CN108182430B - Double-area lane line identification system and method - Google Patents
Double-area lane line identification system and method Download PDFInfo
- Publication number
- CN108182430B CN108182430B CN201810163493.9A CN201810163493A CN108182430B CN 108182430 B CN108182430 B CN 108182430B CN 201810163493 A CN201810163493 A CN 201810163493A CN 108182430 B CN108182430 B CN 108182430B
- Authority
- CN
- China
- Prior art keywords
- lane line
- double
- area
- identification
- judging
- 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
Images
Classifications
-
- 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
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a double-area lane line identification system, which is characterized in that: the device comprises a double-area lane line identification unit, a double-area lane line judgment unit and a double-area lane line result deviation calculation unit. The invention also relates to a double-area lane line identification method. According to the double-region lane line identification system and method, the lane lines on the left and right of the main vehicle are respectively detected in different regions, so that the identification precision of the driving assistance system is improved, the noise interference of identification is reduced, and the identification efficiency is enhanced.
Description
Technical Field
The invention belongs to the field of driving assistance systems and automatic driving, relates to a lane line identification system and method, and particularly relates to a dual-area lane line identification system and method capable of improving identification accuracy and identification efficiency of a driving assistance system.
Background
Structured road detection is one of the important issues in the research of driving assistance systems (automatic driving systems). Only if the information of the lane line is accurately known, the position and the direction of the vehicle relative to the lane can be accurately obtained. In real urban traffic, the most common is structured roads, and the structured road profile should be relatively regular, which means a standardized road with clear lane line marks and road boundaries. The road area and the non-road area should be obviously divided by drawing a lane line.
The existing lane line detection process is roughly divided into two steps: detecting the edge points of the lane lines and fitting the edge curves of the lane lines. A set of two unobstructed lane lines is typically visible in a distance directly in front of the vehicle in which the camera is mounted, and existing systems only initialize a detection zone in this area for lane line detection. Once the lane lines are extracted, a straight line fitting algorithm may be used to determine the lane line equation and separate the lane lines from the overall region. In order to reduce the amount of calculation and to reduce the disturbance of environmental factors on the road (shadows, cracks, other objects such as other vehicle disturbances, etc.), it is necessary to use some dynamic processing method of the sequence of images: (1) limiting lane line search to dynamic frame regions including lane lines at each frame of the image sequence; (2) defining the boundary point of the lane line as a point in a search area, wherein the jump of the boundary gray value is maximum and exceeds a preset threshold; (3) if the number of the detected lane line boundary points exceeds a certain number, the approximate lane line is obtained through a straight line fitting algorithm, and otherwise, a historical frame and a current frame are used for deduction.
The existing lane identification technology is based on the area identification of the whole double lanes. Even if the identification result of the vehicle line on one side is stable and continuous and the result is good, the influence caused by the shading of the physical environment object of the vehicle line on the other side, the change of the transmission angle of the light source position, meteorological phenomena and the like can cause the failure of the whole road line identification algorithm. Therefore, the reliability of automatic identification of the lane is reduced, and the practicability of the whole driving assistance system is also reduced. Through a search for a patent publication, no patent publication that is the same as the present patent application is found.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a double-area lane line identification system which can improve the identification precision of a driving auxiliary system and an intelligent driving system, reduce the identification noise interference and enhance the identification efficiency.
The present invention is directed to overcome the disadvantages of the prior art, and to provide a method for recognizing a lane line in a dual area, which judges whether or not the lane line information measured from a camera is recognized, compensates the lane line information, and improves the reliability of the lane line information.
The technical problem to be solved by the invention is realized by the following technical scheme:
the utility model provides a two regional lane line identification system which characterized in that: the device comprises a double-region lane line identification unit, a double-region lane line judgment unit and a double-region lane line result deviation calculation unit, wherein the double-region lane line identification unit comprises a camera matting module, a double-region dividing module and a lane line edge detection module; the double-region lane line judging unit comprises a reliability pre-judging module and a double-region identification logic judging module; the double-area lane line result deviation calculating unit comprises a double-area movable processing module, a continuity tracking module and an identification result high-precision processing module.
And the double-area lane line judging unit receives the information of the double-area lane line identifying unit, and judges that the lane line is not identified when the information of both areas does not pass the credibility prejudgment.
And the information of the two areas of the double-area lane line judging unit is judged in advance through the reliability and enters a double-area identification logic judging module.
And the two-region lane line judging unit judges that the lane line is identified and outputs the lane line to the intelligent driving system when the two-region information is lower than a set deviation threshold value.
And the information of the two areas of the double-area lane line judging unit exceeds a set deviation threshold value, the double-area lane line judging unit judges that the double-area lane line judging unit enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, and the area with a larger continuity value is judged as the lane line identification and is output to the intelligent driving system.
And the double-area lane line judging unit judges that the area information of one area passes the reliability prediction and the other area does not pass the reliability prediction, judges that the information enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, if the information is larger than the continuity judging threshold, judges that the lane line is identified and outputs the lane line to the intelligent driving system, and if the information is smaller than the continuity judging threshold, judges that the lane line is not identified.
A method for identifying under a dual-region lane is characterized in that: the method comprises the following steps:
1) the camera is scratched and is divided into areas: setting a front watching angle and an optimal watching range by using a front watching model, removing unnecessary fields, reducing unnecessary operation burden and determining a region in front of a lane line;
2) dividing a double-area lane line: dividing the area in front of the lane line determined in the step 1) into a left lane line area and a right lane line area;
3) and (3) detecting the edges of the lane lines in the double areas: judging whether the identification object is a white line or a yellow line by using the image pigment component, and then detecting an edge by using 2-valued processing, Sobel filtering, differential processing and Hough transformation;
4) judging the lane lines of the two regions: judging whether the output results of the two areas meet the reliability, and pre-judging according to the reliability;
if the two areas are judged to pass the reliability, entering a double-area identification logic judgment module:
if the deviation does not exceed the set deviation threshold, judging lane line recognition, and outputting the output result to an intelligent driving system as a reference lane input signal;
if the deviation exceeds the set deviation threshold value, the vehicle enters a continuity tracking module and is continuously compared with the result of the previous n periods, and the area with the larger continuity value is judged to be lane line identification and output to an intelligent driving system;
if the detection results of the two areas are not reliable (such as recognition error and obvious noise) the signals are used as input signals of the intelligent driving system which do not meet the automatic driving conditions;
if one region passes the reliability judgment and the other region does not pass the reliability judgment, judging that the vehicle enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, if the vehicle is larger than the continuity judgment threshold, judging that the lane line is identified and outputting the lane line to the intelligent driving system, and if the vehicle is smaller than the continuity judgment threshold, judging that the lane line is not identified;
5) adjusting the deviation of the result of the double-area lane line: when the lane enters a route with larger curvature and the lane line deviates from the detection area of the initial state, the detection area can be automatically moved in real time, meanwhile, the deviation result generated by the movement area and the main vehicle is automatically corrected, and the deviation result is output as an automatic driving control deviation value after the deviation optimization processing.
The invention has the advantages and beneficial effects that:
1. the invention provides a double-region lane line identification system, which achieves the purpose of improving the identification precision of a driving assistance system by respectively detecting lane lines on the left and right sides of a main vehicle in regions.
2. The invention provides a novel method based on double-region identification aiming at the limitation of identifying the lane line in a single region of interest, and solves the problems of overall identification efficiency and robustness of an identification system when the identification effect of the lane line on one side is poor.
Drawings
FIG. 1 is a schematic diagram of a dual-zone lane line recognition system according to the present invention;
FIG. 2 is a flow chart of the dual zone lane line identification of the present invention;
FIG. 3 is a flow chart of the dual zone lane line determination of the present invention;
FIG. 4 is a flowchart of the dual-zone lane line result deviation calculation of the present invention.
Detailed Description
The present invention is further illustrated by the following specific examples, which are intended to be illustrative, not limiting and are not intended to limit the scope of the invention.
A double-area lane line identification system is characterized in that: the device comprises a double-region lane line identification unit, a double-region lane line judgment unit and a double-region lane line result deviation calculation unit, wherein the double-region lane line identification unit comprises a camera matting module, a double-region dividing module and a lane line edge detection module; the double-region lane line judging unit comprises a reliability pre-judging module and a double-region identification logic judging module; the double-area lane line result deviation calculating unit comprises a double-area movable processing module, a continuity tracking module and an identification result high-precision processing module.
The double-area lane line judging unit receives the information of the double-area lane line identifying unit, and judges that the lane line is not identified when the information of the two areas does not pass the credibility prejudgment.
And the information of the two areas of the double-area lane line judging unit is pre-judged according to the reliability and enters a double-area identification logic judging module.
And the two-region lane line judging unit judges that the lane line is identified and outputs the lane line to the intelligent driving system when the two-region information is lower than a set deviation threshold value.
And the information of the two areas of the double-area lane line judging unit exceeds a set deviation threshold value, the double-area lane line judging unit judges that the double-area lane line judging unit enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, and the area with a larger continuity value is judged as the lane line identification and is output to the intelligent driving system.
And the double-region lane line judging unit judges that the region information of one region passes the reliability prejudgment and the region information of the other region does not pass the reliability prejudgment, judges that the lane line enters the continuity tracking module and is continuously compared with the result of the previous n periods, if the lane line is larger than the continuity judging threshold, judges that the lane line is identified and outputs the lane line to the intelligent driving system, and if the lane line is smaller than the continuity judging threshold, judges that the lane line is not identified.
A method for identifying under a dual-region lane is characterized in that: the method comprises the following steps:
1) the camera is scratched and is divided into areas: setting a front watching angle and an optimal watching range by using a front watching model, removing unnecessary fields, reducing unnecessary operation burden and determining a region in front of a lane line;
2) dividing a double-area lane line: dividing the area in front of the lane line determined in the step 1) into a left lane line area and a right lane line area;
3) and (3) detecting the edges of the lane lines in the double areas: the method comprises the steps of judging whether an identification object is a white line or a yellow line by using image pigment components, detecting a lane line edge by using 2-valued processing, Sobel filtering, differential processing and Hough transformation, and evaluating and judging the lane line according to set lane specific attributes to finally judge the lane line. Meanwhile, the luminance threshold value is automatically adjusted by using the color discrimination result of the lane line;
4) judging the lane lines of the two regions: judging whether the output results of the two areas meet the reliability, and pre-judging according to the reliability;
a. if the two areas are judged to pass the reliability, entering a double-area identification logic judgment module:
(1) if the deviation does not exceed the set deviation threshold, judging lane line recognition, and outputting the output result to an intelligent driving system as a reference lane input signal;
(2) if the deviation exceeds the set deviation threshold value, the vehicle enters a continuity tracking module and is continuously compared with the result of the previous n periods, and the area with the larger continuity value is judged to be lane line identification and output to an intelligent driving system;
b. if the detection results of the two areas are not reliable (such as recognition error and obvious noise) the signals are used as input signals of the intelligent driving system which do not meet the automatic driving conditions;
c. if one region passes the reliability judgment and the other region does not pass the reliability judgment, judging that the vehicle enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, if the vehicle is larger than the continuity judgment threshold, judging that the lane line is identified and outputting the lane line to the intelligent driving system, and if the vehicle is smaller than the continuity judgment threshold, judging that the lane line is not identified;
5) adjusting the deviation of the result of the double-area lane line:
when the lane enters a route with larger curvature and the lane line deviates from the detection area of the initial state, the double-area movable processing module can automatically move the detection area in real time and automatically correct the deviation result with the main vehicle, which is generated due to the movement area; wherein the deviation value of the left lane line is E-L, and the deviation value of the right lane line is E-R.
If the deviation value of the left lane line and the deviation value of the right lane line are smaller than the reference threshold value E _ ref, outputting the average deviation value E ═ Ave (E _ L, E _ R);
if the deviation value of the left lane line and the deviation value of the right lane line are greater than or equal to the reference threshold value E _ ref, entering a continuity tracking module;
the continuity tracking module comprehensively judges unlocking lane lines by using a plurality of detection methods, so that the misdetection rate is reduced. That is, the automatic tracking function of the lane line is realized by using the current frame and history frame variation amounts of the two elements of the hough transform and the deviation variation amounts of the host vehicle and the lane line of the current frame and the history frame.
When the deviation value of the left lane line and the deviation value of the right lane line are greater than or equal to the reference threshold value E _ ref, it is determined whether the absolute values of the deviation value of the next frame of the left lane line E _ L (i +1) and the current frame deviation value of the left lane line E _ L (i) are less than the historical deviation threshold value of the left lane line E _ L _ ref.
If the deviation value is smaller than the left lane line historical deviation threshold value, a deviation value Ave (E _ L) is output.
And if the left lane line historical deviation threshold is larger than or equal to the left lane line historical deviation threshold, re-entering the calculation of the left lane line deviation value and the right lane line deviation value of the next frame.
On the basis of lane line identification, the identification result high-precision processing module converts image coordinates into road coordinates, and calculates the deviation between the control point of the host vehicle and a white line to be used as an automatic control quantity. The accuracy of deviation calculation is improved by using result comparison of double regions, reliability judgment and statistical judgment of previous and subsequent results. Finally, the deviation is output as an automatic driving control deviation value after the deviation optimization processing.
Although the embodiments of the present invention and the accompanying drawings are disclosed for illustrative purposes, those skilled in the art will appreciate that: various substitutions, changes and modifications are possible without departing from the spirit and scope of the invention and the appended claims, and therefore the scope of the invention is not limited to the disclosure of the embodiments and the accompanying drawings.
Claims (2)
1. A method for identifying a lane line in a double-area is characterized by comprising the following steps: the identification system adopted by the identification method comprises a double-area lane line identification unit, a double-area lane line judgment unit and a double-area lane line result deviation calculation unit, wherein the double-area lane line identification unit comprises a camera matting module, a double-area dividing module and a lane line edge detection module; the double-region lane line judging unit comprises a reliability pre-judging module and a double-region identification logic judging module; the double-region lane line result deviation calculation unit comprises a double-region movable processing module, a continuity tracking module and an identification result high-precision processing module;
the identification method comprises the following steps:
1) the camera is scratched and is divided into areas: setting a front watching angle and an optimal watching range by using a front watching model, removing unnecessary fields, reducing unnecessary operation burden and determining a region in front of a lane line;
2) dividing a double-area lane line: dividing the area in front of the lane line determined in the step 1) into a left lane line area and a right lane line area;
3) and (3) detecting the edges of the lane lines in the double areas: judging whether the identification object is a white line or a yellow line by using the image pigment component, and then detecting an edge by using 2-valued processing, Sobel filtering, differential processing and Hough transformation;
4) judging the lane lines of the two regions: judging whether the output results of the two areas meet the reliability, and pre-judging according to the reliability;
if the two areas are judged to pass the reliability, entering a double-area identification logic judgment module:
if the deviation does not exceed the set deviation threshold, judging lane line recognition, and outputting the output result to an intelligent driving system as a reference lane input signal;
if the deviation exceeds the set deviation threshold value, the vehicle enters a continuity tracking module and is continuously compared with the result of the previous n periods, and the area with the larger continuity value is judged to be lane line identification and output to an intelligent driving system;
if the detection results of the two areas are not credible, the signal is used as an input signal that the intelligent driving system does not meet the automatic driving condition;
if one region passes the reliability judgment and the other region does not pass the reliability judgment, judging that the vehicle enters the continuity tracking module and is subjected to continuity comparison with the result of the previous n periods, if the vehicle is larger than the continuity judgment threshold, judging that the lane line is identified and outputting the lane line to the intelligent driving system, and if the vehicle is smaller than the continuity judgment threshold, judging that the lane line is not identified;
5) adjusting the deviation of the result of the double-area lane line: when the lane enters a route with larger curvature and the lane line deviates from the detection area of the initial state, the detection area can be automatically moved in real time, meanwhile, the deviation result generated by the movement area and the main vehicle is automatically corrected, and the deviation result is output as an automatic driving control deviation value after the deviation optimization processing.
2. The dual-zone lane line identification method according to claim 1, wherein: and the double-area lane line judging unit receives the information of the double-area lane line identifying unit, and judges that the lane line is not identified when the information of both areas does not pass the credibility prejudgment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810163493.9A CN108182430B (en) | 2018-02-27 | 2018-02-27 | Double-area lane line identification system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810163493.9A CN108182430B (en) | 2018-02-27 | 2018-02-27 | Double-area lane line identification system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108182430A CN108182430A (en) | 2018-06-19 |
| CN108182430B true CN108182430B (en) | 2022-01-14 |
Family
ID=62553239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810163493.9A Active CN108182430B (en) | 2018-02-27 | 2018-02-27 | Double-area lane line identification system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108182430B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109542994B (en) * | 2018-11-19 | 2021-07-06 | 禾多科技（北京）有限公司 | Lane picking algorithm for automatic driving high-precision map |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103802766A (en) * | 2012-11-14 | 2014-05-21 | 现代摩比斯株式会社 | Lane recognition system and method |
| CN104657735A (en) * | 2013-11-21 | 2015-05-27 | 比亚迪股份有限公司 | Lane line detection method and system, as well as lane departure early warning method and system |
| CN106128115A (en) * | 2016-08-01 | 2016-11-16 | 青岛理工大学 | A kind of fusion method based on twin camera detection Traffic Information |
-
2018
- 2018-02-27 CN CN201810163493.9A patent/CN108182430B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103802766A (en) * | 2012-11-14 | 2014-05-21 | 现代摩比斯株式会社 | Lane recognition system and method |
| CN104657735A (en) * | 2013-11-21 | 2015-05-27 | 比亚迪股份有限公司 | Lane line detection method and system, as well as lane departure early warning method and system |
| CN106128115A (en) * | 2016-08-01 | 2016-11-16 | 青岛理工大学 | A kind of fusion method based on twin camera detection Traffic Information |
Non-Patent Citations (2)
| Title |
|---|
| 一种鲁棒性的多车道线检测算法;宣寒宇 等;《计算机科学》;20171130;全文 * |
| 智能车辆结构化道路单目视觉导航技术;王安帅;《中国优秀硕士学位论文全文数据库 信息科技辑》;20081015;摘要，正文第4.5节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108182430A (en) | 2018-06-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101864066B1 (en) | Lane marking detection device, Lane departure determination device, Lane marking detection method and Lane departure determination method | |
| US10891738B2 (en) | Boundary line recognition apparatus and branch road determination apparatus | |
| TWI302879B (en) | Real-time nighttime vehicle detection and recognition system based on computer vision | |
| KR100459476B1 (en) | Apparatus and method for queue length of vehicle to measure | |
| CN110992693B (en) | Deep learning-based traffic congestion degree multi-dimensional analysis method | |
| US20150278615A1 (en) | Vehicle exterior environment recognition device | |
| US9558546B2 (en) | Three-dimensional object detection device, and three-dimensional object detection method | |
| US20130141520A1 (en) | Lane tracking system | |
| Li et al. | Nighttime lane markings recognition based on Canny detection and Hough transform | |
| US8385601B2 (en) | In-vehicle white line recognition apparatus | |
| CN105512720A (en) | Public transport vehicle passenger flow statistical method and system | |
| CN103617412A (en) | Real-time lane line detection method | |
| RU2576315C1 (en) | Device for detecting three-dimensional objects and method of detecting three-dimensional objects | |
| US10726277B2 (en) | Lane line detection method | |
| CN104809449B (en) | Track dotted line line of demarcation automatic testing method suitable for highway video monitoring system | |
| CN105654073A (en) | Automatic speed control method based on visual detection | |
| CN110335467B (en) | Method for realizing highway vehicle behavior detection by using computer vision | |
| JP2020064583A (en) | Vehicle detection method, nighttime vehicle detection method based on dynamic light intensity, and system for the same | |
| WO2014054328A1 (en) | Vehicle detection apparatus | |
| CN104268860A (en) | Lane line detection method | |
| CN107563331A (en) | A kind of road sign line detecting method and system based on geometrical relationship | |
| Wang et al. | Robust lane detection based on gradient-pairs constraint | |
| Jiang et al. | Lane line detection optimization algorithm based on improved Hough transform and R-least squares with dual removal | |
| CN109919062A (en) | A kind of road scene weather recognition methods based on characteristic quantity fusion | |
| CN107037467B (en) | Positioning system and method and intelligent automobile |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20191220 Address after: 361024 unit 64, Room 308, building 1, Yifu Building, No. 600, Ligong Road, Jimei District, Xiamen City, Fujian Province Applicant after: Xiamen Weisen Intelligent Technology Co.,Ltd. Address before: 350317 No. 52, Nan Cun, Gao Dong village, port town, Fuzhou, Fujian Applicant before: Gao Xiujing |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220402 Address after: No.89 Lianjing Road, Siming District, Xiamen City, Fujian Province 361000 Patentee after: Wang Weichang Address before: 361024 unit 64, Room 308, building 1, Yifu Building, No. 600, Ligong Road, Jimei District, Xiamen City, Fujian Province Patentee before: Xiamen Weisen Intelligent Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220801 Address after: 361000 Room 502, No. 78, TIYU Road, Siming District, Xiamen City, Fujian Province Patentee after: Klose (Xiamen) instrument equipment Co.,Ltd. Address before: No.89 Lianjing Road, Siming District, Xiamen City, Fujian Province 361000 Patentee before: Wang Weichang |
|
| TR01 | Transfer of patent right |