CN106525056A - Method for lane line detection by gyro sensor - Google Patents
Method for lane line detection by gyro sensor Download PDFInfo
- Publication number
- CN106525056A CN106525056A CN201610970334.0A CN201610970334A CN106525056A CN 106525056 A CN106525056 A CN 106525056A CN 201610970334 A CN201610970334 A CN 201610970334A CN 106525056 A CN106525056 A CN 106525056A
- Authority
- CN
- China
- Prior art keywords
- lane line
- lane
- line
- detection
- gyro sensor
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 80
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000003993 interaction Effects 0.000 claims abstract description 10
- 238000012552 review Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 8
- 238000013467 fragmentation Methods 0.000 claims description 7
- 238000006062 fragmentation reaction Methods 0.000 claims description 7
- 230000009466 transformation Effects 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims description 3
- 238000013519 translation Methods 0.000 claims description 3
- 239000007787 solid Substances 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 2
- 238000012544 monitoring process Methods 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 4
- 239000000047 product Substances 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000000205 computational method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 230000004888 barrier function Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/36—Input/output arrangements for on-board computers
- G01C21/3626—Details of the output of route guidance instructions
- G01C21/3658—Lane guidance
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Navigation (AREA)
Abstract
The invention belongs to the field of computer vision and intelligent transportation, particularly relates to a method for lane line detection by a gyro sensor and solves problems of poor accuracy in existing lane line calibration and the like. The method includes: A, calibration scene selection, to be more specific, selecting a lane with a left lane line and a right lane line as a calibration scene, and parking or driving a current vehicle between the left lane line and the right lane line of the lane; B, man-machine interaction calibration, to be more specific, respectively working out linear equations of the left lane line and the right lane line in an image coordinate system at the moment; C, vehicular body attitude angle detection, to be more specific, monitoring real-time attitudes of the current vehicle and acquiring real-time attitude data; D, real-time camera angle adjustment, to be more specific, guaranteeing that the shot left lane line and the right lane line are in fixed areas constantly; E, lane line detection, to be more specific, detecting the left lane line and the right lane line in the fixed areas. The method has advantages of high lane line detection accuracy, convenience and quickness in man-machine interaction and great reduction of clutter influences on lane line detection.
Description
Technical field
The invention belongs to computer vision and intelligent transportation field, more particularly to a kind of to enter driving using gyro sensor
The method of diatom detection.
Background technology
In recent years, as the technology in terms of intelligent transportation field and computer vision constantly develops, based on computer vision
Automobile active safety product also day by day enrich.Such as adaptive cruise control system (ACC), lane departur warning (LDW), car
Road keeps auxiliary (LKA) and side object detection (SOD) etc., but these product correlation techniques are realized in the detection of lane line
Basis.
The common method for detecting lane lines based on camera is usually first to carry out pretreatment to image, then substantially determines car
Diatom in-scope, recycles Hough transformation to carry out lane detection, but picture quality is bad, the lane line for detecting
Miscellaneous line is more, and especially the slight crack on the rainy day wheel marking or road can all be misidentified as lane line, or run into barrier
Hinder thing closely to block, when the surface conditions such as blocking of trees or building effects, as camera cannot be collected comprising foot
The image of more than enough lane information and the problems such as there is missing inspection, flase drop, the identification of the lane line for being thus is not accurate enough.Therefore,
The good high robust algorithm of set of environmental strong adaptability, fast operation, transplantability how is designed, is the heat of association area research
One of point.Meanwhile, although the more existing product function admirable with regard to lane detection, makes as algorithm is relative complex
Early stage calibration process complexity is obtained, the Professional knowledge requirement to assembler is of a relatively high, so research and development are a set of beneficial to industrial products
It is also what is be highly desirable to that the algorithm and raising lane detection ground of assembling are accurate.
In order to solve the problems, such as prior art, people have carried out long-term exploration, it is proposed that solution miscellaneous
Scheme.For example, Chinese patent literature discloses a kind of method for detecting lane lines and device [application number：
201310616153.4], the method includes obtaining vehicle front image；Rim detection is carried out to vehicle front image, side is obtained
Edge figure；The position of the lane line obtained according to former frame determines original lane line detection zone in edge graph；By original lane
Line detection zone is divided into N number of little detection zone, and N is more than or equal to 2；According to the edge graph numerical value of each little detection zone and each
The sum of the pixel of little detection zone, determines that certain little detection zone is accurate lane detection region；From accurate lane line
The lane line of present frame is obtained in detection zone.It is not accurate enough that such scheme solves existing lane detection to a certain extent
Problem, but the program still fundamentally cannot solve computationally intensive, subtract the problem of impact of the miscellaneous line to testing result.
The content of the invention
The purpose of the present invention is for the problems referred to above, there is provided a kind of simple and reasonable, high-precision utilization gyroscope is passed
The method that sensor carries out lane detection.
To reach above-mentioned purpose, following technical proposal is present invention employs：This determines lane line using gyro sensor
The method of detection effective range, it is characterised in that this method comprises the steps：
Scene is demarcated in A, selection:Selection is decorated with the track of left-lane line and right-lane line to demarcate scene, by Current vehicle
Stop or travel between the left-lane line and right-lane line in place track, and guarantee the left-lane in Current vehicle place track
Line is visible in camera review with right-lane line；
B, man-machine interaction are demarcated：Default left demarcation straight line of the mobile display in camera review is so that left demarcation straight line
Overlap with the tangent line of left-lane line or left-lane line so as to realize that left-lane line is demarcated, mobile display is pre- in camera review
If right demarcation straight line so that right demarcation straight line is overlapped with the tangent line of right-lane line or right-lane line so as to realize right-lane line mark
It is fixed, now left-lane line and linear equation of the right-lane line in image coordinate system are calculated respectively；
C, detection body gesture angle:The real-time of Current vehicle is monitored by the gyro sensor on Current vehicle
Attitude simultaneously gathers real-time attitude data, three axles of gyro sensor correspond respectively to Current vehicle before and after axis, on
Lower axis and left and right axis；
D, real-time adjustment camera angle：The real-time attitude data that control unit is gathered according to gyro sensor, that is, bow
Angle of elevation alpha, angle of inclination beta, yaw angle γ, in real time to the rightabout adjustment video camera of angle of pitch α, angle of inclination beta, yaw angle γ, with
Keep video camera all the time parallel to horizontal dead ahead, it is ensured that video camera shoot left-lane line fixed with right-lane line all the time
In region；
E, lane detection：Lane detection is carried out using lane detection algorithm in fixed area, so as to detect
Left-lane line and right-lane line.
In the present invention, Current vehicle should travel the centre in place track as far as possible, currently also be not excluded for certain limit
Deviation, it is to be decorated with the highway of left and right lane line to demarcate scene, without the need for reconfiguring demarcation scene again, such as setting-out on the ground
Demarcate etc., it is so more simple and practical than traditional method, it is cost-effective；Fragmentation thresholdization can be reduced because shade or obstacle
Thing etc. blocks the inaccurate problem of lane detection that lane line is caused；Man-machine interaction is convenient and swift, to operator require compared with
It is low, therefore changeable scene is adapted to, it is workable；Using the scope of lane line fluctuation in front and back's two field pictures, sense is set
Emerging region can greatly reduce impact of the clutter to lane detection, and improve the execution efficiency of algorithm；On-line proving is supported, can
The a series of activities for gathering video image and processing is needed to remove off-line calibration from, directly demarcate the slope and intercept of lane line,
It is one simply and easily to demarcate very much and computational methods, a large amount of valuable times of developer, and human and material resources can be saved
Etc. aspect resource consumption.
Carry out in the method for lane detection in above-mentioned utilization gyro sensor, with camera optical axis and video camera figure
The intersection point of image plane is origin, and using origin, laterally bearing of trend, as x-axis bearing of trend, is made with origin longitudinally downward direction to the right
For y-axis bearing of trend, image coordinate system is set up, linear equation of the left-lane line in image coordinate system is y=k after demarcationL·x
+bL, wherein kLIt is the slope of left-lane line, bLIt is left-lane line intercept on the y axis, after demarcation, right-lane line is in image coordinate
Linear equation in system is y=kR·x+bR, wherein kRIt is the slope of right-lane line, bRIt is right-lane line intercept on the y axis,
The linear equation of the front bumper place straight line of Current vehicle is y=c；Described fixed area be by left-lane line after demarcating,
The delta-shaped region that the front bumper place straight line of right-lane line and Current vehicle surrounds after demarcation.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, described demarcation rear left
After lane line and demarcation, the intersecting point coordinate of right-lane line is：
Described fixed area (D) is：
Wherein W is picture traverse；
The final carriage matrix of Current vehicle is：
Wherein, R (x, α) is spin matrix of the gyro sensor around axis before and after Current vehicle, and R (y, β) is top
Spin matrix of the spiral shell instrument around the axis up and down of Current vehicle, R (z, γ) are gyroscopes around the left and right axis of Current vehicle
Spin matrix.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, in above-mentioned steps E,
Described lane detection algorithm includes pretreatment image and detection lane line undated parameter；Described pretreatment image includes
By camera acquisition to present frame camera review or current fixed area image be converted into gray-scale maps, and the left car that will be demarcated
Then diatom neighborhood and right-lane line field is done at fragmentation threshold to area-of-interest respectively respectively as area-of-interest
Reason；Simultaneously undated parameter includes detecting lane line respectively in area-of-interest described detection lane line, if detecting new left car
Diatom then updates the linear equation slope and/or intercept of left-lane line, otherwise retains the value at a moment thereon；If detecting new
Right-lane line then updates the linear equation slope and/or intercept of right-lane line, otherwise retains the value at a moment thereon.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, in above-mentioned steps B,
Manually operating-controlling mechanism control is left demarcates rectilinear translation and/or rotation to adjust the left intercept and/or slope for demarcating straight line, leads to
Cross the right intercept and/or slope for rectilinear translation and/or rotation being demarcated to adjust right demarcation straight line of manual operating-controlling mechanism control；It is described
Manual manipulation mechanism include button and/or knob, or described manual manipulation mechanism includes showing camera review
Touch screen.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, in above-mentioned steps C,
If the maximum magnitude of the slope fluctuation of left-lane line is (- Δ kL,ΔkL), the maximum magnitude of intercept fluctuation is (- Δ bL,ΔbL),
The maximum magnitude of the slope fluctuation of right-lane line is (- Δ kR,ΔkR), the maximum magnitude of intercept fluctuation is (- Δ bR,ΔbR), choosing
Take left-lane line slope and maximum contiguous range is fluctuated as the area-of-interest of left-lane line with intercept, choose right-lane line oblique
Rate fluctuates maximum contiguous range as the area-of-interest of right-lane line with intercept.Be arranged on slope fluctuation range mainly by
In vehicle, photographic head has a little shake, or other some correlative factors in the process of moving, under such as varying environment
Track line width and track spacing are different so that the lane line in camera views can fluctuate near the position of previous moment.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, in above-mentioned steps C,
Respectively by area-of-interest homogenous segmentations or uneven segmentation from top to bottom, per section of height is set to Δ hi；Calculate respectively each
The average Δ h of Δ hi in section area-of-interest, and as the threshold value of this section, it is then by area-of-interest binaryzation, adjacent
When the threshold value of two sections of area-of-interests is different, the marginal slope of adjacent two sections of area-of-interests is 0.Can be with homogenous segmentations
Can be stated using homogenous segmentations with uneven segmentation, for convenience herein, be segmented more in theory, testing result is more accurate, but
The problems such as allowing for the real-time in practical application, institute's segments should not be too many, should be appropriate according to practical situation.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, in above-mentioned steps A,
Described left-lane line and right-lane line are respectively the combination of any one or more in solid line, dotted line, yellow line and white line；When
Vehicle in front is stopped or travels the centre between the left-lane line and right-lane line in place track.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, existed using Hough transformation
Lane line is detected respectively in area-of-interest, and the linear equation slope of left-lane line is updated if new left-lane line is detected
And/or intercept, otherwise retain the value at a moment thereon；The linear equation of right-lane line is updated if new right-lane line is detected
Slope and/or intercept, otherwise retain the value at a moment thereon；If the situation of change lane occurs in vehicle, vehicle is waited to enter
During central authorities' traveling in next track, the new lane line of automatic detection.If bend, then the method detection is cutting for bend
Line.
In above-mentioned utilization gyro sensor determines the method for lane detection effective range, gyro sensor case
On meter panel of motor vehicle, and gather 3 current attitude angle of vehicle body, the respectively angle of pitch, inclination angle and yaw angle.
Compared with prior art, this determines the advantage of the method for lane detection effective range using gyro sensor
It is：
1st, amount of calculation can be reduced with the emerging region of downscaled images sense, reduces the lane detection time, improve algorithm real-time,
Many impacts of the miscellaneous line to testing result, the guardrail of such as both sides of the road, the flower bed edge of both sides of the road, rainy day can also be excluded
The wheel marking, road slight crack etc..
2nd, it is to be decorated with the highway of left and right lane line to demarcate scene, without the need for reconfiguring demarcation scene again, such as on the ground
Setting-out demarcation etc., it is so more simple and practical than traditional method, it is cost-effective；
3rd, fragmentation thresholdization can be reduced because shade or barrier etc. block lane detection that lane line causes not
Accurate problem；
4th, man-machine interaction is convenient and swift, operator is required relatively low, therefore is adapted to changeable scene, operability
By force；
5th, the emerging region of sense is set using the scope of lane line fluctuation in two field pictures in front and back clutter can be greatly reduced to car
The impact of diatom detection, and improve the execution efficiency of algorithm；
6th, on-line proving is supported, a series of activities that off-line calibration needs to gather video image and process can be removed from；
7th, the slope and intercept of lane line are directly demarcated, is one and is simply and easily demarcated very much and computational methods, can be with
Save a large amount of valuable times of developer, and the resource consumption of the aspect such as human and material resources.
Description of the drawings
Fig. 1 is the flow chart of the present invention.
Fig. 2 is the lane detection flow chart of the present invention.
Fig. 3 is by the actual calibrated camera views figure of man-machine interaction.
Fig. 4 is image coordinate system schematic diagram.
Fig. 5 is the emerging administrative division map of sense according to slope with the maximum contiguous range selection of intercept fluctuation.
Fig. 6 is the emerging area segmentation figure of sense.
In figure, wherein Δ kLRepresent the difference of red straight line and two straight slopes of left-lane line；ΔbLRepresent blue straight line with
The difference of two Linear intercepts of left-lane line；ΔkRRepresent the difference of red straight line and two straight slopes of right-lane line；ΔbRRepresent blue
The difference of color straight line and two Linear intercepts of right-lane line；Region in picture between the blue line in two, left side is left-lane Gan Xing areas
Domain ROI；Region between the blue line in two, right side is the emerging region ROI of right lane sense；Δ h represents each section of height；1 represents left
Lane line；11 represent left demarcation straight line；2 represent right-lane line；21 represent right demarcation straight line.
Specific embodiment
The present invention will be further described in detail with reference to the accompanying drawings and detailed description.
As shown in figures 1 to 6, the method that this determines lane detection effective range using gyro sensor, including following steps
Suddenly：
Scene is demarcated in step A, selection:Selection is decorated with the track of left-lane line 1 and right-lane line 2 to demarcate scene, ought
Vehicle in front is stopped or is travelled between the left-lane line 1 and right-lane line 2 in place track, and guarantees Current vehicle place track
Left-lane line 1 it is visible in camera review with right-lane line 2；Here left-lane line 1 and right-lane line 2 is respectively real
The combination of any one or more in line, dotted line, yellow line and white line；Current vehicle should be stopped as far as possible or be travelled and be located
Centre between the left-lane line 1 and right-lane line 2 in track, has currently also been not excluded for a range of deviation.
Step B, man-machine interaction are demarcated：Default left demarcation straight line 11 of the mobile display in camera review is so that left mark
Boning out 11 is overlapped with the tangent line of left-lane line 1 or left-lane line 1 so as to realize that left-lane line 1 is demarcated, and mobile display is in shooting
Default right demarcation straight line 21 in machine image is so that right demarcation straight line 21 is overlapped with the tangent line of right-lane line 2 or right-lane line 2
So as to realize that right-lane line 2 is demarcated, now left-lane line 1 and straight line of the right-lane line 2 in image coordinate system are calculated respectively
Equation；In above-mentioned steps B, the intersection point with camera optical axis and camera image plane is laterally prolonged to the right as origin with origin
Direction is stretched as x-axis bearing of trend, using origin longitudinally downward direction as y-axis bearing of trend, set up image coordinate system, after demarcation
Linear equation of the left-lane line 1 in image coordinate system is y=kL·x+bL, wherein kLIt is the slope of left-lane line, bLIt is left car
Diatom intercept on the y axis, after demarcation, linear equation of the right-lane line 2 in image coordinate system is y=kR·x+bR, wherein kR
It is the slope of right-lane line, bRIt is right-lane line intercept on the y axis, the straight line side of the front bumper place straight line of Current vehicle
Journey is y=c；Fixed area is that the front bumper of right-lane line 2 and Current vehicle is located by left-lane line after demarcating 1, after demarcating
The delta-shaped region that straight line surrounds；I.e. manually the left demarcation straight line 11 of operating-controlling mechanism control translates and/or rotates to adjust
It is whole it is left demarcate straight line 11 intercept and/or slope, manually operating-controlling mechanism control it is right demarcation straight line 21 translate and/or rotate with
The right intercept and/or slope for demarcating straight line 21 of adjustment, manual manipulation mechanism here include button and/or knob, manual manipulation
Mechanism includes the touch screen that can show camera review.
Step C, detection body gesture angle：Current vehicle is monitored by the gyro sensor on Current vehicle
Real-time attitude simultaneously gathers real-time attitude data, and three axles of gyro sensor correspond respectively to the axis in front and back of Current vehicle
Line, upper and lower axis and left and right axis；Described gyro sensor is arranged on meter panel of motor vehicle；Camera acquisition is arrived
Present frame camera review be converted into gray-scale maps, and by 2 field of 1 neighborhood of left-lane line and right-lane line demarcated in step B
Respectively as region of interest ROI, fragmentation thresholdization process is then done respectively to region of interest ROI；If left-lane line 1 is oblique
The maximum magnitude of rate fluctuation is (- Δ kL,ΔkL), the maximum magnitude of intercept fluctuation is (- Δ bL,ΔbL), right-lane line 2 it is oblique
The maximum magnitude of rate fluctuation is (- Δ kR,ΔkR), the maximum magnitude of intercept fluctuation is (- Δ bR,ΔbR), choose left-lane line 1
Slope and intercept fluctuate maximum contiguous range as the region of interest ROI of left-lane line 1, choose 2 slope of right-lane line with
Region of interest ROI of the maximum contiguous range of intercept fluctuation as right-lane line 2, be arranged on slope fluctuation range mainly by
In vehicle, photographic head has a little shake, or other some correlative factors in the process of moving, under such as varying environment
Track line width and track spacing are different so that the lane line in camera views can fluctuate near the position of previous moment；
Respectively by region of interest ROI homogenous segmentations or uneven segmentation from top to bottom, per section of height is set to Δ hi, uniformly can divide
Section uneven can also be segmented, and for convenience, is stated using homogenous segmentations herein, is segmented more in theory, and testing result is more accurate
Really, it is contemplated that the problems such as real-time in practical application, institute's segments should be too many, according to practical situation should suitably be
Can；The average Δ h of Δ hi in each section of region of interest ROI is calculated respectively, and as the threshold value of this section, then will be felt emerging
Interesting region ROI binaryzations, when the threshold value of adjacent two sections of region of interest ROI is different, the boundary of adjacent two sections of region of interest ROI
The slope of line is 0.
Step D, real-time adjustment camera angle：The real-time attitude data that control unit is gathered according to gyro sensor,
I.e. angle of pitch α, angle of inclination beta, yaw angle γ, are imaged to the rightabout adjustment of angle of pitch α, angle of inclination beta, yaw angle γ in real time
Machine, with keep video camera all the time parallel to horizontal dead ahead, it is ensured that video camera shoot left-lane line 1 begin with right-lane line 2
Eventually in fixed area；
Step E, detection lane line undated parameter：Driveway line inspection is entered in fixed area using lane detection algorithm
Survey, include pretreatment image and detection track so as to detect left-lane line 1 and the lane detection algorithm described in right-lane line 2
Line undated parameter；Described pretreatment image includes the present frame camera review for arriving camera acquisition or current fixed area
Area image is converted into gray-scale maps, and using 1 neighborhood of left-lane line demarcated and 2 field of right-lane line as area-of-interest
ROI, then does fragmentation thresholdization process respectively to region of interest ROI；Simultaneously undated parameter includes sense to described detection lane line
Lane line is detected respectively in the ROI of interest region, and the linear equation that left-lane line 1 is updated if new left-lane line 1 is detected is oblique
Rate and/or intercept, otherwise retain the value at a moment thereon；The straight line of right-lane line 2 is updated if new right-lane line 2 is detected
Equation slope and/or intercept, otherwise retain the value at a moment thereon；Car is detected respectively using Hough transformation in area-of-interest
Diatom, updates the linear equation slope and/or intercept of left-lane line if new left-lane line is detected, and otherwise retains thereon one
The value at moment；The linear equation slope and/or intercept of right-lane line are updated if new right-lane line is detected, otherwise retains which
The value at a upper moment；If the situation of change lane occurs in vehicle, when waiting vehicle to enter central authorities' traveling in next track, automatically
The new lane line of detection.If bend, then what the method was detected is the tangent line of bend.
Specifically, in stepb, the left-lane line 1 and intersecting point coordinate of right-lane line 2 is after demarcating after demarcation：
Described fixed area (D) is：
Wherein W is picture traverse；
The final carriage matrix of Current vehicle is：
Wherein, R (x, α) is spin matrix of the gyro sensor around axis before and after Current vehicle, and R (y, β) is top
Spin matrix of the spiral shell instrument around the axis up and down of Current vehicle, R (z, γ) are gyroscopes around the left and right axis of Current vehicle
Spin matrix.
In the present invention, Current vehicle should travel the centre in place track as far as possible, currently also be not excluded for certain limit
Deviation, it is to be decorated with the highway of left and right lane line to demarcate scene, without the need for reconfiguring demarcation scene again, such as setting-out on the ground
Demarcate etc., it is so more simple and practical than traditional method, it is cost-effective；Fragmentation thresholdization can be reduced because shade or obstacle
Thing etc. blocks the inaccurate problem of lane detection that lane line is caused；Man-machine interaction is convenient and swift, to operator require compared with
It is low, therefore changeable scene is adapted to, it is workable；Using the scope of lane line fluctuation in front and back's two field pictures, sense is set
Emerging region can greatly reduce impact of the clutter to lane detection, and improve the execution efficiency of algorithm；On-line proving is supported, can
The a series of activities for gathering video image and processing is needed to remove off-line calibration from, directly demarcate the slope and intercept of lane line,
It is one simply and easily to demarcate very much and computational methods, a large amount of valuable times of developer, and human and material resources can be saved
Etc. aspect resource consumption.
In calibration process, the possibility source of error has：
1st, scene error：Lane line for demarcating is not straight, ground unevenness etc.；
2nd, man-machine interaction error：Left and right lane line in camera views is misaligned with calibration line.By theory analysis with
It is substantial amounts of it was verified that the lane detection method of calibration-type not only principle is simple, simple operation, can be easily integrated in on-vehicle machines and regard
In feel system, on-line proving is also supported, or even can be carried out in vehicle traveling, so as to solve vehicle-mounted taking the photograph to a certain extent
The drifting problem of camera relevant parameter, improves the accuracy of lane detection.
Specific embodiment described herein is only explanation for example spiritual to the present invention.Technology neck belonging to of the invention
The technical staff in domain can be made various modifications or supplement or replaced using similar mode to described specific embodiment
Generation, but without departing from the spiritual of the present invention or surmount scope defined in appended claims.
Although more having used Δ k hereinLRepresent the difference of red straight line and two straight slopes of left-lane line；ΔbLTable
Show the difference of blue straight line and two Linear intercepts of left-lane line；ΔkRRepresent red straight line and two straight slopes of right-lane line it
Difference；ΔbRRepresent the difference of blue straight line and two Linear intercepts of right-lane line；Region in picture between the blue line in two, left side
It is the emerging region ROI of left-lane sense；Region between the blue line in two, right side is the emerging region ROI of right lane sense；Δ h represents each section
Height；1 represents left-lane line；11 represent left demarcation straight line；2 represent right-lane line；21 represent the terms such as right demarcation straight line, but
It is not precluded from the probability using other terms.It is used for the purpose of more easily describing and explaining the present invention's using these terms
Essence；It is all contrary with spirit of the invention to be construed as any additional restriction.
Claims (10)
1. a kind of method that utilization gyro sensor carries out lane detection, it is characterised in that this method comprises the steps：
Scene is demarcated in A, selection:Selection is decorated with the track of left-lane line (1) and right-lane line (2) to demarcate scene, by current vehicle
Stop or travel between the left-lane line (1) and right-lane line (2) in place track, and guarantee Current vehicle place track
Left-lane line (1) it is visible in camera review with right-lane line (2)；
B, man-machine interaction are demarcated：Default left demarcation straight line (11) of the mobile display in camera review is so that left demarcation straight line
(11) overlap with the tangent line of left-lane line (1) or left-lane line (1) so as to realize that left-lane line (1) is demarcated, mobile display is in taking the photograph
Default right demarcation straight line (21) in camera image is so that right demarcation straight line (21) and right-lane line (2) or right-lane line (2)
Tangent line overlap so as to realize that right-lane line (2) is demarcated, calculate respectively now left-lane line (1) and right-lane line (2) in figure
As the linear equation in coordinate system；
C, detection body gesture angle:The real-time attitude of Current vehicle is monitored by the gyro sensor on Current vehicle
And gather real-time attitude data, three axles of gyro sensor correspond respectively to Current vehicle before and after axis, up and down in
Axis and left and right axis；
D, real-time adjustment camera angle：The real-time attitude data that control unit is gathered according to gyro sensor, the i.e. angle of pitch
α, angle of inclination beta, yaw angle γ, in real time to the rightabout adjustment video camera of angle of pitch α, angle of inclination beta, yaw angle γ, to keep
Video camera all the time parallel to horizontal dead ahead, it is ensured that left-lane line (1) and the right-lane line (2) that video camera shoots is all the time solid
Determine in region；
E, lane detection：Lane detection is carried out using lane detection algorithm in fixed area, so as to detect left car
Diatom (1) and right-lane line (2).
2. the method that utilization gyro sensor according to claim 1 carries out lane detection, it is characterised in that to take the photograph
The intersection point of camera optical axis and camera image plane is origin, using origin laterally to the right bearing of trend as x-axis bearing of trend, with
Origin longitudinally downward direction sets up image coordinate system as y-axis bearing of trend, and after demarcation, left-lane line (1) is in image coordinate system
In linear equation be y=kL·x+bL, wherein kLIt is the slope of left-lane line, bLIt is left-lane line intercept on the y axis, mark
After fixed, linear equation of the right-lane line (2) in image coordinate system is y=kR·x+bR, wherein kRIt is the slope of right-lane line, bR
It is right-lane line intercept on the y axis, the linear equation of the front bumper place straight line of Current vehicle is y=c；Described fixation
Region be by left-lane line after demarcating (1), demarcate after right-lane line (2) and Current vehicle front bumper place straight line surround and
Into delta-shaped region.
3. the method that utilization gyro sensor according to claim 2 carries out lane detection, it is characterised in that described
Demarcation after left-lane line (1) and the intersecting point coordinate of right-lane line (2) is after demarcating：
Described fixed area (D) is：
Wherein W is picture traverse；
The final carriage matrix of Current vehicle is：
Wherein, R (x, α) is spin matrix of the gyro sensor around axis before and after Current vehicle, and R (y, β) is gyroscope
Around the spin matrix of the axis up and down of Current vehicle, R (z, γ) is rotation of the gyroscope around the left and right axis of Current vehicle
Matrix.
4. the method that utilization gyro sensor according to claim 3 carries out lane detection, it is characterised in that upper
State in step E, described lane detection algorithm includes pretreatment image and detection lane line undated parameter；Described pre- place
Reason image include by camera acquisition to present frame camera review or current fixed area image be converted into gray-scale maps, and will
Left-lane line (1) neighborhood of demarcation and right-lane line (2) field are respectively as area-of-interest (ROI), then emerging to sense respectively
Interesting region (ROI) does fragmentation thresholdization process；Simultaneously undated parameter includes dividing in area-of-interest (ROI) described detection lane line
Lane line is not detected, the linear equation slope and/or intercept of left-lane line (1) is updated if new left-lane line (1) is detected,
Otherwise retain the value at a moment thereon；The linear equation slope of right-lane line (2) is updated if new right-lane line (2) is detected
And/or intercept, otherwise retain the value at a moment thereon.
5. the method that utilization gyro sensor according to claim 4 carries out lane detection, it is characterised in that set a left side
The maximum magnitude of the slope fluctuation of lane line (1) is (- Δ kL,ΔkL), the maximum magnitude of intercept fluctuation is (- Δ bL,ΔbL),
The maximum magnitude of the slope fluctuation of right-lane line (2) is (- Δ kR,ΔkR), the maximum magnitude of intercept fluctuation is (- Δ bR,Δ
bR), choose left-lane line (1) slope maximum contiguous range is fluctuated as the area-of-interest of left-lane line (1) with intercept
(ROI), choose right-lane line (2) slope maximum contiguous range is fluctuated as the area-of-interest of right-lane line (2) with intercept
(ROI)。
6. the method that utilization gyro sensor according to claim 5 carries out lane detection, it is characterised in that respectively
Area-of-interest (ROI) homogenous segmentations or uneven segmentation from top to bottom, per section of height are set to into Δ hi；Calculate respectively each
The average Δ h of Δ hi in section area-of-interest (ROI), and as the threshold value of this section, then by area-of-interest (ROI) two
Value, when the threshold value of adjacent two sections of area-of-interests (ROI) is different, adjacent two sections of area-of-interests (ROI) it is marginal tiltedly
Rate is 0.
7. the method that the utilization gyro sensor according to claim 4 or 5 or 6 carries out lane detection, its feature exist
In lane line respectively being detected in the area-of-interest (ROI) using Hough transformation, if new left-lane line (1) is detected more
The linear equation slope and/or intercept of new left-lane line (1), otherwise retains the value at a moment thereon；If detecting new right car
Diatom (2) then updates the linear equation slope and/or intercept of right-lane line (2), otherwise retains the value at a moment thereon；If vehicle
There is the situation of change lane, then when waiting vehicle to enter central authorities' traveling in next track, the new lane line of automatic detection.
8. the method that the utilization gyro sensor according to any one in claim 1-7 carries out lane detection, its
It is characterised by, in above-mentioned steps B, manually the left demarcation straight line (11) of operating-controlling mechanism control translates and/or rotates to adjust
The left intercept and/or slope for demarcating straight line (11), manually right demarcation straight line (21) translation of operating-controlling mechanism control and/or rotation
To adjust the right intercept and/or slope for demarcating straight line (21)；Described manual manipulation mechanism includes button and/or knob, or
Described manual manipulation mechanism includes the touch screen that can show camera review.
9. the method that the utilization gyro sensor according to any one in claim 1-7 carries out lane detection, its
Be characterised by, in above-mentioned steps A, described left-lane line (1) and right-lane line (2) be respectively solid line, dotted line, yellow line with it is white
The combination of any one or more in line；Current vehicle is stopped or travels the left-lane line (1) and right car in place track
Centre between diatom (2).
10. the method that the utilization gyro sensor according to any one in claim 1-7 carries out lane detection, its
It is characterised by, described gyro sensor is arranged on meter panel of motor vehicle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610970334.0A CN106525056A (en) | 2016-11-04 | 2016-11-04 | Method for lane line detection by gyro sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610970334.0A CN106525056A (en) | 2016-11-04 | 2016-11-04 | Method for lane line detection by gyro sensor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106525056A true CN106525056A (en) | 2017-03-22 |
Family
ID=58326015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610970334.0A Pending CN106525056A (en) | 2016-11-04 | 2016-11-04 | Method for lane line detection by gyro sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106525056A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107229908A (en) * | 2017-05-16 | 2017-10-03 | 浙江理工大学 | A kind of method for detecting lane lines |
| CN108769849A (en) * | 2018-04-26 | 2018-11-06 | Oppo广东移动通信有限公司 | The control method and Related product of wearable device |
| CN108802785A (en) * | 2018-08-24 | 2018-11-13 | 清华大学 | Vehicle method for self-locating based on High-precision Vector map and monocular vision sensor |
| CN109389650A (en) * | 2018-09-30 | 2019-02-26 | 京东方科技集团股份有限公司 | A kind of scaling method of in-vehicle camera, device, vehicle and storage medium |
| WO2020063817A1 (en) * | 2018-09-30 | 2020-04-02 | 长城汽车股份有限公司 | Traffic lane line fitting method and system |
| CN111010537A (en) * | 2019-12-06 | 2020-04-14 | 苏州智加科技有限公司 | Vehicle control method, device, terminal and storage medium |
| CN111212260A (en) * | 2018-11-21 | 2020-05-29 | 杭州海康威视数字技术股份有限公司 | Method and device for drawing lane line based on surveillance video |
| CN113515973A (en) * | 2020-04-09 | 2021-10-19 | 北京地平线机器人技术研发有限公司 | Data acquisition method, training method, adjustment method and device |
| CN114247081A (en) * | 2019-04-04 | 2022-03-29 | 南京敏思软件有限公司 | Method and system for determining motion energy consumption of intelligent stone lock |
| US12007243B2 (en) | 2018-09-30 | 2024-06-11 | Great Wall Motor Company Limited | Traffic lane line fitting method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010070012A (en) * | 2008-09-17 | 2010-04-02 | Toyota Motor Corp | Lane recognition device |
| CN102288121A (en) * | 2011-05-12 | 2011-12-21 | 电子科技大学 | Method for measuring and pre-warning lane departure distance based on monocular vision |
| CN104008645A (en) * | 2014-06-12 | 2014-08-27 | 湖南大学 | Lane line predicating and early warning method suitable for city road |
| CN104129389A (en) * | 2014-08-06 | 2014-11-05 | 中电海康集团有限公司 | Method for effectively judging and recognizing vehicle travelling conditions and device thereof |
| CN105197014A (en) * | 2014-06-30 | 2015-12-30 | 现代自动车株式会社 | Apparatus and method for recognizing driving lane of vehicle |
-
2016
- 2016-11-04 CN CN201610970334.0A patent/CN106525056A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010070012A (en) * | 2008-09-17 | 2010-04-02 | Toyota Motor Corp | Lane recognition device |
| CN102288121A (en) * | 2011-05-12 | 2011-12-21 | 电子科技大学 | Method for measuring and pre-warning lane departure distance based on monocular vision |
| CN104008645A (en) * | 2014-06-12 | 2014-08-27 | 湖南大学 | Lane line predicating and early warning method suitable for city road |
| CN105197014A (en) * | 2014-06-30 | 2015-12-30 | 现代自动车株式会社 | Apparatus and method for recognizing driving lane of vehicle |
| CN104129389A (en) * | 2014-08-06 | 2014-11-05 | 中电海康集团有限公司 | Method for effectively judging and recognizing vehicle travelling conditions and device thereof |
Non-Patent Citations (1)
| Title |
|---|
| 周金金等: "基于ARM11 的海洋浮标云台稳定控制系统" * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107229908B (en) * | 2017-05-16 | 2019-11-29 | 浙江理工大学 | A kind of method for detecting lane lines |
| CN107229908A (en) * | 2017-05-16 | 2017-10-03 | 浙江理工大学 | A kind of method for detecting lane lines |
| CN108769849A (en) * | 2018-04-26 | 2018-11-06 | Oppo广东移动通信有限公司 | The control method and Related product of wearable device |
| CN108802785A (en) * | 2018-08-24 | 2018-11-13 | 清华大学 | Vehicle method for self-locating based on High-precision Vector map and monocular vision sensor |
| CN109389650B (en) * | 2018-09-30 | 2021-01-12 | 京东方科技集团股份有限公司 | Vehicle-mounted camera calibration method and device, vehicle and storage medium |
| CN109389650A (en) * | 2018-09-30 | 2019-02-26 | 京东方科技集团股份有限公司 | A kind of scaling method of in-vehicle camera, device, vehicle and storage medium |
| WO2020063817A1 (en) * | 2018-09-30 | 2020-04-02 | 长城汽车股份有限公司 | Traffic lane line fitting method and system |
| US12007243B2 (en) | 2018-09-30 | 2024-06-11 | Great Wall Motor Company Limited | Traffic lane line fitting method and system |
| CN111212260B (en) * | 2018-11-21 | 2021-08-20 | 杭州海康威视数字技术股份有限公司 | Method and device for drawing lane line based on surveillance video |
| CN111212260A (en) * | 2018-11-21 | 2020-05-29 | 杭州海康威视数字技术股份有限公司 | Method and device for drawing lane line based on surveillance video |
| CN114247081A (en) * | 2019-04-04 | 2022-03-29 | 南京敏思软件有限公司 | Method and system for determining motion energy consumption of intelligent stone lock |
| CN111010537A (en) * | 2019-12-06 | 2020-04-14 | 苏州智加科技有限公司 | Vehicle control method, device, terminal and storage medium |
| CN113515973A (en) * | 2020-04-09 | 2021-10-19 | 北京地平线机器人技术研发有限公司 | Data acquisition method, training method, adjustment method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106525056A (en) | Method for lane line detection by gyro sensor | |
| CN106919915B (en) | Map road marking and road quality acquisition device and method based on ADAS system | |
| CN107235044B (en) | A kind of restoring method realized based on more sensing datas to road traffic scene and driver driving behavior | |
| KR102493878B1 (en) | controlling host vehicle based on detected spacing between stationary vehicles | |
| Pomerleau | RALPH: Rapidly adapting lateral position handler | |
| CN102682292B (en) | Method based on monocular vision for detecting and roughly positioning edge of road | |
| US10896626B2 (en) | Method, computer readable storage medium and electronic equipment for analyzing driving behavior | |
| CN105678285B (en) | A kind of adaptive road birds-eye view transform method and road track detection method | |
| CN105260699B (en) | A kind of processing method and processing device of lane line data | |
| EP2605185B1 (en) | Detection of obstacles at night by analysis of shadows | |
| DE112020006426T5 (en) | SYSTEMS AND METHODS FOR VEHICLE NAVIGATION | |
| CN109359409A (en) | A kind of vehicle passability detection system of view-based access control model and laser radar sensor | |
| CN109085823A (en) | The inexpensive automatic tracking running method of view-based access control model under a kind of garden scene | |
| CN108519605A (en) | Curb detection method based on laser radar and video camera | |
| CN107646114A (en) | Method for estimating track | |
| CN110203210A (en) | A kind of lane departure warning method, terminal device and storage medium | |
| KR20190132404A (en) | Direct vehicle detection as 3D bounding boxes using neural network image processing | |
| EP2372605A2 (en) | Image processing system and position measurement system | |
| CN106682646A (en) | Method and apparatus for recognizing lane line | |
| CN108068817A (en) | A kind of automatic lane change device and method of pilotless automobile | |
| DE112020002764T5 (en) | SYSTEMS AND METHODS FOR VEHICLE NAVIGATION | |
| CN108108750A (en) | Metric space method for reconstructing based on deep learning and monocular vision | |
| CN107462223A (en) | Driving sight distance self-operated measuring unit and measuring method before a kind of highway is turned | |
| CN105835880A (en) | Lane tracking system | |
| DE112021002001T5 (en) | NAVIGATING A VEHICLE USING AN ELECTRONIC HORIZON |
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 | ||
| AD01 | Patent right deemed abandoned | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20240419 |