CN106803061A - A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest - Google Patents
A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest Download PDFInfo
- Publication number
- CN106803061A CN106803061A CN201611151545.8A CN201611151545A CN106803061A CN 106803061 A CN106803061 A CN 106803061A CN 201611151545 A CN201611151545 A CN 201611151545A CN 106803061 A CN106803061 A CN 106803061A
- Authority
- CN
- China
- Prior art keywords
- line
- lane
- interest
- image
- previous frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Simple and fast method for detecting lane lines the present invention relates to be based on dynamic area-of-interest, to original image pre-process and obtains edge detection results figure;Calculate dynamic area-of-interest:The lane detection result of previous frame image is obtained, the lane detection result generation mask figure according to previous frame image;Made and operated with mask figure and edge detection results figure, obtain only existing the topography of dynamic area-of-interest;Probability Hough transformation detection of straight lines is used in topography;All straight lines are traveled through, its length is calculated, this straight line is skipped if length is less than threshold value;Otherwise calculate its distance for arriving image base midpoint;Straight line collection after traversal sequence, calculates straight slope, color and the extent of deviation with previous frame track line position;If extent of deviation is all in respective scope, using the straight line as this frame lane line;Otherwise continue to search for backward.The present invention can preferably meet requirement of the vision navigation system to versatility, real-time and robustness.
Description
Technical field
The present invention relates to lane detection technology, specially a kind of simple and fast lane line based on dynamic area-of-interest
Detection method.
Background technology
With the development of intelligent vehicle, and people, for increasing that traffic safety is paid close attention to, lane detection is used as intelligence
The important component that vehicle environmental is perceived has become study hotspot.Autonomous navigation system, track keep accessory system,
In lane departure warning system and track lane-changing assistance system, lane detection all plays a key effect, and only understands exactly
The particular location of lane line, autonomous navigation system can just cook up rational driving path, vehicle could correctly make traveling,
Overtake other vehicles, the driving behavior of lane-change.
At present, lane detection strategy mainly has three kinds:The first is based on GPS+MAP, determines oneself at which according to positioning
The data of travel are wanted in individual track, first collection, are made map;This strategy is relatively costly, but good reliability.Second
Kind based on identification, data source come from laser or video, in video according to all kinds of image procossings algorithm identification lane line, algorithm compared with
It is many and complicated, lane line is recognized according to strength information in laser.The third is based on avoidance, and data source comes from laser or video, all
It is detection barrier, it is determined that can traffic areas.In general, every strategy based on video, cost all than relatively low, but to calculating
Method and environmental requirement are higher.Every strategy based on laser, is all high cost when laser price does not lower, but right
Environmental requirement is general.
The content of the invention
The present invention is based on the complicated shortcoming of the tactful high cost of laser, the policing algorithm based on video for prior art,
A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest is proposed, it has realization simple, efficiency high, general
The characteristics of property is good, is very different with other recognizers, can preferably meet vision navigation system to versatility, real-time
With the requirement of robustness.
The technical scheme that the present invention takes is:A kind of simple and fast lane detection side based on dynamic area-of-interest
Method, comprises the following steps:
S1, original image is filtered, the image preprocessing of gray processing, rim detection, obtain edge detection results figure;
S2, the dynamic area-of-interest of calculating:The lane detection result of previous frame image is obtained, according to previous frame image
Lane detection result generates mask figure;Make logical "and" with mask figure and edge detection results figure to operate, obtain only existing dynamic
The topography of state area-of-interest as subsequent treatment data;
S3, in the topography for only existing dynamic area-of-interest use probability Hough transformation detection of straight lines;
The all straight lines obtained by the detection of probability Hough transformation of S4, traversal, calculate its length, if length is less than threshold value,
Then skip this straight line;Otherwise, its distance for arriving image base midpoint is calculated, and the distance at image base midpoint is arrived by it, near
To remote sequence;
S5, the distance by straight line to image base midpoint, the straight line collection from after closely being sorted to remote traversal, calculate the oblique of straight line
Rate, color and the extent of deviation with previous frame track line position;If the slope of the straight line, color and with previous frame lane line position
The extent of deviation put, then using the straight line as the lane line of this frame, stops search all in respective scope;Otherwise continue to
After search for.
Preferably, the process of generation mask figure is described in step S2:If previous frame image detection less than lane line or this
One two field picture is the first two field picture, then take full images Area generation mask figure;If previous frame image only detects a track
Line, then make the outside expansion parallel lines of lane line, takes expansion parallel lines lower zone as mask figure;If previous frame image is examined
Two lane lines are measured, has then made two outside expansion lines of lane line respectively, asked two intersection points of expansion line, two expansion lines
It is intersecting to divide an image into four regions, two regions expanded between line intersection point and image base are taken as mask figure.
Compared with prior art, the present invention has the advantages that:
1st, the present invention is using dynamic area-of-interest algorithm, the continuity for making full use of lane line to be distributed, by previous frame
Testing result determine the dynamic area-of-interest of this detection, can quickly determine track region, reduce hunting zone, from
And greatly reduce the complexity and detection time of subsequent treatment.
2nd, the present invention using based on lane line feature lane line screening technique, using distance as the search of lane line according to
According to being empirically derived track slope range, color gamut first, and determine the deviation range of lane line and previous frame;Then,
Since the midpoint of image base according to distance from closely to remote outwards search, calculate the slope of straight line, color and with previous frame track
The extent of deviation of line position, if the slope of this straight line, color and with the extent of deviation of previous frame track line position all respective
In the range of, then using this straight line as this frame lane line.Without being suspected to be that lane line is calculated one by one to all of, shorten
Calculating time, improve efficiency of algorithm.
3rd, due to equipment (such as camera lens deformity), the influence of environment (such as illumination, rainwater) factor, the picture for being gathered unavoidably can
Produce to the noisy picture noise of testing result, the present invention is using filtering, gray processing, the image preprocessing stream of edge extracting
Journey is effective except making an uproar to carry out to picture.
Brief description of the drawings
Fig. 1 is overhaul flow chart of the invention;
Fig. 2 illustrates to set the process of dynamic area-of-interest.
Specific embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention will be further described, but specific embodiment of the invention is not limited
In this.
Embodiment
Simple and fast method for detecting lane lines of the present invention based on dynamic area-of-interest, based on opencv image procossings
Storehouse, can realize the detection of intelligent vehicle lane line on puduino platforms.The present invention mainly carries out gray scale by track picture
Change, binaryzation, Canny rim detections are reducing picture noise;And the feature being distributed according to track, from the position of previous frame lane line
Confidence breath calculates detection zone of the dynamic area-of-interest as next frame, to improve efficiency;Use probability Hough transformation
(PPHT) all possible lane line is detected within a detection region, is finally fitted lane line with order-1 linear equation, rule of thumb
Obtain track slope range, color gamut, and determine the deviation range of lane line and previous frame, the root since the midpoint of image base
According to distance from closely outwards being searched for remote, slope, color and the extent of deviation with previous frame track line position of straight line are calculated, if
The slope of this straight line, color and with the extent of deviation of previous frame track line position all in respective scope, then this straight line is made
It is the lane line of this frame.As shown in figure 1, the present invention specifically includes following steps:
First, start lane detection system, initialize camera;Original image is continuously acquired by camera;To original
Image carries out gaussian filtering, eliminates picture noise, and it then is converted into gray-scale map from RGB image, finally uses Canny algorithms
Rim detection is carried out to gray-scale map, edge image is obtained.
The present invention is using filtering, gray processing, the image preprocessing flow of rim detection.Due to equipment (such as camera lens deformity),
Environment (such as illumination, rainwater) factor is influenceed, and the picture for being gathered can be produced to the noisy picture noise of testing result unavoidably,
Image preprocessing can carry out effective except making an uproar to picture.Using gaussian filtering (Gaussian filter), this most has the present invention
With wave filter carry out except making an uproar, general principle is each pixel and Gaussian kernel convolution that will be input into array, by convolution
With as output pixel value.In order to reduce original image data amount, amount of calculation is less when being easy to subsequent treatment, and we will obtain
Image carries out gray processing, and RGB image is converted into gray-scale map.Gray processing is carried out for RGB image, popular point says to be exactly to image
Tri- components of RGB be weighted and averagely obtain final gray value.After obtaining gray-scale map, we will on this basis carry out side
Edge detects that the edge detection algorithm that the present invention is used is Canny edge detection algorithms, and it is that John F.Canny were opened in 1986
The multistage edge detection algorithm for issuing, it has the characteristics of low error rate, polarization high and minimum response, by many people
It is considered the optimal algorithm of rim detection, is widely used in rim detection.
2nd, dynamic area-of-interest is calculated:As shown in Fig. 2 obtaining the lane detection result of previous frame image;According to upper
The lane detection result generation mask figure of one two field picture;If previous frame image detection is less than lane line or this two field picture
First two field picture, then take full images Area generation mask figure;If it (is probably leakage that previous frame image only detects a lane line
Sentence or image in an only lane line really), then make the outside expansion parallel lines of lane line, take this line lower zone as covering
Code figure;If previous frame image detection makees two outside expansion lines of lane line respectively to two lane lines, two are asked to open up
The intersection point of show line, two are expanded that lines are intersecting to divide an image into four regions, are taken two and are expanded between line intersection points and image base
That region as mask figure.Make logical "and" with mask figure and Image Edge-Detection result figure to operate, obtain only existing dynamic
The topography of state area-of-interest as subsequent treatment data.
Wherein, mask figure is the single channel binary map being made up of 0-255.For example, mask is doing logical "and" fortune with picture a
During calculation, if it is 255 that the pixel coordinate pair on picture a answers the value of the pixel on mask figure, retain this picture in picture a
The value of vegetarian refreshments, otherwise zero setting.
According to the definition of lane line, it is known that the feature of lane line has:Positioned at track both sides, in the short time, position is not
Become.Lane line is parallel, but in the picture and non-parallel in real track, but intersecting.Two phases of lane line
Picture can be divided into four pieces by intersection point, and take picture bottom one piece just can obtain a dynamic area-of-interest.If this two
Lane line is the nearest lane line in vehicle both sides, then this dynamic area-of-interest must include track.If we are in this base
By area-of-interest edge to external expansion on plinth, then area-of-interest now includes track and lane line, and eliminates week
Enclose independent environment, and then greatly reduce the hunting zone of straight-line detection, shorten the calculating time.It can be seen that, the present invention is used
Dynamic area-of-interest algorithm, the continuity for making full use of lane line to be distributed determines that this is examined by the testing result of previous frame
The dynamic area-of-interest of survey, can quickly determine track region, hunting zone be reduced, so as to greatly reduce subsequent treatment
Complexity and detection time.
3rd, probability Hough (hough) change detection straight line is used in the topography for only existing dynamic area-of-interest.
4th, all straight lines obtained by the detection of probability Hough transformation are traveled through, its length is calculated, if length is less than threshold value,
Then skip this straight line;Otherwise, its distance for arriving image base midpoint is calculated, and the distance at image base midpoint is arrived by it, near
To remote sequence.
Wherein, position of the straight line on coordinate is referred to " to the distance at image base midpoint ", the straight line is being only existed
What the position in the topography and Image Edge-Detection result figure of dynamic area-of-interest was just as, therefore when calculating
Directly can be carried out in topography.Because camera is arranged on vehicle interior, and vehicle is usually to travel in the middle of track,
So distance of the lane line that is travelled of general vehicle to image base midpoint, is most short compared to other lane lines.
5th, by the distance of straight line to image base midpoint, the straight line collection from after closely being sorted to remote traversal calculates the oblique of straight line
Rate, color and the extent of deviation with previous frame track line position, if the slope of this straight line, color and with previous frame lane line position
The extent of deviation put, then using this straight line as the lane line of this frame, stops search all in respective scope;Otherwise continue to
After search for.
The present invention is using the lane line screening technique based on lane line feature.Enter driveway line by probability Hough transformation to examine
Survey, many straight lines for being suspected to be lane line can be obtained, next differentiated using the lane line feature of definition, multi straight of comforming
In find out correct lane line.But due to being suspected to be that lane line is very more sometimes, if a rule goes to calculate, undoubtedly increase many
Time, reduce efficiency of algorithm.For the consideration of driving safety, the track region that we finally obtain is necessarily equal to or is contained in true
Real vehicle road region.So present invention selection is empirically derived track slope range, color first using distance as search foundation
Scope, and determine the deviation range of lane line and previous frame;Then, since the midpoint of image base according to distance from closely to far to
Outer search, calculates slope, color and the extent of deviation with previous frame track line position of straight line, if the slope of this straight line, face
Color and with the extent of deviation of previous frame track line position all in respective scope, then using this straight line as the track of this frame
Line.
Above-described embodiment is the present invention preferably implementation method, but embodiments of the present invention are not by above-described embodiment
Limitation, it is other it is any without departing from Spirit Essence of the invention and the change, modification, replacement made under principle, combine, simplification,
Equivalent substitute mode is should be, is included within protection scope of the present invention.
Claims (5)
1. a kind of simple and fast method for detecting lane lines based on dynamic area-of-interest, it is characterised in that comprise the following steps:
S1, original image is filtered, the image preprocessing of gray processing, rim detection, obtain edge detection results figure;
S2, the dynamic area-of-interest of calculating:The lane detection result of previous frame image is obtained, according to the track of previous frame image
Line testing result generates mask figure;Make logical "and" with mask figure and edge detection results figure to operate, obtain only existing dynamic sense
The topography in interest region as subsequent treatment data;
S3, in the topography for only existing dynamic area-of-interest use probability Hough transformation detection of straight lines;
The all straight lines obtained by the detection of probability Hough transformation of S4, traversal, calculate its length, if length is less than threshold value, jump
Cross this straight line;Otherwise, its distance for arriving image base midpoint is calculated, and the distance at image base midpoint is arrived by it, from closely to remote
Sequence;
S5, the distance by straight line to image base midpoint, the straight line collection from after closely being sorted to remote traversal, the slope of calculating straight line,
Color and the extent of deviation with previous frame track line position;If the slope of the straight line, color and with previous frame track line position
Extent of deviation all in respective scope, then using the straight line as the lane line of this frame, stop search;Otherwise continue backward
Search.
2. the simple and fast method for detecting lane lines based on dynamic area-of-interest according to claim 1, its feature exists
In the process that mask figure is generated described in step S2 is:If previous frame image detection is less than lane line or this two field picture
One two field picture, then take full images Area generation mask figure;If previous frame image only detects a lane line, make lane line
Outside expansion parallel lines, take expansion parallel lines lower zone as mask figure;If previous frame image detection is to two cars
Diatom, then make two outside expansion lines of lane line respectively, seeks two intersection points of expansion line, expands line for two and intersects image stroke
It is divided into four regions, takes two regions expanded between line intersection point and image base as mask figure.
3. the simple and fast method for detecting lane lines based on dynamic area-of-interest according to claim 1, its feature exists
In mask figure described in step S2 is the single channel binary map being made up of 0-255.
4. the simple and fast method for detecting lane lines based on dynamic area-of-interest according to claim 1, its feature exists
In being filtered into gaussian filtering described in step S1.
5. the simple and fast method for detecting lane lines based on dynamic area-of-interest according to claim 1, its feature exists
In rim detection described in step S1 uses Canny edge detection algorithms.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611151545.8A CN106803061A (en) | 2016-12-14 | 2016-12-14 | A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611151545.8A CN106803061A (en) | 2016-12-14 | 2016-12-14 | A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106803061A true CN106803061A (en) | 2017-06-06 |
Family
ID=58984908
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611151545.8A Pending CN106803061A (en) | 2016-12-14 | 2016-12-14 | A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106803061A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107451566A (en) * | 2017-08-02 | 2017-12-08 | 海信集团有限公司 | Display methods, device and the computer-readable recording medium of lane line |
CN108776776A (en) * | 2018-05-25 | 2018-11-09 | 河南思维轨道交通技术研究院有限公司 | A kind of recognition methods for horizontal vertical line segment in image |
CN110147698A (en) * | 2018-02-13 | 2019-08-20 | Kpit技术有限责任公司 | System and method for lane detection |
CN110171263A (en) * | 2019-05-19 | 2019-08-27 | 瑞立集团瑞安汽车零部件有限公司 | A kind of bend identification and overall height adjusting method for ECAS system |
CN110209924A (en) * | 2018-07-26 | 2019-09-06 | 腾讯科技(深圳)有限公司 | Recommended parameter acquisition methods, device, server and storage medium |
CN110667581A (en) * | 2018-07-02 | 2020-01-10 | 上汽通用汽车有限公司 | Automatic lane change control system and automatic lane change control method for vehicle |
CN110930459A (en) * | 2019-10-29 | 2020-03-27 | 北京经纬恒润科技有限公司 | Vanishing point extraction method, camera calibration method and storage medium |
CN111721216A (en) * | 2020-06-29 | 2020-09-29 | 河南科技大学 | Steel wire rope detection device based on three-dimensional image, surface damage detection method and rope diameter calculation method |
CN112365448A (en) * | 2020-10-20 | 2021-02-12 | 天津大学 | Fabric defect detection method in warp knitting process |
CN112434621A (en) * | 2020-11-27 | 2021-03-02 | 武汉极目智能技术有限公司 | Method for extracting characteristics of inner side edge of lane line |
CN116047537A (en) * | 2022-12-05 | 2023-05-02 | 北京中科东信科技有限公司 | Road information generation method and system based on laser radar |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103617412A (en) * | 2013-10-31 | 2014-03-05 | 电子科技大学 | Real-time lane line detection method |
CN103632140A (en) * | 2013-11-27 | 2014-03-12 | 智慧城市***服务(中国)有限公司 | Method and device for detecting lane line |
CN104036253A (en) * | 2014-06-20 | 2014-09-10 | 智慧城市***服务(中国)有限公司 | Lane line tracking method and lane line tracking system |
CN104063691A (en) * | 2014-06-27 | 2014-09-24 | 广东工业大学 | Lane line fast detection method based on improved Hough transform |
CN104268860A (en) * | 2014-09-17 | 2015-01-07 | 电子科技大学 | Lane line detection method |
CN104408460A (en) * | 2014-09-17 | 2015-03-11 | 电子科技大学 | A lane line detecting and tracking and detecting method |
CN105678791A (en) * | 2016-02-24 | 2016-06-15 | 西安交通大学 | Lane line detection and tracking method based on parameter non-uniqueness property |
CN105966398A (en) * | 2016-06-21 | 2016-09-28 | 广州鹰瞰信息科技有限公司 | Method and device for early warning lane departure of vehicle |
-
2016
- 2016-12-14 CN CN201611151545.8A patent/CN106803061A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103617412A (en) * | 2013-10-31 | 2014-03-05 | 电子科技大学 | Real-time lane line detection method |
CN103632140A (en) * | 2013-11-27 | 2014-03-12 | 智慧城市***服务(中国)有限公司 | Method and device for detecting lane line |
CN104036253A (en) * | 2014-06-20 | 2014-09-10 | 智慧城市***服务(中国)有限公司 | Lane line tracking method and lane line tracking system |
CN104063691A (en) * | 2014-06-27 | 2014-09-24 | 广东工业大学 | Lane line fast detection method based on improved Hough transform |
CN104268860A (en) * | 2014-09-17 | 2015-01-07 | 电子科技大学 | Lane line detection method |
CN104408460A (en) * | 2014-09-17 | 2015-03-11 | 电子科技大学 | A lane line detecting and tracking and detecting method |
CN105678791A (en) * | 2016-02-24 | 2016-06-15 | 西安交通大学 | Lane line detection and tracking method based on parameter non-uniqueness property |
CN105966398A (en) * | 2016-06-21 | 2016-09-28 | 广州鹰瞰信息科技有限公司 | Method and device for early warning lane departure of vehicle |
Non-Patent Citations (4)
Title |
---|
席军强: "《车辆信息技术》", 31 December 2013, 北京理工大学出版社 * |
张军平 等: "基于视频图像的车道线识别方法", 《视频应用与工程》 * |
杨喜宁 等: "基于改进Hough变换的车道线检测技术", 《计算机测量与控制》 * |
陈军 等: "基于概率霍夫变换的车道检测技术研究", 《科技通报》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107451566A (en) * | 2017-08-02 | 2017-12-08 | 海信集团有限公司 | Display methods, device and the computer-readable recording medium of lane line |
CN110147698A (en) * | 2018-02-13 | 2019-08-20 | Kpit技术有限责任公司 | System and method for lane detection |
CN108776776A (en) * | 2018-05-25 | 2018-11-09 | 河南思维轨道交通技术研究院有限公司 | A kind of recognition methods for horizontal vertical line segment in image |
CN110667581B (en) * | 2018-07-02 | 2021-04-16 | 上汽通用汽车有限公司 | Automatic lane change control system and automatic lane change control method for vehicle |
CN110667581A (en) * | 2018-07-02 | 2020-01-10 | 上汽通用汽车有限公司 | Automatic lane change control system and automatic lane change control method for vehicle |
CN110209924A (en) * | 2018-07-26 | 2019-09-06 | 腾讯科技(深圳)有限公司 | Recommended parameter acquisition methods, device, server and storage medium |
CN110171263A (en) * | 2019-05-19 | 2019-08-27 | 瑞立集团瑞安汽车零部件有限公司 | A kind of bend identification and overall height adjusting method for ECAS system |
CN110930459A (en) * | 2019-10-29 | 2020-03-27 | 北京经纬恒润科技有限公司 | Vanishing point extraction method, camera calibration method and storage medium |
CN110930459B (en) * | 2019-10-29 | 2023-02-17 | 北京经纬恒润科技股份有限公司 | Vanishing point extraction method, camera calibration method and storage medium |
CN111721216A (en) * | 2020-06-29 | 2020-09-29 | 河南科技大学 | Steel wire rope detection device based on three-dimensional image, surface damage detection method and rope diameter calculation method |
CN112365448A (en) * | 2020-10-20 | 2021-02-12 | 天津大学 | Fabric defect detection method in warp knitting process |
CN112434621A (en) * | 2020-11-27 | 2021-03-02 | 武汉极目智能技术有限公司 | Method for extracting characteristics of inner side edge of lane line |
CN116047537A (en) * | 2022-12-05 | 2023-05-02 | 北京中科东信科技有限公司 | Road information generation method and system based on laser radar |
CN116047537B (en) * | 2022-12-05 | 2023-12-26 | 北京中科东信科技有限公司 | Road information generation method and system based on laser radar |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106803061A (en) | A kind of simple and fast method for detecting lane lines based on dynamic area-of-interest | |
CN110501018B (en) | Traffic sign information acquisition method for high-precision map production | |
Wu et al. | Applying a functional neurofuzzy network to real-time lane detection and front-vehicle distance measurement | |
CN104392212B (en) | The road information detection and front vehicles recognition methods of a kind of view-based access control model | |
CN102682292B (en) | Method based on monocular vision for detecting and roughly positioning edge of road | |
CN104318258B (en) | Time domain fuzzy and kalman filter-based lane detection method | |
Sotelo et al. | A color vision-based lane tracking system for autonomous driving on unmarked roads | |
AU2020102039A4 (en) | A high-precision multi-targets visual detection method in automatic driving scene | |
KR101285106B1 (en) | Obstacle detection method using image data fusion and apparatus | |
Kortli et al. | A novel illumination-invariant lane detection system | |
JP2008168811A (en) | Traffic lane recognition device, vehicle, traffic lane recognition method, and traffic lane recognition program | |
Shunsuke et al. | GNSS/INS/on-board camera integration for vehicle self-localization in urban canyon | |
US20230005278A1 (en) | Lane extraction method using projection transformation of three-dimensional point cloud map | |
CN104915642A (en) | Method and apparatus for measurement of distance to vehicle ahead | |
Kamble et al. | Lane departure warning system for advanced drivers assistance | |
Varun et al. | A road traffic signal recognition system based on template matching employing tree classifier | |
Truong et al. | Lane boundaries detection algorithm using vector lane concept | |
CN109191473B (en) | Vehicle adhesion segmentation method based on symmetry analysis | |
Getahun et al. | A robust lane marking extraction algorithm for self-driving vehicles | |
Manoharan et al. | A robust approach for lane detection in challenging illumination scenarios | |
US20230154013A1 (en) | Computer vision system for object tracking and time-to-collision | |
TWI619099B (en) | Intelligent multifunctional driving assisted driving recording method and system | |
CN113449629B (en) | Lane line false and true identification device, method, equipment and medium based on driving video | |
CN113569803A (en) | Multi-mode data fusion lane target detection method and system based on multi-scale convolution | |
Gohilot et al. | Detection of pedestrian, lane and traffic signal for vision based car navigation |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170606 |