connected region of prospect, judges whether the size of each connected region is less than setting value, if so, then the pixel of correspondence is included into background, if not, then performs steps d).The size of vehicle is fixing.The foreign material such as raindrop, the scraps of paper can be removed after such process.
Step S104, to single pixel, extracts feature according to the change of pixel in its neighborhood and judges illumination variation or prospect change.The characteristic of illumination variation is the change of the overall situation, and change evenly; And prospect change, change greatly in neighborhood, process can eliminate the impact of illumination variation like this.
Step S105, the track that vehicle occurs is fixing, judges whether prospect is in non-vehicle region, if so, then the pixel of correspondence is included into background, performs step S106, if not, directly performs step S106.
Step S106, adopt the pixel re-training mixed Gauss model judging into background, and upgrade the support vector machine classifier of RBF Kernel, adopt the update algorithm (on-line learning algorithm) in Online-SVM algorithm to upgrade the support vector machine classifier of RBF Kernel, the frequency of re-training mixed Gauss model be 15 minutes once.
Step S2, calculates roadway occupancy and the vehicle average velocity of present road according to the foreground pixel obtained.
Claims (5)
1. a congestion in road detection method, is characterized in that, comprises the following steps:
1) video image of the present road obtained is processed, obtains background and the prospect of present road, be specially:
A) to the present frame of video image, adopt each pixel of mixed Gauss model to present frame to carry out modeling, judge whether each pixel belongs to prospect, if so, then perform step b), if not, then perform step f);
B) extract texture features and the color characteristics of image, adopt the support vector machine classifier of RBF Kernel to classify to vehicle and road surface;
C) calculate the connected region of prospect, judge whether the size of each connected region is less than setting value, if so, then the pixel of correspondence is included into background, if not, then performs steps d);
D) to single pixel, extract feature according to the change of pixel in its neighborhood and judge illumination variation or prospect change;
E) judge whether prospect is in non-vehicle region, if so, then the pixel of correspondence be included into background, perform step f), if not, directly perform step f);
F) adopt the pixel re-training mixed Gauss model judging into background, and upgrade the support vector machine classifier of RBF Kernel;
2) roadway occupancy and the vehicle average velocity of present road is calculated according to the foreground pixel obtained.
2. congestion in road detection method according to claim 1, is characterized in that, in described mixed Gauss model, Gauss model number is 5.
3. congestion in road detection method according to claim 1, it is characterized in that, described step a) in, for each pixel, judge K the Gauss model belonged in mixed Gauss model according to its color value, judge whether this Gauss model is that in mixed Gauss model, energy is the strongest, if, then respective pixel is included into background, then respective pixel is included into prospect if not.
4. congestion in road detection method according to claim 1, is characterized in that, described step f) in, adopt the update algorithm in Online-SVM algorithm to upgrade the support vector machine classifier of RBF Kernel.
5. congestion in road detection method according to claim 1, is characterized in that, described step f) in, the frequency of re-training mixed Gauss model be 15 minutes once.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510058449.8A CN104680787B (en) | 2015-02-04 | 2015-02-04 | A kind of congestion in road detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510058449.8A CN104680787B (en) | 2015-02-04 | 2015-02-04 | A kind of congestion in road detection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104680787A true CN104680787A (en) | 2015-06-03 |
CN104680787B CN104680787B (en) | 2017-08-01 |
Family
ID=53315769
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510058449.8A Active CN104680787B (en) | 2015-02-04 | 2015-02-04 | A kind of congestion in road detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104680787B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104933868A (en) * | 2015-06-17 | 2015-09-23 | 苏州大学 | Real-time online traffic state detection method based on traffic monitoring video |
WO2018068311A1 (en) * | 2016-10-14 | 2018-04-19 | 富士通株式会社 | Background model extraction device, and method and device for detecting traffic congestion |
WO2021169174A1 (en) * | 2020-02-29 | 2021-09-02 | 深圳壹账通智能科技有限公司 | Road congestion degree prediction method, apparatus, computer device, and readable storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101807345A (en) * | 2010-03-26 | 2010-08-18 | 重庆大学 | Traffic jam judging method based on video detection technology |
US20100208987A1 (en) * | 2009-02-16 | 2010-08-19 | Institute For Information Industry | Method and system for foreground detection using multi-modality fusion graph cut |
CN202563526U (en) * | 2012-03-22 | 2012-11-28 | 北京尚易德科技有限公司 | Transportation vehicle detection and recognition system based on video |
TW201447819A (en) * | 2013-06-14 | 2014-12-16 | Univ Asia | Moving foreground detection system and method thereof |
-
2015
- 2015-02-04 CN CN201510058449.8A patent/CN104680787B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100208987A1 (en) * | 2009-02-16 | 2010-08-19 | Institute For Information Industry | Method and system for foreground detection using multi-modality fusion graph cut |
CN101807345A (en) * | 2010-03-26 | 2010-08-18 | 重庆大学 | Traffic jam judging method based on video detection technology |
CN202563526U (en) * | 2012-03-22 | 2012-11-28 | 北京尚易德科技有限公司 | Transportation vehicle detection and recognition system based on video |
TW201447819A (en) * | 2013-06-14 | 2014-12-16 | Univ Asia | Moving foreground detection system and method thereof |
Non-Patent Citations (1)
Title |
---|
李世林: "《基于视频的道路交通参数提取方法研究》", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104933868A (en) * | 2015-06-17 | 2015-09-23 | 苏州大学 | Real-time online traffic state detection method based on traffic monitoring video |
WO2018068311A1 (en) * | 2016-10-14 | 2018-04-19 | 富士通株式会社 | Background model extraction device, and method and device for detecting traffic congestion |
WO2021169174A1 (en) * | 2020-02-29 | 2021-09-02 | 深圳壹账通智能科技有限公司 | Road congestion degree prediction method, apparatus, computer device, and readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN104680787B (en) | 2017-08-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Nienaber et al. | Detecting potholes using simple image processing techniques and real-world footage | |
CN103117005B (en) | Lane deviation warning method and system | |
CN102819952B (en) | Method for detecting illegal lane change of vehicle based on video detection technique | |
CN102768804B (en) | Video-based traffic information acquisition method | |
TWI409718B (en) | Method of locating license plate of moving vehicle | |
CN100545867C (en) | Aerial shooting traffic video frequency vehicle rapid checking method | |
CN102005120A (en) | Traffic intersection monitoring technology and system based on video image analysis | |
Bello-Salau et al. | Image processing techniques for automated road defect detection: A survey | |
WO2008020544A1 (en) | Vehicle detection device, vehicle detection method, and vehicle detection program | |
CN105336169A (en) | Method and system for judging traffic jams based on videos | |
Munawar | Image and video processing for defect detection in key infrastructure | |
CN102118561A (en) | Camera movement detection system in monitoring system and method | |
CN105095869A (en) | Pedestrian detection method and apparatus | |
CN103413439B (en) | A kind of passenger vehicle based on video and lorry sorting technique | |
CN105069441A (en) | Moving vehicle detection method based on background updating and particle swarm optimization algorithm | |
CN112036385B (en) | Library position correction method and device, electronic equipment and readable storage medium | |
CN104680787A (en) | Method for detecting congestion condition of roads | |
CN102867415A (en) | Video detection technology-based road jam judgement method | |
CN102542805A (en) | Device for judging traffic jam based on videos | |
CN110660211B (en) | Parking area map improvement using occupancy behavior anomaly detector | |
CN101819681A (en) | Weight number adaptively adjusted weighted average background updating method | |
CN111524121A (en) | Road and bridge fault automatic detection method based on machine vision technology | |
CN103295003A (en) | Vehicle detection method based on multi-feature fusion | |
CN203020184U (en) | Electric vehicle lane marking line detecting device based on double charge coupled devices (CCD) | |
Iwasaki et al. | Real-time robust vehicle detection through the same algorithm both day and night |
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 |