CN107590451A - A kind of method for detecting lane lines - Google Patents
A kind of method for detecting lane lines Download PDFInfo
- Publication number
- CN107590451A CN107590451A CN201710784083.1A CN201710784083A CN107590451A CN 107590451 A CN107590451 A CN 107590451A CN 201710784083 A CN201710784083 A CN 201710784083A CN 107590451 A CN107590451 A CN 107590451A
- Authority
- CN
- China
- Prior art keywords
- lane line
- lane
- maximum
- line
- road
- 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
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
A kind of method for detecting lane lines is related to road Identification detection field, comprises the following steps：Gather the image of lane line；Image containing lane line is handled, extracts the marginal point of road or so lane line；Lane line is configured to the function model of an entirety, using Feedback Neural Network, using the lane line marginal point extracted, parameter Estimation is carried out to function model, obtains the Convex Functions that Open Side Down；The maximum of function is obtained using the upper convexity of Convex Functions, the left side of determining maximum is left-lane line, and the right side of maximum is right-lane line；Left and right lane line is carried out using least square method to be fitted and reconstruct again, reaches the purpose of detection lane line.General location of this method without manually prejudging out left and right lane line in advance as other methods.The accuracy of detection of left and right lane line is improved, there is good universality, in the case where illumination, shade and other pavement markers are disturbed, to bending and straight lane line, can be detected.
Description
Technical field
The present invention relates to road Identification detection field, and in particular to a kind of method for detecting lane line.
Background technology
The artificial intelligence navigation of view-based access control model is an important application of computer vision field, because its method is simple, can
Complicated road environment is adapted to be widely applied in the road Identification system of autonomous vehicle.It is currently based on the lane line inspection of vision
Method of determining and calculating includes feature based and based on the class of model two.The algorithm of feature based mainly utilizes the information such as color, texture, shape
To extract lane line, but when road surface illumination changes, there is situations such as water stain, shade and all will in markings abrasion, road surface
Influence Detection results.Algorithm based on model estimates road model first, then determines model parameter using image information, commonly uses
Road model have straight line model, parabola model, cubic spline curve model.Numerous studies are found：Although straight line model is calculated
Method simple operation speed is fast, but the lane detection application for being not suitable for bending is narrow；Parabola model is easily by complicated road
The influence robustness of environment is bad；Although cubic spline curve model Detection results are relatively good, algorithm complicated calculations amount is big.
The content of the invention
In order to solve problems of the prior art, the invention provides a kind of new method of lane detection, the party
Method not only can to both sides lane line Holistic modeling, and can also be easy differentiation left and right lane line.
The technical proposal for solving the technical problem of the invention is as follows：
A kind of method for detecting lane lines, this method comprise the following steps：
Step 1：Gather the image of lane line；
Step 2：Described image containing lane line is handled, extracts the marginal point of road or so lane line；
Step 3：The lane line is configured to the function model of an entirety, using Feedback Neural Network, uses extraction
The lane line marginal point gone out, parameter Estimation is carried out to function model, obtains the Convex Functions that Open Side Down；
Step 4：The maximum of function is obtained using the upper convexity of the Convex Functions, the left side of determining maximum is a left side
Lane line, the right side of maximum is right-lane line；Using least square method to left and right lane line progress again fitting and again
Structure, reach the purpose of detection lane line.
The beneficial effects of the invention are as follows：Separation of this method using the maximum of points of Convex Functions as left and right lane line,
Without manually prejudging out the general location of left and right lane line in advance as other methods.Reduce the mistake for manually prejudging and bringing
Difference, improve the accuracy of detection of left and right lane line.In addition this method has good universality, on illumination, shade and other road surfaces
In the case of mark interference, to bending and straight lane line, it can detect.
Brief description of the drawings
Fig. 1 is original image.
Fig. 2 is 45 ° of Line filters and 135 ° of Line filters.
Fig. 3 is lane line pattern function curve map.
Fig. 4 is the edge point diagram in left and right track.
Fig. 5 is reconstruction result of the Feedback Neural Network to new function model.
Fig. 6 is least square method to the two-part reconstruction result of new function model or so.
Embodiment
The present invention is described in further details with reference to the accompanying drawings and examples.
A kind of method for detecting lane lines, this method comprise the following steps：
Step 1：The image of lane line is gathered, as shown in figure 1, according to RGB color to the conversion of YUV color spaces
Formula Y=0.299R+0.587G+0.114B, processes the image into gray-scale map.
Step 2：To the gray-scale map containing lane line, filtered respectively with 45 ° and 135 ° of straight line as shown in Figure 2
Device filtering process, by 45 ° of filtering images and 135 ° of filtering image superpositions, and then extract the marginal point of left and right lane line.
Step 3：The lane line of road or so is configured to the function model of an entirety, as shown in figure 3, respectively three
The common lane line of kind, that is, turn right, keep straight on, turn left, use the coordinate value for the marginal point for extracting the left and right lane line, such as figure
Shown in 4, as the input data of three layers of Feedback Neural Network, parameter Estimation is carried out to function model, obtains opening for determination
The downward Convex Functions of mouth, as shown in figure 5, completing the fitting of function curve model.
Step 4：The maximum of function is obtained using the upper convexity of the Convex Functions, as shown in " small circle " in Fig. 5,
Can be left-lane line on the left of determining maximum, the right side of maximum is right-lane line, distinguishes the data on the lane line of left and right
Point, finally left and right lane line is carried out using least square method to be fitted and reconstruct again, as shown in fig. 6, reaching detection lane line
Purpose.
Claims (2)
1. a kind of method for detecting lane lines, it is characterised in that this method comprises the following steps：
Step 1：Gather the image of lane line；
Step 2：Described image containing lane line is handled, extracts the marginal point of road or so lane line；
Step 3：The lane line is configured to the function model of an entirety, using Feedback Neural Network, uses what is extracted
The lane line marginal point, parameter Estimation is carried out to function model, obtains the Convex Functions that Open Side Down；
Step 4：The maximum of function is obtained using the upper convexity of the Convex Functions, the left side of determining maximum is left-lane
Line, the right side of maximum is right-lane line；The left and right lane line is carried out using least square method to be fitted and reconstruct again, arrived
Up to the purpose of detection lane line.
2. a kind of method for detecting lane lines according to claim 1, it is characterised in that in the step 2, to containing car
The gray-scale map of diatom, respectively with 45 ° and 135 ° of Line filter filtering process, by 45 ° of filtering images and 135 ° of filtering images
Fusion, and then extract the marginal point of road or so lane line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710784083.1A CN107590451A (en) | 2017-09-04 | 2017-09-04 | A kind of method for detecting lane lines |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710784083.1A CN107590451A (en) | 2017-09-04 | 2017-09-04 | A kind of method for detecting lane lines |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107590451A true CN107590451A (en) | 2018-01-16 |
Family
ID=61050272
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710784083.1A Pending CN107590451A (en) | 2017-09-04 | 2017-09-04 | A kind of method for detecting lane lines |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107590451A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109308468A (en) * | 2018-09-21 | 2019-02-05 | 电子科技大学 | A kind of method for detecting lane lines |
| CN111007064A (en) * | 2019-12-13 | 2020-04-14 | 常州大学 | Intelligent logging lithology identification method based on image identification |
| CN115019278A (en) * | 2022-07-13 | 2022-09-06 | 北京百度网讯科技有限公司 | Lane line fitting method and device, electronic equipment and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080114532A1 (en) * | 2006-11-14 | 2008-05-15 | Toyota Jidosha Kabushiki Kaisha | Lane detecting apparatus |
| CN104008645A (en) * | 2014-06-12 | 2014-08-27 | 湖南大学 | Lane line predicating and early warning method suitable for city road |
| CN104657727A (en) * | 2015-03-18 | 2015-05-27 | 厦门麦克玛视电子信息技术有限公司 | Lane line detection method |
| CN106778668A (en) * | 2016-12-30 | 2017-05-31 | 明见（厦门）技术有限公司 | A kind of method for detecting lane lines of the robust of joint RANSAC and CNN |
| CN107092862A (en) * | 2017-03-16 | 2017-08-25 | 浙江零跑科技有限公司 | A kind of track edge detection method based on convolutional neural networks |
-
2017
- 2017-09-04 CN CN201710784083.1A patent/CN107590451A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080114532A1 (en) * | 2006-11-14 | 2008-05-15 | Toyota Jidosha Kabushiki Kaisha | Lane detecting apparatus |
| CN104008645A (en) * | 2014-06-12 | 2014-08-27 | 湖南大学 | Lane line predicating and early warning method suitable for city road |
| CN104657727A (en) * | 2015-03-18 | 2015-05-27 | 厦门麦克玛视电子信息技术有限公司 | Lane line detection method |
| CN106778668A (en) * | 2016-12-30 | 2017-05-31 | 明见（厦门）技术有限公司 | A kind of method for detecting lane lines of the robust of joint RANSAC and CNN |
| CN107092862A (en) * | 2017-03-16 | 2017-08-25 | 浙江零跑科技有限公司 | A kind of track edge detection method based on convolutional neural networks |
Non-Patent Citations (2)
| Title |
|---|
| CHI-FENG WU等: "Applying a Functional Neurofuzzy Network to Real-Time Lane Detection and Front-Vehicle Distance Measurement", 《IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART C: APPLICATIONS AND REVIEWS》 * |
| 王春阳: "车道检测方法综述", 《科技视界》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109308468A (en) * | 2018-09-21 | 2019-02-05 | 电子科技大学 | A kind of method for detecting lane lines |
| CN109308468B (en) * | 2018-09-21 | 2021-09-24 | 电子科技大学 | Lane line detection method |
| CN111007064A (en) * | 2019-12-13 | 2020-04-14 | 常州大学 | Intelligent logging lithology identification method based on image identification |
| CN115019278A (en) * | 2022-07-13 | 2022-09-06 | 北京百度网讯科技有限公司 | Lane line fitting method and device, electronic equipment and medium |
| CN115019278B (en) * | 2022-07-13 | 2023-04-07 | 北京百度网讯科技有限公司 | Lane line fitting method and device, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| 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 | |
| CN105700532B (en) | Vision-based transformer substation inspection robot navigation positioning control method | |
| WO2017041396A1 (en) | Driving lane data processing method, device, storage medium and apparatus | |
| Yuan et al. | Robust lane detection for complicated road environment based on normal map | |
| CN103902985B (en) | High-robustness real-time lane detection algorithm based on ROI | |
| CN105426864A (en) | Multiple lane line detecting method based on isometric peripheral point matching | |
| Li et al. | Nighttime lane markings recognition based on Canny detection and Hough transform | |
| CN105260693A (en) | Laser two-dimensional code positioning method | |
| CN104077756B (en) | Direction filtering method based on lane line confidence | |
| CN102982304B (en) | Utilize polarized light image to detect the method and system of vehicle location | |
| CN108280450A (en) | A kind of express highway pavement detection method based on lane line | |
| CN103996198A (en) | Method for detecting region of interest in complicated natural environment | |
| CN107590451A (en) | A kind of method for detecting lane lines | |
| CN107705301A (en) | A kind of highway graticule damage testing method based on unmanned plane highway map picture | |
| CN108171695A (en) | A kind of express highway pavement detection method based on image procossing | |
| CN106887004A (en) | A kind of method for detecting lane lines based on Block- matching | |
| CN104217427A (en) | Method for positioning lane lines in traffic surveillance videos | |
| CN102999916A (en) | Edge extraction method of color image | |
| US11941892B2 (en) | Method and device for providing data for creating a digital map | |
| CN105809149A (en) | Lane line detection method based on straight lines with maximum length | |
| CN108304749A (en) | The recognition methods of road speed line, device and vehicle | |
| CN109389165A (en) | Oil level gauge for transformer recognition methods based on crusing robot | |
| CN103996031A (en) | Self adaptive threshold segmentation lane line detection system and method | |
| CN111652033A (en) | Lane line detection method based on OpenCV |
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: 20180116 |
|
| WD01 | Invention patent application deemed withdrawn after publication |