CN106529404A - Imaging principle-based recognition method for pilotless automobile to recognize road marker line - Google Patents
Imaging principle-based recognition method for pilotless automobile to recognize road marker line Download PDFInfo
- Publication number
- CN106529404A CN106529404A CN201610872287.6A CN201610872287A CN106529404A CN 106529404 A CN106529404 A CN 106529404A CN 201610872287 A CN201610872287 A CN 201610872287A CN 106529404 A CN106529404 A CN 106529404A
- Authority
- CN
- China
- Prior art keywords
- markings
- region
- picture
- pilotless automobile
- 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
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)
- Image Processing (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to an imaging principle-based recognition method for a pilotless automobile to recognize a road marker line. The method comprises the steps of 1, installing a camera device at the longitudinal center in front of a pilotless automobile, acquiring the information of the current road of the pilotless automobile by the camera device, and generating a road surface information image; 2, acquiring an interested area with a road marker line, and extracting the interested area to form an area image; 3, subjecting the area image to marginalization treatment; 4, subjecting the area image to binarization treatment; 5, identifying the area image through the fitting algorithm; on the condition that the road marker line in the area image is too incomplete and damaged to be identified, sending out the alarm sound through an alarm device by a controller. According to the technical scheme of the invention, the image information is acquired by the camera device, so that the road marker line can be effectively identified. Compared with the scanning identification effect of a single-line laser scanning radar, the effect of the above method is more accurate. The occurrence of potential safety hazards of the pilotless automobile during the driving process is reduced.
Description
Technical field
The present invention relates to pilotless automobile technical field, more particularly, to a kind of unmanned vapour based on shooting principle
The recognition methodss of car road pavement markings.
Background technology
Automatic driving vehicle is also referred to as robotic vehicle in the world, belongs to one kind of outdoor mobile robot, is one
Integrate environment sensing, planning and the integrated intelligent system of the multiple function such as decision-making, control, cover machinery, control, sense
The multi-subject knowledges such as device technology, signal processing, pattern recognition, artificial intelligence and computer technology.Pilotless automobile is also weighing apparatus
One of important symbol of the national research strength of amount one and industrial level, has wide application in terms of national defence and public's traffic
Prospect.Automatic driving vehicle originates from military motion, and from the beginning of 20 century 70s, American and Britain, De Deng developed countries begin to
The research of automatic driving vehicle, through development so for many years, automatic driving vehicle also gradually develops into civilian from military field
Field.It is to need to be identified the markings on highway in civil area stage most outstanding feature, using shooting principle
Pilotless automobile the advantages of specifically put into low, simple compared to laser scanning surface radar, can be dropped in terms of environment sensing
Low production cost.
Pilotless automobile is to mark to the main means that markings are recognized by single line laser radar at present both at home and abroad
Will line is scanned, this for normal markings have an accuracy and reliability, but for some are incomplete or fuzzy
Markings, tend not to identification, also result in automobile erroneous judgement, be likely to result in security incident.
The content of the invention
The technical problem to be solved in the present invention is:A kind of pavement marker of the pilotless automobile based on shooting principle is provided
Line recognition methodss, can effectively recognize pavement marker line.
The technical solution adopted for the present invention to solve the technical problems is:A kind of pilotless automobile based on shooting principle
The recognition methodss of road pavement markings, comprise the steps:
Step 1:Camera head is installed at the Herba Plantaginis longitudinal center of pilotless automobile, car is obtained by the camera head
The information of road surface being presently in, and generate information of road surface image;
Step 2:The pavement image information generated to upper step, area-of-interest of the collection with pavement marker line, extracts sense
Interest region forming region picture;
Step 3:Marginalisation process is carried out to region picture;
Step 4:Binary conversion treatment is carried out to region picture;
Step 5:Lane detection:Region picture is identified by fitting algorithm;Markings in the picture of region
When incomplete and breakage can not be recognized, controller sends alarm sound by alarm device.
Further, the marginalisation processing method is:
(1) pavement image f (x, y) of the determination size for M × N in the picture of region;
(2) in coordinate figure f (i, j) of its collected any pixel, the little adjacent block centered on the pixel be designated as B ×
B, i.e. primitive blocks;
(3) take the region unit centered on primitive blocks and be designated as D × D;
(4) seek the meansquaredeviationσ of primitive blocks B × BB, formula is as follows;
Wherein:μ is arithmetic average;
(5) work as meansquaredeviationσBIt is less than threshold value T1 of setting then to export edge image g (i, j)=0 and terminate, otherwise carry out down
One step;
(6) similarity coefficient and correlation coefficient of region picture is obtained with equation below;
Wherein:S is similarity coefficient, and R is correlation coefficient;
biFor the gray value of B × B all pixels point;
diFor the gray value of D × D all pixels point;
(7) if similarity coefficient and correlation coefficient export edge image g (i, j)=0 and simultaneously terminate less than threshold value T2 of setting,
G (i, j)=1 is exported otherwise and to next step;
(8) return to step (2), continue to judge the picture point of region picture, until institute is somewhat complete in the picture of region
Portion has judged.
Further, region picture is identified comprising the steps by fitting algorithm:
(1) set up plane of delineation coordinate system, demarcate top left co-ordinate origin O, horizontal direction is X-direction, vertical direction
For Y direction;
(2) model of left and right road mark line is, equation below:
yl=k_left × xl+b_left;
yr=k_riht × xr+b_right;
Wherein:xl, yl, xr, yrHorizontal stroke, the longitudinal coordinate of left and right road is represented respectively;
K_left, k_right represent the slope of left and right road mark line respectively;
B_left, b_right represent the intercept of left and right road mark line respectively;
(3) above-mentioned formula is converted to l=xcos θ+ysin θ, so as to any bar straight line of X-Y plane is transformed to correspondence
One point in l- θ spaces;
(4) l- θ spatial spreadings are turned to into grid, the centrifugal pump of θ is brought into by each (x, y) point, and obtains each l value;
(5) a little, the big grid of value of calculation corresponds to collinear points, the fitting parameter of its (l, θ) as straight line to statistics;
(6) Straight Line Fitting Parameters in l- θ spaces are converted to the parameter in rectangular coordinate, determine the shape dough-making powder of markings
Product;
(7) after the shape of markings and area recognition, judge that can automobile recognize the markings, if markings energy
Enough to recognize, then vehicle advances according to the markings type of identification, if automobile can not recognize the markings, controller is by reporting to the police
Device sends alarm sound.
The invention has the beneficial effects as follows:The present invention obtains image information using camera head, can effectively recognize pavement marker
Line, recognizes more accurately than single line laser surface sweeping radar scanning, reduces automatic driving vehicle safety in the process of moving hidden
Suffer from.
In addition, present invention improves over marginalisation process and lane mark identification algorithm, binary conversion treatment is by conventional two-value
Change processing means, region picture is processed, vehicle safety coefficient in the process of moving is improve.
Description of the drawings
The present invention is further described below in conjunction with the accompanying drawings.
Fig. 1 is the overall flow figure of the present invention;
Fig. 2 is the marginalisation Processing Algorithm flow chart of the present invention;
Fig. 3 is the fast schematic diagram in the primitive blocks in the present invention and region;
Fig. 4 is the plane coordinate system figure of the present invention;
Specific embodiment
Presently in connection with accompanying drawing, the present invention is further illustrated.These accompanying drawings are simplified schematic diagram only with signal side
The basic structure of the formula explanation present invention, therefore which only shows the composition relevant with the present invention.
As shown in Figure 1, Figure 2, Figure 3, Figure 4, a kind of knowledge of the pilotless automobile road pavement markings based on shooting principle
Other method, comprises the steps:
Step 1:Camera head is installed at the Herba Plantaginis longitudinal center of pilotless automobile, car is obtained by the camera head
The information of road surface being presently in, and generate information of road surface image;
Step 2:The pavement image information generated to upper step, area-of-interest of the collection with pavement marker line, extracts sense
Interest region forming region picture;
Step 3:Marginalisation process is carried out to region picture;
Step 4:Binary conversion treatment is carried out to region picture;
Step 5:Lane detection:Region picture is identified by fitting algorithm;Markings in the picture of region
When incomplete and breakage can not be recognized, controller sends alarm sound by alarm device.
Further, the marginalisation processing method is:
(1) pavement image f (x, y) of the determination size for M × N in the picture of region;
(2) in coordinate figure f (i, j) of its collected any pixel, the little adjacent block centered on the pixel be designated as B ×
B, i.e. primitive blocks;
(3) take the region unit centered on primitive blocks and be designated as D × D;
(4) seek the meansquaredeviationσ of primitive blocks B × BB, formula is as follows;
Wherein:μ is arithmetic average;
(5) work as meansquaredeviationσBIt is less than threshold value T1 of setting then to export edge image g (i, j)=0 and terminate, otherwise carry out down
One step;
(6) similarity coefficient and correlation coefficient of region picture is obtained with equation below;
Wherein:S is similarity coefficient, and R is correlation coefficient;
biFor the gray value of B × B all pixels point;
diFor the gray value of D × D all pixels point;
(7) if similarity coefficient and correlation coefficient export edge image g (i, j)=0 and simultaneously terminate less than threshold value T2 of setting,
G (i, j)=1 is exported otherwise and to next step;
(8) return to step (2), continue to judge the picture point of region picture, until institute is somewhat complete in the picture of region
Portion has judged.
Further, region picture is identified comprising the steps by fitting algorithm:
(1) set up plane of delineation coordinate system, demarcate top left co-ordinate origin O, horizontal direction is X-direction, vertical direction
For Y direction;
(2) model of left and right road mark line is, equation below:
yl=k_left × xl+b_left;
yr=k_riht × xr+b_right;
Wherein:xl, yl, xr, yrHorizontal stroke, the longitudinal coordinate of left and right road is represented respectively;
K_left, k_right represent the slope of left and right road mark line respectively;
B_left, b_right represent the intercept of left and right road mark line respectively;
(3) above-mentioned formula is converted to l=xcos θ+ysin θ, so as to any bar straight line of X-Y plane is transformed to correspondence
One point in l- θ spaces;
(4) l- θ spatial spreadings are turned to into grid, the centrifugal pump of θ is brought into by each (x, y) point, and obtains each l value;
(5) a little, the big grid of value of calculation corresponds to collinear points, the fitting parameter of its (l, θ) as straight line to statistics;
(6) Straight Line Fitting Parameters in l- θ spaces are converted to the parameter in rectangular coordinate, determine the shape dough-making powder of markings
Product;
(7) after the shape of markings and area recognition, judge that can automobile recognize the markings, if markings energy
Enough to recognize, then vehicle advances according to the markings type of identification, if automobile can not recognize the markings, controller is by reporting to the police
Device sends alarm sound.
The present invention obtains image information using camera head, can effectively recognize pavement marker line, than single line laser surface sweeping thunder
It is more accurate up to scanning recognition, reduce automatic driving vehicle potential safety hazard in the process of moving.
In addition, present invention improves over marginalisation process and lane mark identification algorithm, binary conversion treatment is by conventional two-value
Change processing means, region picture is processed, vehicle safety coefficient in the process of moving is improve.Camera head can be adopted
High-speed area array camera.
With the above-mentioned desirable embodiment according to the present invention as enlightenment, by above-mentioned description, relevant staff is complete
Various change and modification can be carried out in the range of without departing from this invention technological thought entirely.The technology of this invention
Property scope is not limited to the content in description, it is necessary to determine its technical scope according to right.
Claims (3)
1. a kind of recognition methodss of the pilotless automobile road pavement markings based on shooting principle, it is characterised in that:Including such as
Lower step:
Step 1:Camera head is installed at the Herba Plantaginis longitudinal center of pilotless automobile, vehicle is obtained by the camera head and is worked as
Front residing information of road surface, and generate information of road surface image;
Step 2:The pavement image information generated to upper step, area-of-interest of the collection with pavement marker line, extracts interested
Region forming region picture;
Step 3:Marginalisation process is carried out to region picture;
Step 4:Binary conversion treatment is carried out to region picture;
Step 5:Lane detection:Region picture is identified by fitting algorithm;When the markings incompleteness in the picture of region
And breakage, when can not recognize, controller sends alarm sound by alarm device.
2. the recognition methodss of the pilotless automobile road pavement markings based on shooting principle according to claim 1, its
It is characterised by:The marginalisation processing method is:
(1) pavement image f (x, y) of the determination size for M × N in the picture of region;
(2) in coordinate figure f (i, j) of its collected any pixel, the little adjacent block centered on the pixel is designated as B × B, i.e.,
Primitive blocks;
(3) take the region unit centered on primitive blocks and be designated as D × D;
(4) seek the meansquaredeviationσ of primitive blocks B × BB, formula is as follows;
Wherein:μ is arithmetic average;
(5) work as meansquaredeviationσBIt is less than threshold value T1 of setting then to export edge image g (i, j)=0 and terminate, otherwise carry out next step;
(6) similarity coefficient and correlation coefficient of region picture is obtained with equation below;
Wherein:S is similarity coefficient, and R is correlation coefficient;
biFor the gray value of B × B all pixels point;
diFor the gray value of D × D all pixels point;
(7) if similarity coefficient and correlation coefficient export edge image g (i, j)=0 and simultaneously terminate, otherwise less than threshold value T2 of setting
Export g (i, j)=1 and to next step;
(8) return to step (2), continue to judge the picture point of region picture, until institute is a little all sentenced in the picture of region
Break.
3. the recognition methodss of the pilotless automobile road pavement markings based on shooting principle according to claim 1, its
It is characterised by:Region picture is identified comprising the steps by fitting algorithm:
(1) plane of delineation coordinate system is set up, top left co-ordinate origin O is demarcated, horizontal direction is X-direction, and vertical direction is Y-axis
Direction;
(2) model of left and right road mark line is, equation below:
yl=k_left × xl+b_left;
yr=k_riht × xr+b_right;
Wherein:xl, yl, xr, yrHorizontal stroke, the longitudinal coordinate of left and right road is represented respectively;
K_left, k_right represent the slope of left and right road mark line respectively;
B_left, b_right represent the intercept of left and right road mark line respectively;
(3) above-mentioned formula is converted to l=xcos θ+ysin θ, so as to any bar straight line of X-Y plane is transformed to correspondence l- θ
One point in space;
(4) l- θ spatial spreadings are turned to into grid, the centrifugal pump of θ is brought into by each (x, y) point, and obtains each l value;
(5) a little, the big grid of value of calculation corresponds to collinear points, the fitting parameter of its (l, θ) as straight line to statistics;
(6) Straight Line Fitting Parameters in l- θ spaces are converted to the parameter in rectangular coordinate, determine shape and the area of markings;
(7) after the shape of markings and area recognition, judge that can automobile recognize the markings, if markings can be known
Not, then vehicle advances according to the markings type of identification, if automobile can not recognize the markings, controller passes through alarm device
Send alarm sound.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610872287.6A CN106529404A (en) | 2016-09-30 | 2016-09-30 | Imaging principle-based recognition method for pilotless automobile to recognize road marker line |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610872287.6A CN106529404A (en) | 2016-09-30 | 2016-09-30 | Imaging principle-based recognition method for pilotless automobile to recognize road marker line |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106529404A true CN106529404A (en) | 2017-03-22 |
Family
ID=58331304
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610872287.6A Pending CN106529404A (en) | 2016-09-30 | 2016-09-30 | Imaging principle-based recognition method for pilotless automobile to recognize road marker line |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106529404A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107895147A (en) * | 2017-11-07 | 2018-04-10 | 龚土婷 | A kind of safe pilotless automobile system |
CN108068817A (en) * | 2017-12-06 | 2018-05-25 | 张家港天筑基业仪器设备有限公司 | A kind of automatic lane change device and method of pilotless automobile |
CN108108703A (en) * | 2017-12-27 | 2018-06-01 | 天津英创汇智汽车技术有限公司 | Deceleration strip missing detection method, device and electronic equipment |
CN108918532A (en) * | 2018-06-15 | 2018-11-30 | 长安大学 | A kind of through street traffic sign breakage detection system and its detection method |
CN109522804A (en) * | 2018-10-18 | 2019-03-26 | 汽-大众汽车有限公司 | A kind of road edge recognition methods and system |
CN112092827A (en) * | 2020-09-23 | 2020-12-18 | 北京百度网讯科技有限公司 | Automatic driving function control method, automatic driving function control device, electronic equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201427553Y (en) * | 2009-06-26 | 2010-03-24 | 长安大学 | Alarm system of vehicle departure from lane |
US20110216938A1 (en) * | 2010-03-03 | 2011-09-08 | Denso Corporation | Apparatus for detecting lane-marking on road |
CN202134079U (en) * | 2011-06-16 | 2012-02-01 | 长安大学 | Unmanned vehicle lane marker line identification and alarm device |
CN203305895U (en) * | 2012-10-12 | 2013-11-27 | 天津图科科技研发有限公司 | System capable of giving alarm when vehicle deviates from lane based on image recognition |
CN104149783A (en) * | 2014-08-27 | 2014-11-19 | 刘红华 | Digital road and self-driving vehicle thereof |
-
2016
- 2016-09-30 CN CN201610872287.6A patent/CN106529404A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201427553Y (en) * | 2009-06-26 | 2010-03-24 | 长安大学 | Alarm system of vehicle departure from lane |
US20110216938A1 (en) * | 2010-03-03 | 2011-09-08 | Denso Corporation | Apparatus for detecting lane-marking on road |
CN202134079U (en) * | 2011-06-16 | 2012-02-01 | 长安大学 | Unmanned vehicle lane marker line identification and alarm device |
CN203305895U (en) * | 2012-10-12 | 2013-11-27 | 天津图科科技研发有限公司 | System capable of giving alarm when vehicle deviates from lane based on image recognition |
CN104149783A (en) * | 2014-08-27 | 2014-11-19 | 刘红华 | Digital road and self-driving vehicle thereof |
Non-Patent Citations (1)
Title |
---|
王秀梅: "基于分形路面破损图像裂纹识别研究", 《中国优秀硕士学位论文全文数据库 工程科技辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107895147A (en) * | 2017-11-07 | 2018-04-10 | 龚土婷 | A kind of safe pilotless automobile system |
CN108068817A (en) * | 2017-12-06 | 2018-05-25 | 张家港天筑基业仪器设备有限公司 | A kind of automatic lane change device and method of pilotless automobile |
CN108108703A (en) * | 2017-12-27 | 2018-06-01 | 天津英创汇智汽车技术有限公司 | Deceleration strip missing detection method, device and electronic equipment |
CN108918532A (en) * | 2018-06-15 | 2018-11-30 | 长安大学 | A kind of through street traffic sign breakage detection system and its detection method |
CN108918532B (en) * | 2018-06-15 | 2021-06-11 | 长安大学 | System and method for detecting damage of expressway traffic sign |
CN109522804A (en) * | 2018-10-18 | 2019-03-26 | 汽-大众汽车有限公司 | A kind of road edge recognition methods and system |
CN109522804B (en) * | 2018-10-18 | 2020-11-06 | 一汽-大众汽车有限公司 | Road edge identification method and system |
CN112092827A (en) * | 2020-09-23 | 2020-12-18 | 北京百度网讯科技有限公司 | Automatic driving function control method, automatic driving function control device, electronic equipment and storage medium |
CN112092827B (en) * | 2020-09-23 | 2022-04-22 | 北京百度网讯科技有限公司 | Automatic driving function control method, automatic driving function control device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106529404A (en) | Imaging principle-based recognition method for pilotless automobile to recognize road marker line | |
CN109460709B (en) | RTG visual barrier detection method based on RGB and D information fusion | |
CN105260699B (en) | A kind of processing method and processing device of lane line data | |
CN102779280B (en) | Traffic information extraction method based on laser sensor | |
CN100403332C (en) | Vehicle lane Robust identifying method for lane deviation warning | |
CN110862033B (en) | Intelligent early warning detection method applied to inclined shaft winch of coal mine | |
CN103714538B (en) | road edge detection method, device and vehicle | |
CN110379168B (en) | Traffic vehicle information acquisition method based on Mask R-CNN | |
CN106127113A (en) | A kind of road track line detecting method based on three-dimensional laser radar | |
CN105488454A (en) | Monocular vision based front vehicle detection and ranging method | |
CN103034843B (en) | Method for detecting vehicle at night based on monocular vision | |
CN102831618A (en) | Hough forest-based video target tracking method | |
CN111369541A (en) | Vehicle detection method for intelligent automobile under severe weather condition | |
CN104134209A (en) | Feature extraction and matching method and feature extraction and matching system in visual navigation | |
CN108717540A (en) | The method and device of pedestrian and vehicle are distinguished based on 2D laser radars | |
CN106683530A (en) | Computerized judging system and method based on three-dimensional laser vision and high-precision lane model | |
CN110188606B (en) | Lane recognition method and device based on hyperspectral imaging and electronic equipment | |
CN102768726A (en) | Pedestrian detection method for preventing pedestrian collision | |
CN109272482A (en) | A kind of urban road crossing vehicle queue detection system based on sequence image | |
CN202134079U (en) | Unmanned vehicle lane marker line identification and alarm device | |
Zhang et al. | Vehicle detection method for intelligent vehicle at night time based on video and laser information | |
Li et al. | Judgment and optimization of video image recognition in obstacle detection in intelligent vehicle | |
CN114530042A (en) | Urban traffic brain monitoring system based on internet of things technology | |
CN106960193A (en) | A kind of lane detection apparatus and method | |
Zhou et al. | Real-time traffic light recognition based on c-hog features |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | 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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170322 |