CN104318761A - Highway-scene-based detection and vehicle detection tracking optimization method - Google Patents

Highway-scene-based detection and vehicle detection tracking optimization method Download PDF

Info

Publication number
CN104318761A
CN104318761A CN201410436667.6A CN201410436667A CN104318761A CN 104318761 A CN104318761 A CN 104318761A CN 201410436667 A CN201410436667 A CN 201410436667A CN 104318761 A CN104318761 A CN 104318761A
Authority
CN
China
Prior art keywords
vehicle
video
highway
lane line
following
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
Application number
CN201410436667.6A
Other languages
Chinese (zh)
Other versions
CN104318761B (en
Inventor
陈键锋
黄仙海
周智恒
庄衍竖
靳晓雨
李腾辉
李学攀
马衡禹
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN201410436667.6A priority Critical patent/CN104318761B/en
Publication of CN104318761A publication Critical patent/CN104318761A/en
Application granted granted Critical
Publication of CN104318761B publication Critical patent/CN104318761B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a highway-scene-based detection and vehicle detection tracking optimization method. The method comprises the following steps: step one, detecting a lane line in a video and marking and recording the lane line information; step two, detecting vehicles in the marked lane line and marking and recording vehicle information; step three, detecting a distance between vehicles in the video and marking and recording the vehicle distance; step four, determining whether the vehicle distance is larger than a given threshold value; if so, entering the step five; and if not, returning to the step one; step five, recording a vehicle distance Di within time t, marking the maximum value and the minimum value, and determining whether the maximum and minimum value difference is less than a given value; if so, entering the step six; if not, returning to the step one; step six, determining scene switching to the highway; and step seven, tracking the marked vehicle. The method has the following advantages: the number of processed frames of the video can be reduced; the calculated amount is reduced; and the real-time performance is improved.

Description

Based on the method that detection and the automobile detecting following of highway scene are optimized
Technical field
The present invention relates to a kind of mechanical vision inspection technology, a kind of method of particularly detection based on highway scene and automobile detecting following optimization.
Background technology
Along with the development of society, the living standard of people's life improves constantly, and automobile is gradually gained popularity.But the thing followed is traffic hazard, and especially pernicious traffic hazard trend constantly rises.Especially on a highway, due to features such as its traffic volume are large, the speed of a motor vehicle is fast, once there is traffic hazard, its huge vehicle flowrate very easily forms tens the even chain traffic hazard of up to a hundred car collisions, the loss caused is also just inestimable, is all often that loss is serious, car crash.For reducing the incidence of traffic hazard, researchist has proposed the concept of intelligent vehicle.
In order to improve the security of automobile and intelligent, require that vehicle can identify the major obstacle thing (as vehicle) on road exactly, follow the tracks of, point out driver simultaneously, allow driver can grasp the situation of road in time, danger potential in driving procedure is made and judges accurately.
But highway and the road conditions of urban highway have obviously different.Usually, on highway, spacing is comparatively large, and road conditions are simpler, the change of the relative position such as objects in front (as vehicle, lane line) is little at short notice, and general spacing is less on urban highway, road conditions are complicated, and the rate of change of objects in front relative position is large.When changing little for Freeway Conditions, change between the successive frame that front end camera collection is returned is very little, under traditional detection algorithm, namely all processes each frame gathered, obviously like this, increase the weight of the unnecessary burden of processor, reduce real-time.And directly unnecessary frame is lost, under the complex road condition in urban district, obviously can not ensure its accuracy of detection.Therefore, in order to improve real-time and the high precision of guarantee detection, we need to apply different detection algorithms at highway and urban highway.But the prerequisite of realizing this goal accurately to judge that Current vehicle travels at highway or urban highway.So a kind of method proposing detection based on highway scene and automobile detecting following optimization is significant.
Summary of the invention
The object of the invention is to overcome the shortcoming of prior art and deficiency, a kind of method that detection based on highway scene and automobile detecting following are optimized is provided, the method accurately can judge that Current vehicle travels at highway or urban highway, improves real-time and the high precision of guarantee detection.
Object of the present invention is achieved through the following technical solutions: a kind of method that detection based on highway scene and automobile detecting following are optimized, and comprises the following steps:
1) detect the lane line in video, mark and record lane line information;
2) detect institute and mark vehicle in lane line, mark also recording information of vehicles;
3) detect the distance of vehicle in video, mark and record spacing D i;
4) judge whether the distance of vehicle is greater than given threshold values D thif had, then enter step 5), if do not had, then return step 1);
5) the spacing D in time t is recorded in i, mark its maximal value D maxwith minimum value D min, and judge most value difference Δ=D max-D minwhether be less than set-point Δ thif had, then enter step 6), if do not had, then return step 1);
6) judge that scene switches to highway;
7) marked vehicle is followed the tracks of.
Preferably, described step 1) in, the detection of lane line is based on randomized hough transform, finds two lane lines corresponding to Current vehicle traveling lane according to the constraint of lane line angle, and the lane line kept left is denoted as Line_L, the lane line of keeping right is denoted as Line_R; Video is by the monocular cam captured in real-time be arranged on vehicle, and camera is arranged on the position in the middle of vehicle front.
Further, described step 3) in, draw vehicle distances by data in frame of video, the process obtaining vehicle distances is as follows:
(1-1) method of rim detection positioning licence plate from video i-th frame is adopted;
(1-2) the length L of car plate is detected 1with vehicle distances L in video image 2;
(1-3) existing automotive license plate regular length is 440mm, by formula D i=L 2 × 440/l 1mm extrapolates vehicle distances D i.
Further, described step 7) in, be by following the tracks of near the direct position frame of video of vehicle being detected when detecting vehicle to the tracking of institute's marked vehicle, and the frame of video handled by following the tracks of is sampled with sampling rate fs from continuous print frame of video.
The present invention has following advantage and effect relative to prior art:
1, compared with prior art, the invention provides a kind of method based on highway scene detection, fill up the vacancy of prior art in highway scene detection.
2, the present invention utilizes spacing to realize the differentiation of highway and non-freeway two kinds of patterns.
3, the present invention is on the basis detecting highway scene, introduces the optimization method of automobile detecting following.When judging that scene has switched to highway, directly follow the tracks of near the position of frame of video vehicle being detected, and the frame of video handled by following the tracks of is sampled with sampling rate fs from continuous print frame of video, which reduce the process frame number of video, decrease calculated amount, improve real-time.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of method of the present invention.
Embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited thereto.
Embodiment
As shown in Figure 1, a kind of method that detection based on highway scene and automobile detecting following are optimized, comprises the following steps:
Step 1, by being arranged on the monocular cam captured in real-time in the middle of vehicle front, get continuous print frame of video, pass through randomized hough transform, two lane lines corresponding to Current vehicle traveling lane are found according to the constraint of lane line angle, and the lane line kept left is denoted as Line_L, the lane line of keeping right is denoted as Line_R;
Step 2, detect institute and mark vehicle in lane line, mark is recording information of vehicles also, and method positioning licence plate from video i-th frame of employing rim detection, extrapolates vehicle distances D by the area S of license plate image i, mark and record spacing D i; Draw vehicle distances by data in frame of video, the process obtaining vehicle distances is as follows:
(1-1) method of rim detection positioning licence plate from video i-th frame is adopted;
(1-2) the length L of car plate is detected 1with vehicle distances L in video image 2;
(1-3) existing automotive license plate regular length is 440mm, by formula D i=0.44L 2/l 1m extrapolates vehicle distances D i.
Step 3, judge the distance D of vehicle iwhether be greater than D th=60m, if do not had, then enters step 2, if had, is recorded in the spacing D in time t=30s i, mark its maximal value D maxwith minimum value D min, and judge most value difference Δ=D max-D minwhether be less than Δ th=20m, if do not had, then returns step 2; If had, judge that scene switches to highway; Marked vehicle is followed the tracks of.
Sample with sampling rate fs from continuous print frame of video and process, directly following the tracks of near the position of frame of video vehicle being detected.
Above-described embodiment is the present invention's preferably embodiment; but embodiments of the present invention are not restricted to the described embodiments; change, the modification done under other any does not deviate from Spirit Essence of the present invention and principle, substitute, combine, simplify; all should be the substitute mode of equivalence, be included within protection scope of the present invention.

Claims (5)

1., based on the detection method of highway scene and an optimization for automobile detecting following, it is characterized in that, comprise the following steps:
Step 1, the lane line detected in video, mark and record lane line information;
Step 2, detect institute and mark vehicle in lane line, mark also recording information of vehicles;
In step 3, detection video, the distance of vehicle, marks and records spacing D i;
Step 4, judge whether the distance of vehicle is greater than given threshold values D th, if so, then enter step 5, otherwise, return step 1;
Step 5, the spacing D be recorded in time t i, mark its maximal value D maxwith minimum value D min, and judge most value difference Δ=D max-D minwhether be less than set-point Δ th, if so, then enter step 6, otherwise, return step 1;
Step 6, judgement scene switch to highway;
Step 7, marked vehicle to be followed the tracks of.
2. according to claim 1 based on the detection method of highway scene and the optimization of automobile detecting following, it is characterized in that, in step 1, the detection of described lane line is based on randomized hough transform, two lane lines corresponding to Current vehicle traveling lane are found according to the constraint of lane line angle, and the lane line kept left is denoted as Line_L, the lane line of keeping right is denoted as Line_R.
3. according to claim 1 based on the detection method of highway scene and the optimization of automobile detecting following, it is characterized in that, in step 1, described video is by the monocular cam captured in real-time be arranged on vehicle, and camera is arranged on the position in the middle of vehicle front.
4. according to claim 1ly to it is characterized in that based on the detection method of highway scene and the optimization of automobile detecting following, in step 3, protect vehicle distance be drawn by data in frame of video, the distance obtaining described vehicle comprises the following steps:
(1-1) method of rim detection positioning licence plate from video i-th frame is adopted;
(1-2) the length L of car plate is detected 1with vehicle distances L in video image 2;
(1-3) existing automotive license plate regular length is 440mm, by formula D i=L 2 × 440/l 1mm extrapolates vehicle distances D i.
5. according to claim 1 based on the detection method of highway scene and the optimization of automobile detecting following, it is characterized in that, in step 7, the described tracking to institute's marked vehicle is by following the tracks of near the direct position frame of video of vehicle being detected when detecting vehicle, and the frame of video handled by following the tracks of gets with sampling rate fs sampling from continuous print frame of video.
CN201410436667.6A 2014-08-29 2014-08-29 Highway-scene-based detection and vehicle detection tracking optimization method Active CN104318761B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410436667.6A CN104318761B (en) 2014-08-29 2014-08-29 Highway-scene-based detection and vehicle detection tracking optimization method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410436667.6A CN104318761B (en) 2014-08-29 2014-08-29 Highway-scene-based detection and vehicle detection tracking optimization method

Publications (2)

Publication Number Publication Date
CN104318761A true CN104318761A (en) 2015-01-28
CN104318761B CN104318761B (en) 2017-02-08

Family

ID=52373985

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410436667.6A Active CN104318761B (en) 2014-08-29 2014-08-29 Highway-scene-based detection and vehicle detection tracking optimization method

Country Status (1)

Country Link
CN (1) CN104318761B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107792068A (en) * 2016-09-06 2018-03-13 德尔福技术有限公司 Automated vehicle lane changing control system
CN107832788A (en) * 2017-11-01 2018-03-23 青岛理工大学 A kind of vehicle distance measurement method based on monocular vision and Car license recognition
CN109003338A (en) * 2018-06-22 2018-12-14 南京慧尔视智能科技有限公司 A kind of Roadside Parking self-clocking charging method and device
CN113269705A (en) * 2020-02-14 2021-08-17 中国石油天然气集团有限公司 Video-based explosive feeding depth detection method and device

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003346282A (en) * 2002-05-29 2003-12-05 Toshiba Corp System and method for monitoring road
CN1928945A (en) * 2006-09-20 2007-03-14 烟台麦特电子有限公司 Traffic road condition gathering promulgation method based on GPRS
US20100042315A1 (en) * 2006-10-18 2010-02-18 Tomoya Ikeuchi Map information processing apparatus
CN102289948A (en) * 2011-09-02 2011-12-21 浙江大学 Multi-characteristic fusion multi-vehicle video tracking method under highway scene
CN102542835A (en) * 2010-11-10 2012-07-04 罗伯特·博世有限公司 Navigation system and navigation method with traffic jam recognition function
CN203134149U (en) * 2012-12-11 2013-08-14 武汉高德红外股份有限公司 Vehicle auxiliary driving system based on different wave band imaging fusion image processing

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003346282A (en) * 2002-05-29 2003-12-05 Toshiba Corp System and method for monitoring road
CN1928945A (en) * 2006-09-20 2007-03-14 烟台麦特电子有限公司 Traffic road condition gathering promulgation method based on GPRS
US20100042315A1 (en) * 2006-10-18 2010-02-18 Tomoya Ikeuchi Map information processing apparatus
CN102542835A (en) * 2010-11-10 2012-07-04 罗伯特·博世有限公司 Navigation system and navigation method with traffic jam recognition function
CN102289948A (en) * 2011-09-02 2011-12-21 浙江大学 Multi-characteristic fusion multi-vehicle video tracking method under highway scene
CN203134149U (en) * 2012-12-11 2013-08-14 武汉高德红外股份有限公司 Vehicle auxiliary driving system based on different wave band imaging fusion image processing

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
佘永业 等: "一种交通拥堵自动检测方法", 《计算机与现代化》 *
王创新 等: "一种高速密集视频监控场景背景重构方法", 《数据采集与处理》 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107792068A (en) * 2016-09-06 2018-03-13 德尔福技术有限公司 Automated vehicle lane changing control system
CN107792068B (en) * 2016-09-06 2020-04-21 安波福技术有限公司 Automatic vehicle lane change control system
CN107832788A (en) * 2017-11-01 2018-03-23 青岛理工大学 A kind of vehicle distance measurement method based on monocular vision and Car license recognition
CN107832788B (en) * 2017-11-01 2021-07-23 青岛中汽特种汽车有限公司 Vehicle distance measuring method based on monocular vision and license plate recognition
CN109003338A (en) * 2018-06-22 2018-12-14 南京慧尔视智能科技有限公司 A kind of Roadside Parking self-clocking charging method and device
CN113269705A (en) * 2020-02-14 2021-08-17 中国石油天然气集团有限公司 Video-based explosive feeding depth detection method and device

Also Published As

Publication number Publication date
CN104318761B (en) 2017-02-08

Similar Documents

Publication Publication Date Title
US11084494B2 (en) Method for detecting safety of driving behavior, apparatus, device and storage medium
CN108437986B (en) Vehicle driving assistance system and assistance method
CN104008645B (en) One is applicable to the prediction of urban road lane line and method for early warning
CN102632839B (en) Back sight image cognition based on-vehicle blind area early warning system and method
CN103942960B (en) A kind of vehicle lane change detection method and device
CN110400478A (en) A kind of road condition notification method and device
CN104554259B (en) Active automatic Pilot accessory system and method
CN101135558B (en) Vehicle anti-collision early warning method and apparatus based on machine vision
CN102332209B (en) Automobile violation video monitoring method
CN102956118B (en) Device and method for traffic sign recognition
CN110517521A (en) A kind of lane departure warning method based on road car fusion perception
CN105216792A (en) Obstacle target in surrounding environment is carried out to the method and apparatus of recognition and tracking
CN101984478B (en) Abnormal S-type driving warning method based on binocular vision lane marking detection
WO2018059586A1 (en) A vehicle identification method, device, and vehicle
CN105620489A (en) Driving assistance system and real-time warning and prompting method for vehicle
CN103832282A (en) Automobile intelligent speed limiting system and automobile speed limiting method
CN202071799U (en) Intelligent automobile safety warning device
CN101131321A (en) Real-time safe interval measurement method and device used for vehicle anti-collision warning
CN105825185A (en) Early warning method and device against collision of vehicles
CN103287372A (en) Automobile anti-collision safe protecting method based on image processing
CN104494598A (en) Road-crossing driving control method for intelligent vehicles
CN102778223A (en) License number cooperation target and monocular camera based automobile anti-collision early warning method
CN106541946A (en) Speed controller
CN110126730A (en) Vehicle lane change based reminding method and system
CN104978856A (en) Automobile intelligent steering automatic prompting device and method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant