CN103902985B - High-robustness real-time lane detection algorithm based on ROI - Google Patents
High-robustness real-time lane detection algorithm based on ROI Download PDFInfo
- Publication number
- CN103902985B CN103902985B CN201410148832.8A CN201410148832A CN103902985B CN 103902985 B CN103902985 B CN 103902985B CN 201410148832 A CN201410148832 A CN 201410148832A CN 103902985 B CN103902985 B CN 103902985B
- Authority
- CN
- China
- Prior art keywords
- image
- lane line
- real
- roi
- line information
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a high-robustness real-time lane detection algorithm based on ROI. In order to achieve curve detection, an image is divided into an upper portion and a lower portion, namely a close shot area and a long shot area, and solutions are obtained through Hough and a hyperbolic curve pair model respectively. The whole image is mainly filtered through a transverse gradient operator so that the calculation speed can be increased and the purpose of real-time detection can be achieved; an area of gradient direction angles of a lane boundary image is counted through a sliding ROI window strategy, boundary noise of the abnormal gradient direction angles is eliminated, and therefore the accuracy of lane detection is guaranteed; in the hyperbolic curve pair model adopted in the long shot area, parameters in a close shot model are mainly used as initial parameters, a parameter K is finally determined through a search strategy, and the curve portion is detected. The high-robustness real-time lane detection algorithm is good in robustness in the complex road condition and the complex environment and lane counseling information can be provided in real time.
Description
Technical field
The present invention relates to driving safety monitoring field is and in particular to a kind of real-time track of strong robustness based on ROI is detectd
Method of determining and calculating.
Background technology
Recently, what ITS became comes into vogue, because people more pay close attention to vehicle safety aspect.There are a lot of bases with regard to ITS
In the research topic of vision, including obstacle detection, pedestrian's avoidance, lane line departure warning, anticollision etc..At these, there is challenge
In the task of property, lane detection is one of most important part of ITS.The main body of this technology is from complexity by some characteristics
Environment in isolate lane line information;However, the most of lane detection technology having existed to vile weather and
The impact of woods shade is very sensitive.Lane detection be an inner important component part of intelligent transportation system (ITS) it is proposed that
A kind of bend detecting system under complex environment, for example:Tree shade, complex road condition, there is road sign on road.
Current research has achieved the detection of lane line with different technology, for example:Algorithm based on color, it is based on
The algorithm at edge, the algorithm based on block diagram, the algorithm based on visual angle effect, the algorithm based on model.However, based on color
Algorithm with only having used an overall threshold value or specific colouring information goes to extract lane line and indicates, but this may have shade or
Impact is received when person's bad weather.For the algorithm based on edge, difficult point is the edge that abates the noise.Carry despite document
Go out algorithm and be discussed in detail whole problem, but also deposit much not enough.For the algorithm based on column, they are only capable of in car
Before do not have could run when other vehicles and its road sign.Based on the algorithm of visual angle effect, difficult point is to correct photographic head, and
And this algorithm is on slope and inefficent in the case of having wind.Lane line is replaced with model based on the algorithm of model, it for
When noise is more and lane line information disappears, there is more preferable robustness.In order to receive more structurally sound model parameter, should
With more preferable N6ise deletion algorithm.
Content of the invention
In order to solve the above problems, the invention provides a kind of strong robustness based on ROI real-time track detecting algorithm, first
First detect first image to extract adaptive threshold, subsequently with the transverse gradients operator with adaptive threshold to lane line
Information filters and uses fuzzy strategy of being searched to repair lane line information it is contemplated that Hough operand, need exclusion lane line noise and
Redundancy information, the method based on slip ROI and image association that employs extracts effective lane line;Subsequently, adopt in close shot region
Straight way is scanned for computing and is shown with the Hough conversion method of fast zoom table;Use hyperbola to mould in distant view region
Type simultaneously accepts the result calculating hyperbola of Hough computing to parameter, determines K parameter numerical value using search strategy simultaneously, simulates
Bend track, last output display entirety track.Can there is woods shade, other lane markings, Qi Tache in this method
Etc. accurately extract lane line information under complex situations.
In order to achieve the above object, the present invention is achieved by the following technical solutions:
The invention provides algorithm is detected in a kind of real-time track of the strong robustness based on ROI, described algorithm specifically include with
Lower step, step one, first image of detecting first are to extract initial adaptive threshold;
Step 2 simultaneously sets up coordinate system on the image of picked-up, and the upper left corner is initial point, laterally for x-axis, is longitudinally y-axis;Lean on
Recent photo as bottom part be referred to as close shot region, away from part be referred to as distant view region;
Step 3, driveway line information mistake is entered using the transverse gradients operator with adaptive threshold to image greyscale figure
Filter, enters runway information first and extracts:Before carrying out track rim detection, RGB image is converted into gray-scale figure, from RGB model
The formula changing into gray-scale figure is:;Wherein R, G, B are the component in [0,255] for the numeric distribution;Then adopt
The mode of the shade of transverse gradients operator enters driveway line detecting, and threshold value adopts adaptive threshold, and its shade is:
Step 4, open up ROI and extract effective track information:
The Grad of step 4.1 image and its direction are expressed as:
Wherein,WithRepresent the gradient in transverse direction and longitudinal direction respectively, its computing shade is as follows:
Step 4.2 willScope is set to, and be quantified as 180 angles, i.e. each angle
Spend and be, calculated direction angleIf,Do not existInterior, then it is taken as noise lane line information and disappear
Remove;
Step 4.3 carries out preliminary hough conversion using the pixel retaining, and primarily determines that lane line information, and along tentatively
Lane line information arranges the window of the slip that several have certain ROI scope, and to sliding window execution step 4.2 one by one,
Analysis exclusion is probably the isolated noise beyond lane line;
Step 4.4, from the beginning of second image, opens the little spy of image middle position amount of movement using lane line information in front and back
Property, this image and front several images being compared, extracting maximally effective lane line information to reduce subsequent line track
The operand of Hough conversion in matching;
Step 5, straight way and bend matching:
(1) straight way matching
Rectilinear stretch is fitted using Hough in image close shot region, its formula is as follows:
Matching adopts Hough conversion method, comprises the following steps that:
A) initially set up、Each angle andDatabase table, for Hough change search, to reduce computing
Amount;
B) obtain lane line information figure E and data set S,, wherein m is lane line information
Pixel number;
C) arrangeAs initial search point, calculate as d) parameter, and computer is set, be subsequently calculated by rows
Next point;
If d) currentMeet, then, setting, until;
E) retain maximumAnd E, fitting a straight line finishes;
(2)Bend matching
In image distant view region, using hyperbola to model, its mathematic(al) representation is:
Wherein,For vertical coordinate in image plane for the horizon, K,WithFor bend parameter;
Standard type according to Hough space cathetusHave:
Contrast can draw:,,,
Each parameter can be calculated by straight line Hough conversion portion, setting If, matched curve pointWith sliding window central pointMeet(Single
Position is pixel), then;
(3)The overall lane line in real time of output:By step(1)And step(2)Gained rectilinear stretch and bending track combine simultaneously
Display, the overall real-time lane line of output;
Step 6, each sliding window parameter of calculating simultaneously obtain adaptive threshold:From the beginning of second image, according to sliding window
Mouth opens up ROI scope as image subregion, and with this scope, the gray-scale figure after conversion is carried out adaptive using the method for Otsu
Threshold value is answered to extract, formula is expressed as:
Wherein t is required adaptive threshold,For i-th background than row,For i-th background mean value,For i-th object ratio,For i-th object average,For i-th image sub-district area image average, i is to slide
Dynamic window number, parameter calculation formula is as follows:
Wherein,The ratio of the pixel being k for GTG numerical value in i-th sliding window,For i-th sliding window
GTG numerical value is the pixel number of k,For pixel total number in i-th sliding window;
Threshold operation is:
Wherein, Bin(X, y)For imaging point, f(X, y)Respective coordinates for original pixel;
Step 7, from the beginning of second image, every image is first carried out step 6, and then execution step three is to step
Five, export overall lane line in real time.
Further improvement of the present invention is:Described algorithm also includes:To with transverse gradients operator mistake in step 3
Lane line information after filter carries out searching for repair strategy generally, searches for the shade that repair strategy adopted generally and is:
With current image lane line information data set it is, previous isIf A, B, C are not zero, D, E wherein have one
Individual be not zero andThen update.
Detect algorithm as a kind of real-time track of strong robustness based on ROI, the present invention has following some beneficial effects:
1. achieve the real-time detecting of the lane line in embedded system, operand is compared to current most calculations
Method has significant raising.
2. robust performance is strong, in complicated road environment, for example:Tree shade, road sign, wagon flow are many, have when bend
Extremely strong stability, and also have under the weather environment such as rainy day and night and same stablize new energy.
3. algorithm is not limited to current main flow algorithm, using the less algorithm to processor load, significantly relatively low processes
The energy consumption of device is so that whole system is simplified and stable.
The straight in real time bend detection algorithm under complex environment that the present invention provides, has in complicated road conditions and environment
Preferably robustness, can provide track information in real time.
Brief description
Fig. 1 system total algorithm flow chart
Fig. 2 adaptive thresholding algorithm flow chart
Fig. 3 carries the Embedded System Structure figure of algorithm
Specific embodiment
In order to deepen, to understanding of the present utility model, below in conjunction with drawings and Examples, this utility model to be made further
Describe in detail, this embodiment is only used for explaining this utility model, protection domain of the present utility model is not constituted and limits.
As Figure 1-3, present embodiments provide a kind of strong robustness based on ROI real-time track detecting algorithm, described
Algorithm specifically includes following steps:
Step one, first image of detecting first are to extract initial adaptive threshold;
Step 2 simultaneously sets up coordinate system on the image of picked-up, and the upper left corner is initial point, laterally for x-axis, is longitudinally y-axis;Lean on
Recent photo as bottom part be referred to as close shot region, away from part be referred to as distant view region;Due to lane line in real process
Appear in certain regional extent, here, we only process a part for image, to reduce operand.In close shot region
Interior, Hough search arithmetic can be carried out in a subsequent step, and draw rectilinear stretch;In distant view region, can be with reference to Hough
The parameter of computing simultaneously searches for K parameter numerical value, to draw bend track.
Step 3, driveway line information mistake is entered using the transverse gradients operator with adaptive threshold to image greyscale figure
Filter, before carrying out track rim detection, needs first RGB image to be converted into gray-scale figure, traffic lane line is typically white, yellow
Or red, to form a sharp contrast with road surface, in order to preserve this characteristic of traffic lane line, change into ash from RGB model
The employing formula of rank figure is:;Wherein R, G, B are the component in [0,255] for the numeric distribution;
The performance of rim detection is important for system, and algorithm has in atrocious weather and poor environment
Under there is robustness, potential lane line message can be remained;Simultaneously because this algorithm is to use in embedded systems,
So algorithm can not take excessive operand, in order to avoid impact is produced on real-time.Not using canny algorithm the reason is that it is right
The load excessive of processor, here enters driveway line detecting by the way of the shade claiming to become transverse gradients operator, and threshold value is adopted
With adaptive threshold, its shade is:
For obtaining the maximum amount of lane line information, improve system robustness, to the car after filtering with transverse gradients operator
Diatom information carries out searching for repair strategy generally, searches for the shade that repair strategy adopted generally and is:
With current image lane line information data set it is, previous isIf A, B, C are not zero, D, E wherein have one
Individual be not zero andThen update.
Step 4, open up ROI and extract effective track information:
The Grad of step 4.1 image and its direction are expressed as:
Wherein,WithRepresent the gradient in transverse direction and longitudinal direction respectively, its computing shade is as follows:
Step 4.2 willScope is set to, and be quantified as 180 angles, i.e. each angle
For, calculated direction angleIf,Do not existInterior, then it is taken as noise lane line and seek advice from and eliminate;
Step 4.3 carries out preliminary hough conversion using the pixel retaining, and primarily determines that lane line information, and along tentatively
Lane line information arranges the window of the slip that several have certain ROI scope, and to sliding window execution step 4.2 one by one,
Analysis exclusion is probably the isolated noise beyond lane line;
Step 4.4, from the beginning of second image, opens the little spy of image middle position amount of movement using lane line consulting in front and back
Property, this image and front several images being compared, extracting maximally effective lane line information to reduce subsequent line track
The operand of Hough conversion in matching;
Step 5, straight way and bend matching:
(1) straight way matching
Rectilinear stretch is fitted using Hough in image close shot region, its formula is as follows:
Matching adopts Hough conversion method, comprises the following steps that:
A) initially set up、Each angle andDatabase table, for Hough change search, to reduce computing
Amount;
B) obtain lane line information figure E and data set S,, wherein m is lane line information
Pixel number;
C) arrangeAs initial search point, calculate as d) parameter, and computer is set, be subsequently calculated by rows
Next point;
If d) currentMeet, then, setting, until;
E) retain maximumAnd E, fitting a straight line finishes;
(2)Bend matching
In image distant view region, using hyperbola to model, its mathematic(al) representation is:
Wherein,For vertical coordinate in image plane for the horizon, K,WithFor bend parameter;
Standard type according to Hough space cathetusHave:
Contrast can draw:,,,
Each parameter can be calculated by straight line Hough conversion portion, setting If, matched curve pointWith sliding window central pointMeet(Single
Position is pixel), then;
(3)The overall lane line in real time of output:By step(1)And step(2)Gained rectilinear stretch and bending track combine simultaneously
Display, the overall real-time lane line of output;
Step 6, each sliding window parameter of calculating simultaneously obtain adaptive threshold:From the beginning of second image, according to sliding window
Mouth opens up ROI scope as image subregion, and with this scope, the gray-scale figure after conversion is carried out adaptive using the method for Otsu
Threshold value is answered to extract, formula is expressed as:
Wherein t is required adaptive threshold,For i-th background than row,For i-th background mean value,For i-th object ratio,For i-th object average,For i-th image sub-district area image average, i is to slide
Dynamic window number, parameter calculation formula is as follows:
Wherein,The ratio of the pixel being k for GTG numerical value in i-th sliding window,For i-th sliding window
GTG numerical value is the pixel number of k,For pixel total number in i-th sliding window;
Threshold operation is:
Wherein, Bin(X, y)For imaging point, f(X, y)Respective coordinates for original pixel;
Step 7, from the beginning of second image, every image is first carried out step 6, and then execution step three is to step
Five, export overall lane line in real time.
The present embodiment is a kind of straight in real time bend detection method under complex environment, initially with adaptive threshold
Transverse gradients operator image greyscale figure is entered driveway line information filter, and using search for generally strategy repair filter after car
Diatom information, improves the accuracy of system robustness and lane detection;Then slip ROI window policy is adopted to count lane boundary
Information figure gradient direction angle, the border noise information of rejecting abnormalities gradient direction angle, an image lane line money before and after simultaneously associating
News, reduce redundancy track line boundary information, to reduce Hough translation operation amount.Rectilinear stretch part proposes a kind of fast zoom table
Hough conversion method, for straight line detecting;Racetrack portion adopts hyperbola to model, and its initial parameter value is undertaken in straight line
Lane portion, and its K parameter is finally determined by search strategy, realize the detecting of racetrack portion.This algorithm is in complicated road conditions
And there is in environment preferable robustness, track information can be provided in real time.
Claims (2)
1. a kind of strong robustness based on ROI real-time track detecting algorithm it is characterised in that:Described algorithm specifically includes following step
Suddenly:
Step one, first image of detecting first are to extract initial adaptive threshold;
Step 2 simultaneously sets up coordinate system on the image of picked-up, and the upper left corner is initial point, laterally for x-axis, is longitudinally y-axis;Near shadow
As bottom part be referred to as close shot region, away from part be referred to as distant view region;
Step 3, using the transverse gradients operator with adaptive threshold, image greyscale figure is entered with driveway line information filter, first
Advanced runway information is extracted:Before carrying out track rim detection, RGB image is converted into gray-scale figure, changes into from RGB model
The formula of gray-scale figure is:;Wherein R, G, B are the component in [0,255] for the numeric distribution;Then adopt laterally
The mode of the shade of gradient operator enters driveway line detecting, and threshold value adopts adaptive threshold, and its shade is:
Step 4, open up ROI and extract effective track information:
The Grad of step 4.1 image and its direction are expressed as:
Wherein,WithRepresent the gradient in transverse direction and longitudinal direction respectively, its computing shade is as follows:
Step 4.2 willScope is set to, and it is quantified as 180 angles, that is, each angle is, calculated direction angleIf,Do not existInterior, then it is taken as noise lane line information and eliminate;
Step 4.3 carries out preliminary hough conversion using the pixel retaining, and primarily determines that lane line information, and along preliminary track
Line information arranges the window of the slip that several have certain ROI scope, and to sliding window execution step 4.2 one by one, analyzes
Exclusion is probably the isolated noise beyond lane line;
Step 4.4, from the beginning of second image, opens the little characteristic of image middle position amount of movement using lane line information in front and back, will
This image is compared with front several images, extracts maximally effective lane line information to reduce in the matching of subsequent line track
The operand of Hough conversion;
Step 5, straight way and bend matching:
(1) straight way matching
Rectilinear stretch is fitted using Hough in image close shot region, its formula is as follows:
Matching adopts Hough conversion method, comprises the following steps that:
A) initially set up、Each angle andDatabase table, for Hough change search, to reduce operand;
B) obtain lane line information figure E and data set S,, wherein m is lane line information pixel
Point number;
C) arrangeAs initial search point, calculate as d) parameter, and computer is set, be subsequently calculated by rows the next one
Point;
If d) currentMeet, then, setting, until;
E) retain maximumAnd E, fitting a straight line finishes;
(2)Bend matching
In image distant view region, using hyperbola to model, its mathematic(al) representation is:
Wherein,For vertical coordinate in image plane for the horizon, K,WithFor bend parameter;
Standard type according to Hough space cathetusHave:
Contrast can draw:,,,
Each parameter can be calculated by straight line Hough conversion portion, setting If, matched curve pointWith sliding window central pointMeet(Single
Position is pixel), then;
(3)The overall lane line in real time of output:By step(1)And step(2)Gained rectilinear stretch and bending track combine and show,
The overall real-time lane line of output;
Step 6, each sliding window parameter of calculating simultaneously obtain adaptive threshold:From the beginning of second image, made according to sliding window
Open up ROI scope for image subregion, and adaptive thresholding is carried out to the gray-scale figure after conversion using the method for Otsu with this scope
Value is extracted, and formula is expressed as:
Wherein t is required adaptive threshold,For i-th background than row,For i-th background mean value,For
I object ratio,For i-th object average,For i-th image sub-district area image average, i is sliding window
Number, parameter calculation formula is as follows:
Wherein,The ratio of the pixel being k for GTG numerical value in i-th sliding window,For i-th sliding window GTG
Numerical value is the pixel number of k,For pixel total number in i-th sliding window;
Threshold operation is:
Wherein, Bin(X, y)For imaging point, f(X, y)Respective coordinates for original pixel;
Step 7, from the beginning of second image, every image is first carried out step 6, and then execution step three is to step 5, real
When the overall lane line of output.
2. a kind of strong robustness based on ROI according to claim 1 real-time track detecting algorithm it is characterised in that:Institute
State algorithm also to include:Step 3 carries out searching for generally repairing plan to the lane line information after filtering with transverse gradients operator
Slightly, searching for the shade that repair strategy adopted generally is:
With current image lane line information data set it is, previous isIf A, B, C are not zero, D, E wherein have one not
Be zero andThen update.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410148832.8A CN103902985B (en) | 2014-04-15 | 2014-04-15 | High-robustness real-time lane detection algorithm based on ROI |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410148832.8A CN103902985B (en) | 2014-04-15 | 2014-04-15 | High-robustness real-time lane detection algorithm based on ROI |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103902985A CN103902985A (en) | 2014-07-02 |
CN103902985B true CN103902985B (en) | 2017-02-15 |
Family
ID=50994297
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410148832.8A Active CN103902985B (en) | 2014-04-15 | 2014-04-15 | High-robustness real-time lane detection algorithm based on ROI |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103902985B (en) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105353373B (en) * | 2015-12-16 | 2018-04-20 | 武汉大学 | One kind is based on Hough transform Ground Penetrating Radar target extraction method and device |
CN106778668B (en) * | 2016-12-30 | 2019-08-09 | 明见(厦门)技术有限公司 | A kind of method for detecting lane lines of robust that combining RANSAC and CNN |
CN107180228A (en) * | 2017-05-02 | 2017-09-19 | 开易(北京)科技有限公司 | A kind of grad enhancement conversion method and system for lane detection |
CN107392139B (en) * | 2017-07-18 | 2020-10-20 | 海信集团有限公司 | Lane line detection method based on Hough transform and terminal equipment |
CN107590438A (en) * | 2017-08-16 | 2018-01-16 | 中国地质大学(武汉) | A kind of intelligent auxiliary driving method and system |
CN109543493B (en) * | 2017-09-22 | 2020-11-20 | 杭州海康威视数字技术股份有限公司 | Lane line detection method and device and electronic equipment |
CN107901907B (en) * | 2017-09-30 | 2019-12-20 | 惠州市德赛西威汽车电子股份有限公司 | Lane line detection method based on color block statistics |
CN108830165A (en) * | 2018-05-22 | 2018-11-16 | 南通职业大学 | A kind of method for detecting lane lines considering front truck interference |
CN109087326A (en) * | 2018-09-18 | 2018-12-25 | 辽宁工业大学 | Otsu algorithm based on local auto-adaptive |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5835028A (en) * | 1997-05-05 | 1998-11-10 | Bender; Lee | Lane marker position sensor and alarm |
CN101639983A (en) * | 2009-08-21 | 2010-02-03 | 任雪梅 | Multilane traffic volume detection method based on image information entropy |
CN101826257A (en) * | 2010-03-29 | 2010-09-08 | 北京市公安局公安交通管理局 | Method for detecting motor way travel condition in real time |
CN103366584A (en) * | 2013-06-20 | 2013-10-23 | 银江股份有限公司 | Real-time traffic flow detection-based self-adaptive tide lane control method |
CN103617412A (en) * | 2013-10-31 | 2014-03-05 | 电子科技大学 | Real-time lane line detection method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4162618B2 (en) * | 2004-03-12 | 2008-10-08 | 株式会社豊田中央研究所 | Lane boundary judgment device |
-
2014
- 2014-04-15 CN CN201410148832.8A patent/CN103902985B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5835028A (en) * | 1997-05-05 | 1998-11-10 | Bender; Lee | Lane marker position sensor and alarm |
CN101639983A (en) * | 2009-08-21 | 2010-02-03 | 任雪梅 | Multilane traffic volume detection method based on image information entropy |
CN101826257A (en) * | 2010-03-29 | 2010-09-08 | 北京市公安局公安交通管理局 | Method for detecting motor way travel condition in real time |
CN103366584A (en) * | 2013-06-20 | 2013-10-23 | 银江股份有限公司 | Real-time traffic flow detection-based self-adaptive tide lane control method |
CN103617412A (en) * | 2013-10-31 | 2014-03-05 | 电子科技大学 | Real-time lane line detection method |
Non-Patent Citations (1)
Title |
---|
基于线性抛物线模型的车道检测与跟踪方法;洪敏等;《长江大学学报(自然科学版)》;20130305;第10卷(第7期);第53-58页 * |
Also Published As
Publication number | Publication date |
---|---|
CN103902985A (en) | 2014-07-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103902985B (en) | High-robustness real-time lane detection algorithm based on ROI | |
CN104008387B (en) | Lane line detection method based on feature point piecewise linear fitting | |
CN105678285B (en) | A kind of adaptive road birds-eye view transform method and road track detection method | |
CN105261020B (en) | A kind of express lane line detecting method | |
CN107330376B (en) | Lane line identification method and system | |
CN105426864B (en) | One kind being based on the matched Multi-lane Lines Detection method of equidistant marginal point | |
CN104392212B (en) | The road information detection and front vehicles recognition methods of a kind of view-based access control model | |
Kong et al. | General road detection from a single image | |
CN103500322B (en) | Automatic lane line identification method based on low latitude Aerial Images | |
CN103077617B (en) | Based on crossing intelligent traffic light supervisory systems and the method for computer vision | |
Li et al. | Nighttime lane markings recognition based on Canny detection and Hough transform | |
CN106887004A (en) | A kind of method for detecting lane lines based on Block- matching | |
CN102982304B (en) | Utilize polarized light image to detect the method and system of vehicle location | |
CN107025432A (en) | A kind of efficient lane detection tracking and system | |
CN101980245B (en) | Adaptive template matching-based passenger flow statistical method | |
CN104766046A (en) | Detection and recognition algorithm conducted by means of traffic sign color and shape features | |
CN104029680A (en) | Lane departure warning system and method based on monocular camera | |
CN104156731A (en) | License plate recognition system based on artificial neural network and method | |
CN105404857A (en) | Infrared-based night intelligent vehicle front pedestrian detection method | |
CN103324930A (en) | License plate character segmentation method based on grey level histogram binaryzation | |
CN105528609A (en) | Vehicle license plate location method based on character position | |
CN103793708A (en) | Multi-scale license plate precise locating method based on affine correction | |
CN102999749A (en) | Intelligent safety belt regulation violation event detecting method based on face detection | |
CN107705301A (en) | A kind of highway graticule damage testing method based on unmanned plane highway map picture | |
CN105117726A (en) | License plate positioning method based on multi-feature area accumulation |
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 |