CN108647572A - A kind of lane departure warning method based on Hough transformation - Google Patents
A kind of lane departure warning method based on Hough transformation Download PDFInfo
- Publication number
- CN108647572A CN108647572A CN201810292955.7A CN201810292955A CN108647572A CN 108647572 A CN108647572 A CN 108647572A CN 201810292955 A CN201810292955 A CN 201810292955A CN 108647572 A CN108647572 A CN 108647572A
- Authority
- CN
- China
- Prior art keywords
- line
- lane line
- lane
- straight line
- collection
- 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.)
- Granted
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 Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of lane departure warning method based on Hough transformation, this method carries out region of interest regional partition to image first;Then gray processing region of interest area image;Edge detection is carried out to gray level image using improved unidirectional gradient operator and obtains edge image;Binary conversion treatment is carried out to image by Otsu algorithm;Candidate straight line collection is obtained using Hough transformation, and left and right track straight line collection is divided into according to slope, two straight line collection is constrained with vanishing point random sample consensus algorithm is combined to screen respectively, obtain best left and right lane line parameter;Two Kalman filter are built to the best lane line in left and right respectively into line trace;Finally utilize left and right track line slope estimation relative cross offset and its variation tendency, identification deviation and early warning.The present invention can effectively identify the deviation problem in driving driving, improve the robustness and real-time of Lane Departure Warning System.
Description
Technical field
The present invention relates to technical field of computer vision, and in particular to a kind of lane departure warning side based on Hough transformation
Method.
Background technology
With the continuous development of society with the continuous improvement of national economy, automobile number is growing day by day.Due to China human mortality
Radix is too big, and automobile number sharply increases the high rate for resulting in traffic accident, and national economy is made to be subject to huge damage
It loses, even more threatens personal safety.According to correlation study the results show that the cause minority of traffic accident by bad weather and
Road itself causes, and the problem of being more by driver itself is caused, and deviation is one of main cause.
Current Lane Departure Warning System is mostly based on computer vision technique exploitation, generally only clear in lane line
Highway on competence exertion effectively act on.Detection error is may result in if there is chaff interferent or lane line are fuzzy, greatly
The big accuracy for reducing system, is susceptible to false alarm.Moreover, the factors such as weather and illumination are also easy to influence system
Accuracy.
Invention content
The purpose of the present invention is to solve drawbacks described above in the prior art, provide a kind of track based on Hough transformation
Deviate method for early warning, to further increase the robustness and real-time of Lane Departure Warning System.Pass through image preprocessing, track
Detection, lane departure warning series of steps, the present invention can effectively identify the deviation problem in driving driving, method tool
There are higher robustness and real-time.
The purpose of the present invention can be reached by adopting the following technical scheme that:
A kind of lane departure warning method based on Hough transformation, which is characterized in that the lane departure warning method
Include the following steps:
S1, video image is acquired by camera, divides area-of-interest;
S2, the video image acquired in step S1 is pre-processed, including image gray processing, edge detection, the calculation of big Tianjin
Method determines adaptive threshold, image binaryzation;
S3, candidate straight line collection is obtained to binaryzation edge image progress Hough transformation in step S2, and is divided into according to slope
Left and right track straight line collection;
S4, to the left and right track straight line collection that is obtained in step S3 using vanishing point constraint and random sample consensus algorithm into
Row screens twice, obtains best left and right lane line parameter;
The parameter (ρ, θ) for the left and right lane line that 2 S5, structure Kalman filter obtain in tracking step S4 respectively, and
2 missing inspection frame counters are respectively set, when present frame detects lane line, corresponding missing inspection frame counter resets, while more
New corresponding Kalman filter;When that can not detect lane line, corresponding missing inspection frame counter is just added one, is counted when reaching
When number threshold values T, just thinks that the lane line has been lost, freeze corresponding Kalman filter;When lane line is detected again
When, reinitialize corresponding Kalman filter;
S6, it is being only able to detect a lane line, is estimating the slope of an other lane line;Utilize left and right track
Line slope estimates relative cross offset and its variation tendency, and deviation is identified according to relative cross offset and its variation tendency.
Further, in the step S3, for the straight line collection that Hough transformation detects, in pole coordinate parameter space
(ρ, θ) parameter indicate, be that straight line within the scope of 10 °~70 ° is classified as left-lane line set by polar angle θ, polar angle θ is 110 °~
Straight line within the scope of 170 ° is classified as right-lane line set, remaining straight line is deleted.
Further, with vanishing point to constrain the process screened to two straight line collection in the step S4 as follows:
Remember that left and right lane line collection is respectively:
L={ l1(ρ1,θ1),l2(ρ2,θ2),.....,lm(ρm,θm)}
R={ r1(ρ1,θ1),r2(ρ2,θ2),.....,rn(ρn,θn)}
Wherein, L is left-lane line collection, and R is right-lane line collection, and m, n respectively represent the straight line number that left and right lane line is concentrated
Amount, (ρ, θ) are the parameter of line correspondence;
To lane line concentrate arbitrary line to li(ρi,θi)、rj(ρj,θj), utilize following system of linear equations:
Obtain the intersecting point coordinate (u of 2 straight linesk,vk), the set of m and n straight line can obtain m*n friendship by above-mentioned formula pairing
Point searches for other intersection points in its M*M neighborhood to each intersection point and records neighborhood number of intersections;
Compare the neighborhood number of intersections of m*n intersection point, the most intersection point of neighborhood number of intersections is vanishing point, and neighborhood is regarded as going out
Point effective coverage, selects the straight line pair in neighborhood, re-establishes new left and right lane line collection, it is complete that vanishing point constraint carries out preliminary screening
At.
Further, mistake results of preliminary screening screened with random sample consensus algorithm in the step S4
Journey is as follows:
Left and right lane line set is handled successively, and a parameter for selecting lane line to concentrate arbitrary first is the straight of (ρ, θ)
Line, note are l with the corresponding vertical line of area-of-interest lower boundary intersection point by this straight linei, can be respectively in the hope of slope-
Tan θ+Δ A and-tan θ-Δ A and two straight lines by vanishing point, are denoted as l+And l-, straight line l+And l-With liIntersection point image coordinate
It is denoted as (u respectively+,v+) and (u-,v-), calculate remaining straight line and l in the setiIntersection point image coordinate (u, v), record intersection point is horizontal
Coordinate u is fallen in u-And u+Between straight line quantity, be denoted as Ni.The straight line that lane line is concentrated constantly is taken out to repeat the above steps
Until all straight lines are traversed, N is recordediIt is worth maximum straight line, the best lane line as corresponding lane line collection.
Further, the step S6 processes are as follows:
When present frame can detect left and right lane line, then right-lane line slope is-tan θr, left-lane line slope be-
tanθl, the lateral shift distance at note automotive run-off-road center is that relative cross offset d, d are logical with the ratio of lane width half
Cross following formula calculating:
Wherein d ∈ [- 1,1], when vehicle is normally placed in the middle, d values fluctuate near 0 when driving, when d gradually increases, judgement
Vehicle moves to left, and when d values are gradually reduced, judgement vehicle moves to right, and has following relationship between left and right track line slope:
Wherein k1, k2For constant, u0For vanishing point abscissa;
Work as u0When close to 0, constant k1Estimated using following formula:
k1=tan θl-tanθr;
Work as u0When not being 0, constant k2Estimated using following formula:
When present frame can detect left and right lane line, the position for the line that gone out according to the ordinate preservation level of vanishing point calculates
Constant k1And k2, and preserve;When present frame can only detect a lane line, according to level go out line and lane line intersection point it is true
Determine vanishing point, then according to relation formula between above-mentioned left and right track line slope, utilizes k1, k2And u0Value find out another track
The slope of line, and then d is calculated, still it can differentiate vehicle relative cross offset by the method for two-way traffic line;
Interframe smothing filtering is carried out to relative cross offset d using sliding window filter, calculation formula is as follows:
Wherein, N is filtering window size, WiFor filter coefficient, diFor the relative cross offset of the i-th frame;
Then it calculatesInter-frame difference, formula is as follows:
NoteDeviation threshold values be dT, deviation counter is CdIfIt then resets and deviates counter CdIt is 0, sentences
Not Wei vehicle be in straight-going state, ifAndMore than 0, then CdSubtract one;IfAndLess than 0, then
CdIt resets;IfAndLess than 0, then CdAdd one;IfAndMore than 0, then CdIt resets, CdRange
For [- Cmax,Cmax], it goes beyond the scope and takes boundary value, wherein CmaxFor upper boundary values, if Cd>CT, then it is judged as that vehicle is deviated to the right vehicle
Road;If Cd<-CT, then it is judged as that vehicle is deviated to the left track.
Further, in the step S6, if left and right lane line can't detect, without differentiating, processing is next
Frame data are until detecting lane line.
The present invention has the following advantages and effects with respect to the prior art:
1, the present invention using vanishing point constraint combine the lane line collection that random sample consensus algorithm detects Hough transformation into
Row screening, improves the accuracy rate of lane detection.
2, the present invention tracks left and right lane line using 2 independent Kalman filter, improves the reliable of lane line tracking
Property.
3, the present invention identifies deviation using relative cross offset and its variation tendency, needs not rely on camera correction
Parameter, while can remain to effectively identify in the case where only detecting a lane line, improve the robustness of identification.
Description of the drawings
Fig. 1 is the process step figure of the lane departure warning method disclosed by the invention based on Hough transformation;
Fig. 2 is the gray level image schematic diagram after image gray processing in the embodiment of the present invention;
Fig. 3 is area-of-interest (ROI) the image schematic diagram divided in the embodiment of the present invention;
Fig. 4 is the edge binary images signal handled by unidirectional gradient operator and Otsu algorithm in the embodiment of the present invention
Figure;
Fig. 5 is line correspondences ρ and θ the parameter schematic diagram of Hough transformation in the embodiment of the present invention;
Fig. 6 is the track detected after Hough transformation and random sample consensus algorithm process in the embodiment of the present invention
Line image schematic diagram;
Fig. 7 is the flow diagram of Kalman filter tracking.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
The every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Embodiment
Referring to Fig.1, the lane departure warning method proposed by the present invention based on Hough transformation comprises the steps of:
S1, video image, segmentation area-of-interest (hereinafter referred to as ROI region) are acquired by camera.It will be under image
It is area-of-interest that half part, which delimited, and reduce irrelevant region influences caused by algorithm.
S2, the image collected frame is handled, converts RGB color image to gray level image, as a result such as Fig. 2 institutes
Show, calculation formula is as follows:
I (u, v)=0.299*IR(u,v)+0.587*IG(u,v)+0.114*IB(u,v)
The gray level image of area-of-interest is intercepted, the results are shown in Figure 3.Lane line is detected with improved unidirectional gradient operator
Edge, calculation formula are as follows:
f(u,V)=2 × I (u,v)-I(u+S,v)-I(u-S,v)-|I(u+S,v)-I(u-S,v)|
Wherein, parameter S controls the track line width of detection, and S is 15 in the present embodiment.The image obtained for edge detection
Adaptive threshold is calculated with Otsu algorithm, the threshold value is used in combination to carry out image binaryzation operation, the feature of prominent lane line.After processing
The results are shown in Figure 4.
S3, Hough transformation is carried out to binary edge map and removes unreasonable straight line, obtained straight line collection is divided
Class is divided into left-lane line straight line collection and right-lane line straight line collection.It is illustrated in figure 5 Hough transformation cathetus corresponding ρ and θ ginseng
Number schematic diagram.For the straight line collection that Hough transformation detects, polar angle θ is classified as left-lane line for the straight line within the scope of 10 °~70 °
Set, polar angle θ are that the straight line within the scope of 110 °~170 ° is classified as right-lane line set, remaining straight line is deleted.
S4, two straight line collection are constrained with vanishing point and are screened in conjunction with random sample consensus algorithm, remember left and right lane line
Collection is respectively:
L={ l1(ρ1,θ1),l2(ρ2,θ2),.....,lm(ρm,θm)}
R={ r1(ρ1,θ1),r2(ρ2,θ2),.....,rn(ρn,θn)}
Wherein, L is left-lane line collection, and R is right-lane line collection, and m, n respectively represent the straight line number that left and right lane line is concentrated
Amount, (ρ, θ) are the parameter of line correspondence.To lane line concentrate arbitrary line to li(ρi,θi)、rj(ρj,θj), using such as offline
Property equation group:
It can obtain the intersecting point coordinate (u of 2 straight linesk,vk), there are m and the set of n straight line to be matched by above-mentioned formula respectively
M*n intersection point can be obtained, other intersection points are searched in its M*M neighborhood to each intersection point and records neighborhood number of intersections.Compare m*n
The neighborhood number of intersections of intersection point, the most intersection point of neighborhood number of intersections are vanishing point, and neighborhood is regarded as vanishing point effective coverage, and selection is adjacent
Straight line pair in domain re-establishes new left and right lane line collection, and first screening is completed.In the present embodiment, M takes 5.
Further left and right lane line set is handled successively with random sample consensus algorithm.Lane line is arbitrarily selected first
The parameter concentrated is the straight line of (ρ, θ), and note is l with the corresponding vertical line of lower boundary intersection point interested by this straight linei,
It can be respectively-tan θ+Δ A and-tan θ-Δ A and two straight lines by vanishing point in the hope of slope, be denoted as l+And l-, straight line l+With
l-With liIntersection point image coordinate be denoted as (u respectively+,v+) and (u-,v-), calculate remaining straight line and l in the setiIntersection point image
Coordinate (u, v), record intersection point abscissa u are fallen in u-And u+Between straight line quantity, be denoted as Ni.Constantly take out what lane line was concentrated
Straight line steps be repeated alternatively until that all straight lines are traversed, and record NiIt is worth maximum straight line, as corresponding track
The best lane line of line collection.The results are shown in Figure 6.In the present embodiment, Δ A is 0.5.
2 S5, structure Kalman filter, track the parameter (ρ, θ) of left-lane line and right-lane line respectively, and are them
Missing inspection frame counter is respectively set, the tracking of left and right lane line is mutual indepedent.In the present embodiment, the parameter vector of pick-up diatom
[ρ,θ]TAs the state vector of Kalman filter, the state-transition matrix of Kalman filter is:
The process noise covariance matrix initialisation of Kalman filter is as follows:
The measurement noise covariance matrix initialization of Kalman filter is as follows:
When detecting lane line in the current frame, corresponding missing inspection frame counter resets, while updating corresponding Kalman's filter
Wave device.When that can not detect a wherein lane line, corresponding missing inspection frame counter is just added one, when reaching count threshold T
When, just think that the lane line has been lost, and freeze corresponding Kalman filter.When lane line is detected again, weight
The new corresponding Kalman filter of initialization, flow are as shown in Figure 7.In the present embodiment, T takes 5.
S6, when present frame can detect left and right lane line, then right-lane line slope be-tan θr, left-lane line slope
For-tan θl, the lateral shift distance at note automotive run-off-road center is relative cross offset d with the ratio of lane width half,
D can be calculate by the following formula:
Wherein d ∈ [- 1,1], when vehicle is normal placed in the middle when driving, d values fluctuate near 0.When d gradually increases, can sentence
Determine vehicle to move to left.When d values are gradually reduced, it can determine that vehicle moves to right.There is following relationship between left and right track line slope
Wherein k1, k2For constant, u0For vanishing point abscissa.
Work as u0When close to 0, constant k1Following formula can be utilized to estimate:
k1=tan θl-tanθr
Work as u0When not being 0, constant k2Following formula can be utilized to estimate:
When present frame can detect left and right lane line, the position for the line that gone out according to the ordinate preservation level of vanishing point calculates
Constant k1And k2, and preserve;When present frame can only detect a lane line, going out the intersection point of line and lane line according to level can
To determine vanishing point, then according to relation formula between above-mentioned left and right lane line, k is utilized1, k2And u0Value find out another track
The slope of line, and then d is calculated, still it can differentiate vehicle relative cross offset by the method for two-way traffic line.
Interframe smothing filtering is carried out to d using sliding window filter, calculation formula is as follows:
Wherein, N is filtering window size, WiFor filter coefficient, diFor the relative cross offset of the i-th frame.The present embodiment
Middle N values are 5, filter coefficient WiIt is 1/N.
Then it calculatesInter-frame difference, formula is as follows:
NoteDeviation threshold values be dT, deviation counter is Cd.IfIt then resets and deviates counter CdIt is 0, sentences
Not Wei vehicle be in straight-going state.IfAndMore than 0, then CdSubtract one;IfAndLess than 0, then
CdIt resets;IfAndLess than 0, then CdAdd one;IfAndMore than 0, then CdIt resets.CdRange
For [- Cmax,Cmax], it goes beyond the scope and takes boundary value, wherein CmaxFor upper boundary values.If Cd>CT, then it is judged as that vehicle is deviated to the right vehicle
Road;If Cd<-CT, then it is judged as that vehicle is deviated to the left track.In the present embodiment, dT=0.5, CT=2, Cmax=5.
If left and right lane line can't detect, without differentiating, handle next frame data is until detecting lane line
Only.
The above embodiment is a preferred embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, it is other it is any without departing from the spirit and principles of the present invention made by changes, modifications, substitutions, combinations, simplifications,
Equivalent substitute mode is should be, is included within the scope of the present invention.
Claims (6)
1. a kind of lane departure warning method based on Hough transformation, which is characterized in that the lane departure warning method packet
Include the following steps:
S1, video image is acquired by camera, divides area-of-interest;
S2, the video image acquired in step S1 is pre-processed, including image gray processing, edge detection, Otsu algorithm are true
Determine adaptive threshold, image binaryzation;
S3, Hough transformation is carried out to binaryzation edge image in step S2 obtain candidate straight line collection, and be divided into according to slope left and right
Track straight line collection;
S4, two are carried out using vanishing point constraint and random sample consensus algorithm to the left and right track straight line collection obtained in step S3
Secondary screening obtains best left and right lane line parameter;
The parameter (ρ, θ) for the left and right lane line that 2 S5, structure Kalman filter obtain in tracking step S4 respectively, and respectively
2 missing inspection frame counters are set, and when present frame detects lane line, corresponding missing inspection frame counter resets, while update pair
The Kalman filter answered;When that can not detect lane line, corresponding missing inspection frame counter is just added one, when reaching counting valve
When value T, just thinks that the lane line has been lost, freeze corresponding Kalman filter;When lane line is detected again, weight
Newly initialize corresponding Kalman filter;
S6, it is being only able to detect a lane line, is estimating the slope of an other lane line;It is oblique using left and right lane line
Rate estimates relative cross offset and its variation tendency, and deviation is identified according to relative cross offset and its variation tendency.
2. a kind of lane departure warning method based on Hough transformation according to claim 1, which is characterized in that described
In step S3, for the straight line collection that Hough transformation detects, indicated with (ρ, θ) parameter in pole coordinate parameter space, by polar angle θ
It is classified as left-lane line set for the straight line within the scope of 10 °~70 °, polar angle θ is that the straight line within the scope of 110 °~170 ° is classified as right vehicle
Diatom set, remaining straight line are deleted.
3. a kind of lane departure warning method based on Hough transformation according to claim 1, which is characterized in that described
It is as follows that with vanishing point the process screened is constrained to two straight line collection in step S4:
Remember that left and right lane line collection is respectively:
L={ l1(ρ1,θ1),l2(ρ2,θ2),.....,lm(ρm,θm)}
R={ r1(ρ1,θ1),r2(ρ2,θ2),.....,rn(ρn,θn)}
Wherein, L is left-lane line collection, and R is right-lane line collection, and m, n respectively represent the straight line quantity that left and right lane line is concentrated, (ρ,
It is θ) parameter of line correspondence;
To lane line concentrate arbitrary line to li(ρi,θi)、rj(ρj,θj), utilize following system of linear equations:
Obtain the intersecting point coordinate (u of 2 straight linesk,vk), the set of m and n straight line can obtain m*n intersection point by above-mentioned formula pairing,
Other intersection points are searched in its M*M neighborhood to each intersection point and record neighborhood number of intersections;
Compare the neighborhood number of intersections of m*n intersection point, the most intersection point of neighborhood number of intersections is vanishing point, and neighborhood, which is regarded as vanishing point, to be had
Region is imitated, the straight line pair in neighborhood is selected, re-establishes new left and right lane line collection, vanishing point constraint carries out preliminary screening completion.
4. a kind of lane departure warning method based on Hough transformation according to claim 3, which is characterized in that described
The process screened with random sample consensus algorithm to results of preliminary screening in step S4 is as follows:
Left and right lane line set is handled successively, and a parameter for selecting lane line to concentrate arbitrary first is the straight line of (ρ, θ), note
It is l by this straight line and the corresponding vertical line of area-of-interest lower boundary intersection pointi, can be respectively-tan θ+Δ in the hope of slope
A and-tan θ-Δ A and two straight lines by vanishing point, are denoted as l+And l-, straight line l+And l-With liIntersection point image coordinate remember respectively
For (u+,v+) and (u-,v-), calculate remaining straight line and l in the setiIntersection point image coordinate (u, v), record intersection point abscissa u
It falls in u-And u+Between straight line quantity, be denoted as Ni.It constantly takes out the straight line that lane line is concentrated and steps be repeated alternatively until institute
There is straight line to be traversed, records NiIt is worth maximum straight line, the best lane line as corresponding lane line collection.
5. a kind of lane departure warning method based on Hough transformation according to claim 1, which is characterized in that described
Step S6 processes are as follows:
When present frame can detect left and right lane line, then right-lane line slope is-tan θr, left-lane line slope is-tan θl,
The lateral shift distance at note automotive run-off-road center is that relative cross offset d, d pass through following formula with the ratio of lane width half
It calculates:
Wherein d ∈ [- 1,1], when vehicle is normally placed in the middle, d values fluctuate near 0 when driving, when d gradually increases, judge vehicle
It moves to left, when d values are gradually reduced, judgement vehicle moves to right, and has following relationship between left and right track line slope:
Wherein k1, k2For constant, u0For vanishing point abscissa;
Work as u0When close to 0, constant k1Estimated using following formula:
k1=tan θl-tanθr;
Work as u0When not being 0, constant k2Estimated using following formula:
When present frame can detect left and right lane line, the position for the line that gone out according to the ordinate preservation level of vanishing point, computational constant
k1And k2, and preserve;When present frame can only detect a lane line, gone out according to the horizontal intersection point determination for going out line and lane line
Point utilizes k then according to relation formula between above-mentioned left and right track line slope1, k2And u0Value find out another lane line
Slope, and then d is calculated, still it can differentiate vehicle relative cross offset by the method for two-way traffic line;
Interframe smothing filtering is carried out to relative cross offset d using sliding window filter, calculation formula is as follows:
Wherein, N is filtering window size, WiFor filter coefficient, diFor the relative cross offset of the i-th frame;
Then it calculatesInter-frame difference, formula is as follows:
NoteDeviation threshold values be dT, deviation counter is CdIfIt then resets and deviates counter CdIt is 0, is determined as
Vehicle is in straight-going state, ifAndMore than 0, then CdSubtract one;IfAndLess than 0, then CdIt is multiple
Position;IfAndLess than 0, then CdAdd one;IfAndMore than 0, then CdIt resets, CdRanging from [-
Cmax,Cmax], it goes beyond the scope and takes boundary value, wherein CmaxFor upper boundary values, if Cd>CT, then it is judged as that vehicle is deviated to the right track;
If Cd<-CT, then it is judged as that vehicle is deviated to the left track.
6. a kind of lane departure warning method based on Hough transformation according to claim 5, which is characterized in that described
In step S6, if left and right lane line can't detect, without differentiating, handle next frame data is until detecting lane line
Only.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810292955.7A CN108647572B (en) | 2018-04-04 | 2018-04-04 | Lane departure early warning method based on Hough transform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810292955.7A CN108647572B (en) | 2018-04-04 | 2018-04-04 | Lane departure early warning method based on Hough transform |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108647572A true CN108647572A (en) | 2018-10-12 |
CN108647572B CN108647572B (en) | 2022-03-29 |
Family
ID=63745186
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810292955.7A Active CN108647572B (en) | 2018-04-04 | 2018-04-04 | Lane departure early warning method based on Hough transform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108647572B (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109263659A (en) * | 2018-07-28 | 2019-01-25 | 上海商汤智能科技有限公司 | Intelligent driving control method and device, vehicle, electronic equipment, medium, product |
CN109435842A (en) * | 2018-10-19 | 2019-03-08 | 四川长虹电器股份有限公司 | The detection of parking stall deviation and auxiliary prompting method based on reversing automobile image |
CN109492598A (en) * | 2018-11-19 | 2019-03-19 | 辽宁工业大学 | A kind of highway automobile run-off-road line initiative recognition and method for early warning based on machine vision |
CN109782764A (en) * | 2019-01-21 | 2019-05-21 | 湖北汽车工业学院 | A kind of unmanned logistics distribution system of intelligent solar, control method and dispensing vehicle |
CN109871776A (en) * | 2019-01-23 | 2019-06-11 | 昆山星际舟智能科技有限公司 | The method for early warning that round-the-clock lane line deviates |
CN109886122A (en) * | 2019-01-23 | 2019-06-14 | 珠海市杰理科技股份有限公司 | Method for detecting lane lines, device, computer equipment and storage medium |
CN109993099A (en) * | 2019-03-27 | 2019-07-09 | 西安航空职业技术学院 | A kind of lane line drawing recognition methods based on machine vision |
CN110705342A (en) * | 2019-08-20 | 2020-01-17 | 上海阅面网络科技有限公司 | Lane line segmentation detection method and device |
CN111079598A (en) * | 2019-12-06 | 2020-04-28 | 深圳市艾为智能有限公司 | Lane line detection method based on image texture and machine learning |
CN111241911A (en) * | 2019-12-11 | 2020-06-05 | 华侨大学 | Self-adaptive lane line detection method |
CN112380956A (en) * | 2020-11-10 | 2021-02-19 | 苏州艾氪英诺机器人科技有限公司 | Lane judgment method |
CN112373477A (en) * | 2020-11-23 | 2021-02-19 | 重庆长安汽车股份有限公司 | Redundancy control method for automatic driving system, automobile, controller, and computer-readable storage medium |
CN113033418A (en) * | 2021-03-25 | 2021-06-25 | 南京工程学院 | Lane departure early warning method suitable for mobile terminal |
CN114037977A (en) * | 2022-01-07 | 2022-02-11 | 深圳佑驾创新科技有限公司 | Road vanishing point detection method, device, equipment and storage medium |
CN114128461A (en) * | 2021-10-27 | 2022-03-04 | 江汉大学 | Control method of plug seedling transplanting robot and plug seedling transplanting robot |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1810907A1 (en) * | 2006-01-19 | 2007-07-25 | Valeo Vision | Method of adjusting the orientation of a camera installed on a vehicle and system for implementing this method |
US20130100286A1 (en) * | 2011-10-21 | 2013-04-25 | Mesa Engineering, Inc. | System and method for predicting vehicle location |
CN104318258A (en) * | 2014-09-29 | 2015-01-28 | 南京邮电大学 | Time domain fuzzy and kalman filter-based lane detection method |
CN105160309A (en) * | 2015-08-24 | 2015-12-16 | 北京工业大学 | Three-lane detection method based on image morphological segmentation and region growing |
KR101603293B1 (en) * | 2015-01-26 | 2016-03-16 | 한국기술교육대학교 산학협력단 | Lane recognition method |
WO2016178335A1 (en) * | 2015-05-07 | 2016-11-10 | 株式会社日立製作所 | Lane detection apparatus and lane detection method |
-
2018
- 2018-04-04 CN CN201810292955.7A patent/CN108647572B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1810907A1 (en) * | 2006-01-19 | 2007-07-25 | Valeo Vision | Method of adjusting the orientation of a camera installed on a vehicle and system for implementing this method |
US20130100286A1 (en) * | 2011-10-21 | 2013-04-25 | Mesa Engineering, Inc. | System and method for predicting vehicle location |
CN104318258A (en) * | 2014-09-29 | 2015-01-28 | 南京邮电大学 | Time domain fuzzy and kalman filter-based lane detection method |
KR101603293B1 (en) * | 2015-01-26 | 2016-03-16 | 한국기술교육대학교 산학협력단 | Lane recognition method |
WO2016178335A1 (en) * | 2015-05-07 | 2016-11-10 | 株式会社日立製作所 | Lane detection apparatus and lane detection method |
CN105160309A (en) * | 2015-08-24 | 2015-12-16 | 北京工业大学 | Three-lane detection method based on image morphological segmentation and region growing |
Non-Patent Citations (3)
Title |
---|
HOU CHANGZHENG等: "An Efficient Lane Markings Detection and Tracking Method based on Vanishing Point Constraints", 《2016 35TH CHINESE CONTROL CONFERENCE (CCC)》 * |
唐阳山等: "基于canny和霍夫变换的车道线识别算法研究", 《汽车实用技术》 * |
江国来等: "自适应模型和固定模型结合的肤色分割算法", 《计算机应用》 * |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109263659A (en) * | 2018-07-28 | 2019-01-25 | 上海商汤智能科技有限公司 | Intelligent driving control method and device, vehicle, electronic equipment, medium, product |
CN109435842A (en) * | 2018-10-19 | 2019-03-08 | 四川长虹电器股份有限公司 | The detection of parking stall deviation and auxiliary prompting method based on reversing automobile image |
CN109492598A (en) * | 2018-11-19 | 2019-03-19 | 辽宁工业大学 | A kind of highway automobile run-off-road line initiative recognition and method for early warning based on machine vision |
CN109782764A (en) * | 2019-01-21 | 2019-05-21 | 湖北汽车工业学院 | A kind of unmanned logistics distribution system of intelligent solar, control method and dispensing vehicle |
CN109871776A (en) * | 2019-01-23 | 2019-06-11 | 昆山星际舟智能科技有限公司 | The method for early warning that round-the-clock lane line deviates |
CN109886122A (en) * | 2019-01-23 | 2019-06-14 | 珠海市杰理科技股份有限公司 | Method for detecting lane lines, device, computer equipment and storage medium |
CN109871776B (en) * | 2019-01-23 | 2023-04-14 | 昆山星际舟智能科技有限公司 | All-weather lane line deviation early warning method |
CN109993099A (en) * | 2019-03-27 | 2019-07-09 | 西安航空职业技术学院 | A kind of lane line drawing recognition methods based on machine vision |
CN110705342A (en) * | 2019-08-20 | 2020-01-17 | 上海阅面网络科技有限公司 | Lane line segmentation detection method and device |
CN111079598A (en) * | 2019-12-06 | 2020-04-28 | 深圳市艾为智能有限公司 | Lane line detection method based on image texture and machine learning |
CN111079598B (en) * | 2019-12-06 | 2023-08-08 | 深圳市艾为智能有限公司 | Lane line detection method based on image texture and machine learning |
CN111241911A (en) * | 2019-12-11 | 2020-06-05 | 华侨大学 | Self-adaptive lane line detection method |
CN111241911B (en) * | 2019-12-11 | 2023-03-24 | 华侨大学 | Self-adaptive lane line detection method |
CN112380956A (en) * | 2020-11-10 | 2021-02-19 | 苏州艾氪英诺机器人科技有限公司 | Lane judgment method |
CN112380956B (en) * | 2020-11-10 | 2024-02-27 | 苏州艾氪英诺机器人科技有限公司 | Lane judging method |
CN112373477A (en) * | 2020-11-23 | 2021-02-19 | 重庆长安汽车股份有限公司 | Redundancy control method for automatic driving system, automobile, controller, and computer-readable storage medium |
CN113033418A (en) * | 2021-03-25 | 2021-06-25 | 南京工程学院 | Lane departure early warning method suitable for mobile terminal |
CN114128461A (en) * | 2021-10-27 | 2022-03-04 | 江汉大学 | Control method of plug seedling transplanting robot and plug seedling transplanting robot |
CN114037977B (en) * | 2022-01-07 | 2022-04-26 | 深圳佑驾创新科技有限公司 | Road vanishing point detection method, device, equipment and storage medium |
CN114037977A (en) * | 2022-01-07 | 2022-02-11 | 深圳佑驾创新科技有限公司 | Road vanishing point detection method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN108647572B (en) | 2022-03-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108647572A (en) | A kind of lane departure warning method based on Hough transformation | |
Hoffman et al. | Vehicle detection fusing 2D visual features | |
CN107025432B (en) | A kind of efficient lane detection tracking and system | |
CN105488454B (en) | Front vehicles detection and ranging based on monocular vision | |
CN104408460B (en) | A kind of lane detection and tracking detection method | |
CN105931220B (en) | Traffic haze visibility detecting method based on dark channel prior Yu minimum image entropy | |
CN110310255B (en) | Point switch notch detection method based on target detection and image processing | |
CN103714538B (en) | road edge detection method, device and vehicle | |
CN109285179A (en) | A kind of motion target tracking method based on multi-feature fusion | |
CN106529493A (en) | Robust multi-lane line detection method based on perspective drawing | |
CN103617412A (en) | Real-time lane line detection method | |
CN106682586A (en) | Method for real-time lane line detection based on vision under complex lighting conditions | |
CN104182756B (en) | Method for detecting barriers in front of vehicles on basis of monocular vision | |
CN104268860B (en) | A kind of method for detecting lane lines | |
CN106887004A (en) | A kind of method for detecting lane lines based on Block- matching | |
CN104318258A (en) | Time domain fuzzy and kalman filter-based lane detection method | |
CN105608429B (en) | Robust method for detecting lane lines based on difference excitation | |
CN103886594B (en) | Road surface line laser rut detection and recognition methods and disposal system | |
CN107097794B (en) | System and method for detecting lane line of road | |
CN108921076B (en) | Pavement crack disease self-adaptive constant false alarm detection method based on image | |
CN103902985B (en) | High-robustness real-time lane detection algorithm based on ROI | |
CN105719283A (en) | Road surface crack image detection method based on Hessian matrix multi-scale filtering | |
CN110992693A (en) | Deep learning-based traffic congestion degree multi-dimensional analysis method | |
CN104992429A (en) | Mountain crack detection method based on image local reinforcement | |
CN117094914B (en) | Smart city road monitoring system based on computer vision |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |