CN102314599A - Identification and deviation-detection method for lane - Google Patents
Identification and deviation-detection method for lane Download PDFInfo
- Publication number
- CN102314599A CN102314599A CN201110306984A CN201110306984A CN102314599A CN 102314599 A CN102314599 A CN 102314599A CN 201110306984 A CN201110306984 A CN 201110306984A CN 201110306984 A CN201110306984 A CN 201110306984A CN 102314599 A CN102314599 A CN 102314599A
- Authority
- CN
- China
- Prior art keywords
- lane
- image
- track
- point
- kalman
- 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
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to an identification and deviation-detection method for a lane, which comprises the following steps: (1) acquiring a lane image and carrying out pretreatment on the lane image; (2) carrying out Canny operator edge detection on the lane image which is subjected to the pretreatment to obtain lane edge images; (3) determining the position of a lane mark according to the obtained lane edge images and a Kalman predictor-based lane tracking method, selecting Kalman prediction areas, filtering out a set of effective points by using a distance discrimination method, and extracting lane parameters on the basis of optimizing the set of effective points; (4) extracting the lane mark by using the Hough conversion with linear fitting according to the obtained lane parameters; and (5) counting the number of background points and lane mark points in the Kalman prediction areas by using the starting point position and the dynamic prediction of a lane which are determined in the step (3), and solving the ratio of the background points to the lane mark points. With the adoption of the identification and deviation-detection method for the lane provided by the invention, the monitoring for lane condition can be rapidly and stably realized.
Description
Technical field
The present invention relates to the Research on Lane Detection Based on Hough Transform field, particularly relate to a kind of lane identification and depart from detection method.
Background technology
Fatigue driving is one of important hidden danger of current traffic safety.The driver is when fatigue, and therefore its perception to surrounding environment, situation judgement and the ability of controlling of vehicle all had decline in various degree be easy to occurrence of traffic accident.In anti-tired safe driving intelligence system, the extraction of lane line with handle as judging the whether important indicator of fatigue of people, be the key link of total system.Therefore, lane line is separated from the picture of track, and handled in real time; Calculate parameter, confirm the current state of vehicle in the track, thereby vehicle is carried out the real-time and effective monitoring; And then driver's state made effective judgement, through reminding the generation that avoids traffic accident.
Retrieval through to prior art finds that Chinese invention patent " a kind of methods, devices and systems of definite deviation " application number is 201010033839.7, and publication No. is CN 101804814A.This patent discloses a kind of detection method of deviation, at first carriageway image is carried out rim detection, obtains the gradient magnitude and the gradient direction of each point in the carriageway image, and then the gradient direction of definite lane boundary.The gradient direction of gradient magnitude and gradient direction and lane boundary that utilizes each pixel obtains the lane boundary straight line with the said border of fitting a straight line.In this recognition methods, lane line is through possibly occurring the situation that straight line too much can't good fit after the conversion, and when having certain width and curvature in the face of the track, said method exists than mistake, state that can not the current track of better recognition.Chinese invention patent " a kind of simulated roadway recognition method of cutting apart based on statistical threshold " application number is 200710168943.5, and publication number is CN 101187976A.Arbitrary number strong point pixel value is made the method for difference in this patent utilization image, adopts black and white boundary threshold value to carry out cutting apart of image, and the percentage that accounts for image through the statistics track is recently calculated threshold value.This method fuzzy deviation possibly occur in the process of handling black and white boundary selection of threshold, the practical application time error is bigger, and can not realize real-time and self-adaptation effect.Chinese invention patent " a kind of be used for the method that the gray level image rapid multi-threshold value is cut apart ", application number is 200810064059.1, publication number is CN 101236607A.It is a kind of based on histogrammic Thresholding Method for Grey Image Segmentation that this invents proposition, because multiobject existence, the grey level histogram that this method is used has multimodal, and the gray scale of two therefore adjacent peak mid points correspondences is as the threshold value of Threshold Segmentation.Because the existence of ruffling, this method is poor in the resistivity that has interference noise or uneven illumination in the face of image, is prone in the selection of threshold than mistake, and application is very limited.Jap.P. " TRAFFIC LANE BOUNDARY DECISION DEVICE "; Application number is JP2005258846A; This patent proposes a kind of judgement RM of track, limits the track and is in the perfect condition, and is relatively harsher to environment requirement; To the variation of light, the reason of weather etc. do not propose effective solution.United States Patent (USP) " Vehicle and Lane Mark Detection Device ", application number is US2009167864A1.This patent proposes a kind of carriageway image processing mode based on the CCD image-forming principle, this mode image resolution ratio low with the light situation of change under, the error of dynamic fuzzy appears easily; Cause that identification error is bigger; And this method is not accomplished real time effect when handling carrying out image threshold segmentation.
Summary of the invention
Technical matters to be solved by this invention provides a kind of lane identification and departs from detection method, makes its fast and stable ground realize the monitoring to the track situation.
The technical solution adopted for the present invention to solve the technical problems is: provide a kind of lane identification to depart from detection method, may further comprise the steps:
(1) obtains carriageway image, and said carriageway image is carried out pre-service;
(2) carry out the Canny operator edge detection to carrying out pretreated carriageway image, obtained the track edge image;
(3) according to the track tracking of the track edge image that obtains based on Kalman's fallout predictor; Determine the position of lane line; Selection card Germania estimation range, the service range diagnostic method filters out effective point set, extracts the track parameter on the basis after point set is optimized at last;
(4), utilize the Hough conversion of band fitting a straight line to extract lane line according to the track parameter that obtains;
(5) utilize the definite starting point position of step (3) and the performance prediction in track, the number of statistics background dot and lane line point in the Kalman estimation range, and ask the ratio between background dot and the lane line point.
Pre-service in the said step (1) also comprises following substep:
(11) carriageway image that obtains being carried out ROI handles;
(12) carriageway image after the ROI processing is carried out ashing treatment;
(13) carriageway image after the ashing treatment being carried out medium filtering handles;
(14) to carriageway image degree of the comparing enhancement process behind the medium filtering;
(15) carriageway image after contrast is strengthened is divided into the m level, and the probability that other pixel of each grade is occurred comes out with histogrammic embodied and analyzes, wherein, and m＞1;
(16) utilize object and the difference of background on gray scale in the carriageway image, obtain binaryzation segmentation threshold first based on priori, and adopt adaptive mode to obtain the threshold value of next time cutting apart automatically, so as to confirming each point in the carriageway image.
Hough conversion in the said step (4) is further comprising the steps of:
(41) confirm polar coordinate system, said track parameter is corresponded in the said polar coordinate system;
(42) each pixel on the carriageway image is carried out the Hough conversion, the polar angle that traversal is had a few calculates the utmost point footpath of being had a few, and directly and in the parameter array of the point of polar angle adds 1 in corresponding same pole;
(43) set the straight length threshold value, obtain the pole coordinate parameter of straight line conversion;
(44) mark straight line according to parameter at carriageway image; If in the pole coordinate parameter that obtains, exist many lane lines to cause the wide or multiple barrier situation of lane line; Then take the mode of track fitting a straight line that wide straight line or multiple barrier are carried out the track match, extract lane line.
Pass through in the said step (5) ratio of last piece image is confirmed Kalman's threshold value of piece image down, and according to obtain the said ratio of piece image down according to this Kalman's threshold value.
Kalman estimation range in the said step (3) is for being the center with said lane line, and width is the zone of five said lane lines.
Beneficial effect
Owing to adopted above-mentioned technical scheme; The present invention compared with prior art; Have following advantage and good effect: the present invention adopts adaptive approach to obtain the threshold value of Hough conversion; And having added the fitting a straight line algorithm, accurate in locating goes out linear position, can carry out better recognition to the lane line that certain width and curvature are arranged.The present invention also adopts a kind of brand-new dynamic threshold binarization method based on priori, can adapt to the variation of state of weather, the adaptive threshold value of obtaining carrying out image threshold segmentation.The present invention utilizes the result of last time location to limit the image-region of the lane line search in current track, the scope that can dwindle the sensitizing range according to the multidate information of motion like this, and both positioning car diatom reliably can improve accuracy, real-time again.
Description of drawings
Fig. 1 is a process flow diagram of the present invention;
Fig. 2 is the pretreated process flow diagram of image among the present invention;
Fig. 3 is the process flow diagram of Hough conversion among the present invention.
Embodiment
Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiment only to be used to the present invention is described and be not used in the restriction scope of the present invention.Should be understood that in addition those skilled in the art can do various changes or modification to the present invention after the content of having read the present invention's instruction, these equivalent form of values fall within the application's appended claims institute restricted portion equally.
Embodiment of the present invention relates to a kind of lane identification and departs from detection method, and is as shown in Figure 1, is divided into following step; The pre-service of carriageway image, Canny operator edge detection, Kalman filter forecasting; The Hough conversion of band fitting a straight line, adaptive mode is chosen the threshold value of Hough conversion.
The pre-treatment step of carriageway image is as shown in Figure 2 among the present invention, comprises image is carried out ROI processing, ashing treatment, The disposal of gentle filter, contrast enhancement processing, the regional dynamic threshold binary conversion treatment that reaches based on priori of histogram analysis.
The ROI processing is based on CCD and thereby the cmos imaging principle is divided into area-of-interest and non-area-of-interest modeling with picture.At first, camera obtains the image based on CCD or CMOS, obtain image after, image is carried out dividing processing, data rule of thumb, the sky composition generally accounts for 5/12 of picture, so we get the basis of below 7/12 part of picture as Flame Image Process.Image is carried out ROI handle, be intended to get approx a separation Y
mAs the separation of nearly visual area with visual area far away.Wherein, nearly visual area is area-of-interest, and visual area far away is non-area-of-interest.In myopia is felt the zone, the approximate straight line of regarding as in track.
Then, carry out ashing treatment, ashing treatment is an intermediary with the LAB pattern, and the RGB picture that camera is obtained converts the LAB pattern into, and then generates corresponding equivalent RGB GTG.That is to say, convert the RGB picture into the LAB pattern, in the LAB pattern, discolor then, and then return the RGB picture and generate the RGB GTG of an equivalence, change to gray space according to this GTG more at last, and generate corresponding gray scale K.The value of gray scale K is 0～255.
There are many noises in image through obtaining after the ashing, adopts median filtering method to handle.Replace the value of any in digital picture or the Serial No. with the Mesophyticum of each point value in the neighborhood (3*3) of this point, the pixel value around letting is near actual value, thereby eliminates isolated noise spot.Utilize from left to right, the sleiding form of structure from top to bottom, the size of display screen interior pixel according to pixel value sorted, generate the data sequence of dull rising the (or decline).But The disposal of gentle filter adopts the mode nonlinear smoothing ground of medium filtering to remove the noise spot that exists after the ashing, and protects object boundary to make it not fuzzy simultaneously.
Degree of comparing enhancement process because the plumpness of gained gray level image is lower behind the medium filtering, adopts the mode of enhancing contrast ratio that the plumpness of image is improved then, helps system to differentiate the position of lane line more clearly.
Afterwards, adopt the histogram analysis zone, histogram analysis is divided into the m level with image, and gray-scale value is that the pixel of i has n, and the probability that other pixel of each grade is occurred comes out with histogrammic embodied and analyzes.
In priori dynamic threshold binary conversion treatment; Utilize object and the difference of its background on gray scale in the image; Be regarded as having the combination in two types of zones of different grey-scale to image, choose an appropriate threshold, so as to confirming each point in the image.At first adopt and carry out first binaryzation based on the method for priori and cut apart, and then confirm segmentation threshold first.When vehicle gets into the track, manually choose the zone of unit area respectively in lane line and track respectively, calculate the average W and the B of gray scale in two zones.Each color grey value characteristics of reference template (being the standard road image of priori) is combined into the vectorial A=[q of a 1 * m
1, q
W..., q
M-1, q
m], each component among so vectorial A is represented the proportion of this color characteristic component in the reference template.The average W of gray scale and B utilize " trough method " to carry out first binaryzation to cut apart in two zones choosing first according to priori.Trough between W and B is that minimum point is illustrated in this gray-scale value, and the probability of picture point is minimum, and then can be used as the foundation of cutting apart, and the gray-scale value of this point is designated as K.Through experimental verification, under illumination and Changes in weather situation, the gray-scale map grey scale change is no more than 18 GTG values, with this benchmark as the next frame image binaryzation.The image that binaryzation is first cut apart is mapped in the original image with the mode of Y=X mapping, gets in the original image gray-scale value of each point of white point in the corresponding binary image, obtains average gray K with the mode of arithmetic mean
W1Being presented as the k level when averaging in the grey level histogram if gray scale counts, is the center with this value so, k in the histogram ± 1, and k ± 2 ..., k ± 9 a group respective items contrasts in twos, obtains the greater in two comparing results respectively, confirms that the group of bigger group of gray-scale value is counted k
m, obtain one group of data k
M1, k
M2..., k
M9, and then confirm this and organize the tonal range of data, be designated as K
W2With K
W2As the benchmark of the binaryzation second time, between K
W2Point in the scope is as the white point of the binaryzation second time, K
W2Outside point as black color dots.Go round and begin again, realized that the adaptivity binary-state threshold is definite.
Be not difficult to find, adopt a kind of dynamic threshold binarization method, can adapt to the variation of state of weather, the adaptive threshold value of obtaining carrying out image threshold segmentation based on priori.
The Canny operator edge detection comes down to do level and smooth computing with an accurate Gaussian function, with the first order differential operator location derivative maximal value of band direction, judges according to three decision principles of Canny operator whether this point is marginal point then then.Specifically; Definition according to Canny; Center edge point is the convolution of operator and the image maximal value in the zone on the gradient direction on the edge of, like this, just can judge on the gradient direction of every bit that whether this intensity be that the maximal value of its neighborhood confirms whether this point is marginal point; When a pixel satisfies following three conditions, then be considered to the edge of image point: 1) edge strength of this point is greater than the edge strength along two adjacent image point points of this gradient direction; 2) with this gradient direction on adjacent 2 direction difference less than 45 degree; 3) be that edge strength maximum value in 3 * 3 neighborhoods at center is less than certain threshold value with this point.Thereby obtain the track edge image through the Canny operator edge detection.
Image is in gatherer process, and intensity of illumination, barrier block, roadside trees and Uneven road are smooth and the camera shake that causes all can impact lane line information in the image.Under such situation, the lane line Parameter Extraction will produce bigger error, the phenomenon in track also possibly occur can't tracking because of the track turn condition.The present invention has introduced the position of determining lane line based on the track tracking of Kalman's fallout predictor; After confirming the lane line position at the vehicle drive initial stage; With the lane line is the center, gets width and be five lane lines (zone that in image, is about 50 pixels) as the Kalman estimation range, follows the trail of and position that predict lane possibly occur; The service range diagnostic method filters out effective point set then, extracts the track parameter on the basis after point set is optimized at last.Through such prediction mode, can better obtain the situation of change in track, accurate than other lane detection modes.
In lane line straight line extraction algorithm, the Hough conversion is one of the most frequently used method, and its advantage is that noiseproof feature is good, and algorithm is stable.Pixel in the image space can be represented with the straight line in the parameter space through the mode of projection.As shown in Figure 3, at first confirm a polar coordinate system, i.e. two-dimensional array buffer zone of initialization is used to deposit parameter plane ρ, and the value of θ is changed to 0 with all data in the array earlier.Then each pixel of road image is carried out the Hough conversion, the θ angle (zone of traversal can be selected as required) that traversal is had a few, promptly polar angle calculates all ρ values, i.e. utmost point footpath.At the identical ρ of correspondence, add 1 in the parameter array of the point of θ.At last, find the more a little bigger position of parameter array on the parameter plane, more a little bigger choosing can be set different threshold values according to the conversion needs and choose, and this position is exactly the parameter of straight line on the corresponding road image.In the parameter that obtains, possibly exist many lane lines to cause situation such as the wide or polygon of lane line, take the mode of track match to carry out the track match here about wide straight line, multiple barrier.After all points of Hough transfer pair added up, a number threshold value that adds up is bad to be confirmed.If threshold value is excessive, when lane line has dotted line, the accumulated value of straight line that has been interrupted some effects of lane line, this moment, lane line maybe be by omission.If threshold value is too small, other trade lines outside the current lane line, and road boundary etc. also can be considered to lane line, need take certain strategy to carry out the extraction of lane line.
Adopt adaptive mode to choose the threshold value of Hough conversion at last, that is to say, through above Kalman filtering prediction steps; After confirming starting point, can obtain performance prediction about the track, selected be the center with the track; Width is in the image sensitizing range of five lane widths; The number of the pixel of assert in statistics background dot and the lane line scope that obtains through the Hough conversion, and ask its ratio T, can know according to ratio T whether this track departs from.Ratio T through to last piece image is definite, and draws Kalman's threshold value Δ H through the Kalman's future position relation with following piece image, obtains down the ratio T ' of piece image then according to this Kalman's threshold value Δ H.
This shows that the present invention adopts adaptive approach to obtain the threshold value of Hough conversion, and has added the fitting a straight line algorithm, accurate in locating goes out linear position, can carry out better recognition to the lane line that certain width and curvature are arranged.Utilize the result of last time location to limit the image-region of the lane line search in current track simultaneously, the scope that can dwindle the sensitizing range according to the multidate information of motion like this, both positioning car diatom reliably can improve accuracy, real-time again.
Claims (5)
1. a lane identification departs from detection method, it is characterized in that, may further comprise the steps:
(1) obtains carriageway image, and said carriageway image is carried out pre-service;
(2) carry out the Canny operator edge detection to carrying out pretreated carriageway image, obtained the track edge image;
(3) according to the track tracking of the track edge image that obtains based on Kalman's fallout predictor; Determine the position of lane line; Selection card Germania estimation range, the service range diagnostic method filters out effective point set, extracts the track parameter on the basis after point set is optimized at last;
(4), utilize the Hough conversion of band fitting a straight line to extract lane line according to the track parameter that obtains;
(5) utilize the definite starting point position of step (3) and the performance prediction in track, the number of statistics background dot and lane line point in the Kalman estimation range, and ask the ratio between background dot and the lane line point.
2. lane identification according to claim 1 departs from detection method, it is characterized in that, the pre-service in the said step (1) also comprises following substep:
(11) carriageway image that obtains being carried out ROI handles;
(12) carriageway image after the ROI processing is carried out ashing treatment;
(13) carriageway image after the ashing treatment being carried out medium filtering handles;
(14) to carriageway image degree of the comparing enhancement process behind the medium filtering;
(15) carriageway image after contrast is strengthened is divided into the m level, and the probability that other pixel of each grade is occurred comes out with histogrammic embodied and analyzes, wherein, and m＞1;
(16) utilize object and the difference of background on gray scale in the carriageway image, obtain binaryzation segmentation threshold first based on priori, and adopt adaptive mode to obtain the threshold value of next time cutting apart automatically, so as to confirming each point in the carriageway image.
3. lane identification according to claim 1 departs from detection method, it is characterized in that, the Hough conversion in the said step (4) is further comprising the steps of:
(41) confirm polar coordinate system, said track parameter is corresponded in the said polar coordinate system;
(42) each pixel on the carriageway image is carried out the Hough conversion, the polar angle that traversal is had a few calculates the utmost point footpath of being had a few, and directly and in the parameter array of the point of polar angle adds 1 in corresponding same pole;
(43) set the straight length threshold value, obtain the pole coordinate parameter of straight line conversion;
(44) mark straight line according to parameter at carriageway image; If in the pole coordinate parameter that obtains, exist many lane lines to cause the wide or multiple barrier situation of lane line; Then take the mode of track fitting a straight line that wide straight line or multiple barrier are carried out the track match, extract lane line.
4. lane identification according to claim 1 departs from detection method; It is characterized in that; Pass through in the said step (5) ratio of last piece image is confirmed Kalman's threshold value of piece image down, and according to obtain the said ratio of piece image down according to this Kalman's threshold value.
5. lane identification according to claim 1 departs from detection method, it is characterized in that, the Kalman estimation range in the said step (3) is for being the center with said lane line, and width is the zone of five said lane lines.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110306984A CN102314599A (en) | 2011-10-11 | 2011-10-11 | Identification and deviation-detection method for lane |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110306984A CN102314599A (en) | 2011-10-11 | 2011-10-11 | Identification and deviation-detection method for lane |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102314599A true CN102314599A (en) | 2012-01-11 |
Family
ID=45427749
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110306984A Pending CN102314599A (en) | 2011-10-11 | 2011-10-11 | Identification and deviation-detection method for lane |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102314599A (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102854878A (en) * | 2012-09-14 | 2013-01-02 | 苏州工业园区永动工业设备有限公司 | Infrared photoelectric guiding automatic guided vehicle (AGV) device and control method thereof |
| CN102997853A (en) * | 2012-11-22 | 2013-03-27 | 华中科技大学 | Device and method for detecting ice and snow thickness |
| CN103714538A (en) * | 2013-12-20 | 2014-04-09 | 中联重科股份有限公司 | Road edge detection method, device and vehicle |
| CN103761450A (en) * | 2014-02-24 | 2014-04-30 | 中国石油大学(华东) | Dynamic process fault forecasting method based on fuzzy self-adaptive prediction |
| CN104063877A (en) * | 2014-07-16 | 2014-09-24 | 中电海康集团有限公司 | Hybrid judgment identification method for candidate lane lines |
| CN104063691A (en) * | 2014-06-27 | 2014-09-24 | 广东工业大学 | Lane line fast detection method based on improved Hough transform |
| CN104091166A (en) * | 2014-07-07 | 2014-10-08 | 大连理工大学 | Dynamic threshold method for lane line edge extraction |
| CN104704532A (en) * | 2012-09-14 | 2015-06-10 | 西门子公司 | A Hough transform approach to gap measurement in blade inspection |
| CN105760812A (en) * | 2016-01-15 | 2016-07-13 | 北京工业大学 | Hough transform-based lane line detection method |
| CN106326850A (en) * | 2016-08-18 | 2017-01-11 | 宁波傲视智绘光电科技有限公司 | Fast lane line detection method |
| CN106354135A (en) * | 2016-09-19 | 2017-01-25 | 武汉依迅电子信息技术有限公司 | Lane keeping system and method based on Beidou high-precision positioning |
| CN107179197A (en) * | 2016-03-11 | 2017-09-19 | 深圳市航盛电子股份有限公司 | A kind of lane shift warning threshold recognition capability analog detection method and test system |
| CN107451539A (en) * | 2017-07-13 | 2017-12-08 | 华域汽车***股份有限公司 | A kind of lane departure warning method and system |
| CN108304749A (en) * | 2017-01-13 | 2018-07-20 | 比亚迪股份有限公司 | The recognition methods of road speed line, device and vehicle |
| CN109035684A (en) * | 2018-04-14 | 2018-12-18 | 包仁妹 | Position of human body deviates measurement method |
| CN109635737A (en) * | 2018-12-12 | 2019-04-16 | 中国地质大学（武汉） | Automobile navigation localization method is assisted based on pavement marker line visual identity |
| CN110889342A (en) * | 2019-11-13 | 2020-03-17 | 江苏理工学院 | Deceleration strip identification method |
| CN111611830A (en) * | 2019-02-26 | 2020-09-01 | 阿里巴巴集团控股有限公司 | Lane line identification method, device and system and vehicle |
| CN111652033A (en) * | 2019-12-12 | 2020-09-11 | 苏州奥易克斯汽车电子有限公司 | Lane line detection method based on OpenCV |
| CN112287888A (en) * | 2020-11-20 | 2021-01-29 | 中国铁建电气化局集团第二工程有限公司 | Track turning identification method based on prediction weight |
| CN112329631A (en) * | 2020-11-05 | 2021-02-05 | 浙江点辰航空科技有限公司 | Method for carrying out traffic flow statistics on expressway by using unmanned aerial vehicle |
| CN112784633A (en) * | 2019-11-07 | 2021-05-11 | 北京四维图新科技股份有限公司 | Method and device for processing lane boundary, electronic equipment and storage medium |
| CN113408407A (en) * | 2018-10-31 | 2021-09-17 | 百度在线网络技术（北京）有限公司 | Electronic map lane line correction method, device and computer readable storage medium |
| CN113723500A (en) * | 2021-08-27 | 2021-11-30 | 四川启睿克科技有限公司 | Image data expansion method based on feature similarity and linear smoothing combination |
| CN113793356A (en) * | 2018-09-30 | 2021-12-14 | 百度在线网络技术（北京）有限公司 | Lane line detection method and device |
| CN118097581A (en) * | 2024-04-28 | 2024-05-28 | 山东领军智能交通科技有限公司 | Road edge recognition control method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201427553Y (en) * | 2009-06-26 | 2010-03-24 | 长安大学 | Alarm system of vehicle departure from lane |
| CN101992778A (en) * | 2010-08-24 | 2011-03-30 | 上海科世达-华阳汽车电器有限公司 | Lane deviation early warning and driving recording system and method |
-
2011
- 2011-10-11 CN CN201110306984A patent/CN102314599A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201427553Y (en) * | 2009-06-26 | 2010-03-24 | 长安大学 | Alarm system of vehicle departure from lane |
| CN101992778A (en) * | 2010-08-24 | 2011-03-30 | 上海科世达-华阳汽车电器有限公司 | Lane deviation early warning and driving recording system and method |
Non-Patent Citations (3)
| Title |
|---|
| 《微处理机》 20110630 徐后杰 等 基于单目视觉的车道线分离警告算法研究 第73-74页 1-5 , * |
| 徐后杰 等: "基于单目视觉的车道线分离警告算法研究", 《微处理机》 * |
| 鄢治国 等: "一种基于自适应二值化阈值的焊缝边缘特征提取方法", 《焊接学报》 * |
Cited By (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102854878B (en) * | 2012-09-14 | 2015-07-22 | 苏州工业园区永动工业设备有限公司 | Infrared photoelectric guiding automatic guided vehicle (AGV) device and control method thereof |
| CN102854878A (en) * | 2012-09-14 | 2013-01-02 | 苏州工业园区永动工业设备有限公司 | Infrared photoelectric guiding automatic guided vehicle (AGV) device and control method thereof |
| CN104704532A (en) * | 2012-09-14 | 2015-06-10 | 西门子公司 | A Hough transform approach to gap measurement in blade inspection |
| CN102997853A (en) * | 2012-11-22 | 2013-03-27 | 华中科技大学 | Device and method for detecting ice and snow thickness |
| CN103714538A (en) * | 2013-12-20 | 2014-04-09 | 中联重科股份有限公司 | Road edge detection method, device and vehicle |
| CN103714538B (en) * | 2013-12-20 | 2016-12-28 | 中联重科股份有限公司 | road edge detection method, device and vehicle |
| CN103761450A (en) * | 2014-02-24 | 2014-04-30 | 中国石油大学(华东) | Dynamic process fault forecasting method based on fuzzy self-adaptive prediction |
| CN103761450B (en) * | 2014-02-24 | 2016-08-24 | 中国石油大学(华东) | A kind of dynamic process fault prediction method based on fuzzy self-adaption prediction |
| CN104063691A (en) * | 2014-06-27 | 2014-09-24 | 广东工业大学 | Lane line fast detection method based on improved Hough transform |
| CN104063691B (en) * | 2014-06-27 | 2017-08-25 | 广东工业大学 | Lane line quick determination method based on improved Hough transform |
| CN104091166A (en) * | 2014-07-07 | 2014-10-08 | 大连理工大学 | Dynamic threshold method for lane line edge extraction |
| CN104091166B (en) * | 2014-07-07 | 2017-05-03 | 大连理工大学 | Dynamic threshold method for lane line edge extraction |
| CN104063877A (en) * | 2014-07-16 | 2014-09-24 | 中电海康集团有限公司 | Hybrid judgment identification method for candidate lane lines |
| CN104063877B (en) * | 2014-07-16 | 2017-05-24 | 中电海康集团有限公司 | Hybrid judgment identification method for candidate lane lines |
| CN105760812B (en) * | 2016-01-15 | 2019-06-07 | 北京工业大学 | A kind of method for detecting lane lines based on Hough transform |
| CN105760812A (en) * | 2016-01-15 | 2016-07-13 | 北京工业大学 | Hough transform-based lane line detection method |
| CN107179197B (en) * | 2016-03-11 | 2023-11-07 | 深圳市航盛电子股份有限公司 | Lane deviation early warning limit recognition capability simulation test method and test system |
| CN107179197A (en) * | 2016-03-11 | 2017-09-19 | 深圳市航盛电子股份有限公司 | A kind of lane shift warning threshold recognition capability analog detection method and test system |
| CN106326850A (en) * | 2016-08-18 | 2017-01-11 | 宁波傲视智绘光电科技有限公司 | Fast lane line detection method |
| CN106354135A (en) * | 2016-09-19 | 2017-01-25 | 武汉依迅电子信息技术有限公司 | Lane keeping system and method based on Beidou high-precision positioning |
| CN108304749A (en) * | 2017-01-13 | 2018-07-20 | 比亚迪股份有限公司 | The recognition methods of road speed line, device and vehicle |
| CN107451539A (en) * | 2017-07-13 | 2017-12-08 | 华域汽车***股份有限公司 | A kind of lane departure warning method and system |
| CN107451539B (en) * | 2017-07-13 | 2020-11-27 | 华域汽车***股份有限公司 | Lane departure early warning method and system |
| CN109035684A (en) * | 2018-04-14 | 2018-12-18 | 包仁妹 | Position of human body deviates measurement method |
| CN113792690A (en) * | 2018-09-30 | 2021-12-14 | 百度在线网络技术（北京）有限公司 | Lane line detection method and device |
| CN113793356A (en) * | 2018-09-30 | 2021-12-14 | 百度在线网络技术（北京）有限公司 | Lane line detection method and device |
| CN113408407A (en) * | 2018-10-31 | 2021-09-17 | 百度在线网络技术（北京）有限公司 | Electronic map lane line correction method, device and computer readable storage medium |
| CN113408407B (en) * | 2018-10-31 | 2023-06-23 | 百度在线网络技术（北京）有限公司 | Electronic map lane line correction method, electronic map lane line correction equipment and computer readable storage medium |
| CN109635737A (en) * | 2018-12-12 | 2019-04-16 | 中国地质大学（武汉） | Automobile navigation localization method is assisted based on pavement marker line visual identity |
| CN111611830A (en) * | 2019-02-26 | 2020-09-01 | 阿里巴巴集团控股有限公司 | Lane line identification method, device and system and vehicle |
| CN112784633A (en) * | 2019-11-07 | 2021-05-11 | 北京四维图新科技股份有限公司 | Method and device for processing lane boundary, electronic equipment and storage medium |
| CN112784633B (en) * | 2019-11-07 | 2024-04-19 | 北京四维图新科技股份有限公司 | Lane boundary processing method and device, electronic equipment and storage medium |
| CN110889342A (en) * | 2019-11-13 | 2020-03-17 | 江苏理工学院 | Deceleration strip identification method |
| CN111652033A (en) * | 2019-12-12 | 2020-09-11 | 苏州奥易克斯汽车电子有限公司 | Lane line detection method based on OpenCV |
| CN112329631A (en) * | 2020-11-05 | 2021-02-05 | 浙江点辰航空科技有限公司 | Method for carrying out traffic flow statistics on expressway by using unmanned aerial vehicle |
| CN112287888A (en) * | 2020-11-20 | 2021-01-29 | 中国铁建电气化局集团第二工程有限公司 | Track turning identification method based on prediction weight |
| CN112287888B (en) * | 2020-11-20 | 2024-06-18 | 中国铁建电气化局集团第二工程有限公司 | Track turning recognition method based on predictive weight |
| CN113723500A (en) * | 2021-08-27 | 2021-11-30 | 四川启睿克科技有限公司 | Image data expansion method based on feature similarity and linear smoothing combination |
| CN113723500B (en) * | 2021-08-27 | 2023-06-16 | 四川启睿克科技有限公司 | Image data expansion method based on combination of feature similarity and linear smoothing |
| CN118097581A (en) * | 2024-04-28 | 2024-05-28 | 山东领军智能交通科技有限公司 | Road edge recognition control method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102314599A (en) | Identification and deviation-detection method for lane | |
| CN107045629B (en) | Multi-lane line detection method | |
| CN109657632B (en) | Lane line detection and identification method | |
| Aly | Real time detection of lane markers in urban streets | |
| CN101739829B (en) | Video-based vehicle overspeed monitoring method and system | |
| CN102708356B (en) | Automatic license plate positioning and recognition method based on complex background | |
| CN101950350B (en) | Clear path detection using a hierachical approach | |
| CN109255350B (en) | New energy license plate detection method based on video monitoring | |
| CN104217215A (en) | Classification and identification method for foggy water surface image and clear water surface image | |
| CN106022243B (en) | A kind of retrograde recognition methods of the car lane vehicle based on image procossing | |
| CN103324930A (en) | License plate character segmentation method based on grey level histogram binaryzation | |
| CN102411710A (en) | Vehicle type recognition method based on vehicle face features | |
| CN108921076B (en) | Pavement crack disease self-adaptive constant false alarm detection method based on image | |
| KR101224027B1 (en) | Method for dectecting front vehicle using scene information of image | |
| CN202134079U (en) | Unmanned vehicle lane marker line identification and alarm device | |
| CN109886168B (en) | Ground traffic sign identification method based on hierarchy | |
| CN104915642B (en) | Front vehicles distance measuring method and device | |
| JP2013058199A (en) | Method and system for detecting vehicle position by employing polarization image | |
| CN113239733B (en) | Multi-lane line detection method | |
| CN101369312B (en) | Method and equipment for detecting intersection in image | |
| Ma et al. | An all-weather lane detection system based on simulation interaction platform | |
| CN111753749A (en) | Lane line detection method based on feature matching | |
| CN111652033A (en) | Lane line detection method based on OpenCV | |
| KR101794632B1 (en) | The logo pattern recognizing method in number plate of electric vehicle | |
| CN109800641B (en) | Lane line detection method based on threshold value self-adaptive binarization and connected domain analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120111 |