CN109101743A - A kind of construction method of high-precision road net model - Google Patents
A kind of construction method of high-precision road net model Download PDFInfo
- Publication number
- CN109101743A CN109101743A CN201810986942.XA CN201810986942A CN109101743A CN 109101743 A CN109101743 A CN 109101743A CN 201810986942 A CN201810986942 A CN 201810986942A CN 109101743 A CN109101743 A CN 109101743A
- Authority
- CN
- China
- Prior art keywords
- lane
- section
- point
- linear
- road
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of construction methods of high-precision road net model, define high-precision road net model first, then extract road net data, including lane network layer is extracted and section network layer is extracted；Then section network layer is carried out, lane network layer incidence relation calculates；Finally construct high-precision road net model；The present invention indicates the corresponding road net model of HDmap with high-precision road net model HDRNM (high definition road network model), HDRNM from data content to data model on road net model all than section grade it is abundant.
Description
Technical field
The invention belongs to accurately diagram technology fields, are related to a kind of construction method of road net model, and in particular to a kind of
Serve the structure of unpiloted high-precision road net model (high definition road network model, HDRNM)
Construction method.
Background technique
HD Map is a kind of special service in unpiloted map, the application of high-precision map to it is unmanned increasingly
Important, high-precision road network is the most important component content in high-precision map.As developing for intelligent transportation is burning hot with ADAS
Development, HD map (High Definition map) cause very big concern (document 1-5) in academia and industry.
The cartographic information auxiliary intelligent vehicle that HD map is capable of providing fining realizes high accuracy positioning (document 6-8), can solve spy
The problem of condition of pledging love lower sensor fails, makes up the deficiency of environment sensing equipment, and the difficulty (text of intelligent vehicle perception is effectively reduced
Offer 9-11)；According to the map with the priori knowledge of dynamic information, based on global path planning provide optimal driving path and
It is reasonable to travel tactful (document 12-14), it effectively realizes and drives vehicle active safety, reduce the complexity (document of vehicle drive
15).Therefore, the generation of HD map becomes particularly significant, and HD map is currently in great demand stage (document 16).Road
Network data is the expression to real world road model, and high definition road network is the important of HD map
Component part.
But currently, in the research of high-precision road network automatically generated while that pays close attention to that road extraction and section extract grinds
Study carefully and is not much.
The generation of high-precision road network is studied, and the equipment of main focus utilization crowdsourcing or intelligent vehicle realizes the extraction of road network
(document 17-19), the generation (document 20,21) in high-precision section, the extraction (document 22) at high-precision crossing etc..On high-precision road
On the studying a question of the model of net, research is concentrated mainly on the format (document 23) in the expression of high-precision road network, high-precision road
The expression (document 24) of mouth, the research of the expansion such as road model (document 25,26), but rare people is concerned about lane and road simultaneously
Section branch topology relationship automatically generates.Some researchs have also been made in the forefathers that automatically generate for automatically generating topological relation of road network,
Main method has the crosspoint (document 27) for merging different layers, utilizes an association (document 28), using Hidden Markov
(documents 29) such as Model (HMM) map matching.But the topology that these researchs are not based on lane grade road network is extracted, needle
The topology of lane grade road network is extracted largely by manual method.
HD map data are usually Centimeter Level positioning accuracy of approximate (document 30), high
Precision road network automatically generates the extensive concern for increasingly obtaining scholar.The object that high-precision map is serviced not only includes existing
ADAS system, should also include Unmanned Systems, supplementary security system, bus or train route cooperative system etc..
On the model tormulation of high-precision road network, lane is extracted at present and model has scholar and done many relevant grind
Study carefully.Gi-Poong et al., come simulated roadway, improves the efficiency of road network storage with piecewise polynomials.
The a third-order of approximation an approximated clothoid spline of Chunzhao et al. a kind of
Polynomial expresses lane, expresses the transition curve (document at crossing with the cubic Catmull-Rom spline
It 17), can rapid modeling to lane and crossing.Anning et al. is with the Cubic Hermite spline in lane
Heart line is modeled (document 31), and the software of GIS database is suitble to carry out a series of lane and section modeling.Kichun et
Al. three-dimensional expression (document 25) is carried out to lane grade road network with B-spline curves, ensure that the shape of three-dimensional road network with
accuracy.The geometric expression of lane model is absorbed in this kind research.Tao et al. defines the lane of high-precision road network
Descriptive model, road network is formed into (document 23) by lane segmental arc, lane attribute, crossing and crossing attribute etc., is solved
The expression of lane model in high-precision road network, but it is a lack of the expression of road section information in high-precision road network, and without expression vehicle
Correspondence incidence relation between road and road network.
It is increasingly finer in the content of the expression of high-precision road network.U.S. Federal Highway Administration
And National Highway Traffic and Safety Administration is using lane as road network content, from general
The lane detailed information (document 32) of high-precision road network is enriched in thought.B é taille et al. is further to the geometric form in lane
Shape and topological connection relation are expressed, from the expression of the precision and content of road network all more complete (document 4).Tao et
Al. content (document 24) of the virtual lane as expression is increased in intersection, solves intersection and lacks asking for detailed information
Topic.But these researchs have certain office to unpiloted real-time without providing the attribute expression of dynamic multidimensional
It is sex-limited.
Bibliography:
Document 1.tomtommaps:https: //www.tomtommaps.com/mapdata/.
Document 2.deepmap:https: //www.deepmap.ai/.
Document 3.Nedevschi S, Popescu V, Danescu R, Marita T, Oniga F.Accurate Ego-
Vehicle Global Localization at Intersections Through Alignment of Visual Data
With Digital Map.IEEE Transactions on Intelligent Transportation
Systems.2013；14(2):673-87.
Document 4.B é taille D, Toledo-Moreo R.Creating enhanced maps for lane-level
vehicle navigation.IEEE Transactions on Intelligent Transportation
Systems.2010；11(4):786-98.
Document 5.Rohani M, Gingras D, Gruyer D.A Novel Approach for Improved
Vehicular Positioning Using Cooperative Map Matching and Dynamic Base Station
DGPS Concept.IEEE Transactions on Intelligent Transportation Systems.2016；17
(1):230-9.
Document 6.Suganuma N, Uozumi T, editors.Precise position estimation of
autonomous vehicle based on map-matching.Intelligent Vehicles Symposium；2011.
Document 7.Aeberhard M, Rauch S, Bahram M, Tanzmeister G.Experience, Results
and Lessons Learned from Automated Driving on Germany's Highways.IEEE
Intelligent Transportation Systems Magazine.2015；7(1):42-57.
Document 8.Toledo-Moreo R, Betaille D, Peyret F, Laneurit J.Fusing GNSS, Dead-
Reckoning,and Enhanced Maps for Road Vehicle Lane-Level Navigation.IEEE
Journal of Selected Topics in Signal Processing.2009；3(5):798-809.
Document 9.Driankov D, Saffiotti A.Fuzzy logic techniques for autonomous
vehicle navigation:Physica；2013.
Document 10.Cao G, Damerow F, Flade B, Helmling M, Eggert J, editors.Camera to
map alignment for accurate low-cost lane-level scene
interpretation.Intelligent Transportation Systems(ITSC),2016IEEE 19th
International Conference on；2016:IEEE.
Document 11.Gruyer D, Belaroussi R, Revilloud M.Accurate lateral positioning
from map data and road marking detection:Pergamon Press,Inc.；2016.1-8p.
Document 12.Li H, Nashashibi F, Toulminet G, editors.Localization for
intelligent vehicle by fusing mono-camera,low-cost GPS and map
data.International IEEE Conference on Intelligent Transportation Systems；
2011.
Document 13.Tang B, Khokhar S, Gupta R, editors.Turn prediction at
generalized intersections.Intelligent Vehicles Symposium(IV),2015IEEE；2015:
IEEE.
Document 14.Kim J, Jo K, Chu K, Sunwoo M.Road-model-based and graph-
structure-based hierarchical path-planning approach for autonomous
vehicles.Proceedings of the Institution of Mechanical Engineers,Part D:
Journal of Automobile Engineering.2014；228(8):909-28.
Document 15.Lozano-Perez T.Autonomous robot vehicles:Springer Science&
Business Media；2012.
16. sheep tomahawk Guo Liu Wu of document opens the progress and thinking China engineering science of left high-precision road navigation map
.2018；20(2):99-105.
Document 17.Guo C, Kidono K, Meguro J, Kojima Y, Ogawa M, Naito T.A low-cost
solution for automatic lane-level map generation using conventional in-car
sensors.IEEE Transactions on Intelligent Transportation Systems.2016；17(8):
2355-66.
Document 18.Mattern N, Schubert R, Wanielik G, editors.High-accurate vehicle
localization using digital maps and coherency images.Intelligent Vehicles
Symposium(IV),2010IEEE；2010:IEEE.
Document 19.Gwon G-P, Hur W-S, Kim S-W, Seo S-W.Generation of a precise and
efficient lane-level road map for intelligent vehicle systems.IEEE
Transactions on Vehicular Technology.2017；66(6):4517-33.
Document 20.Gikas V, Stratakos J.A novel geodetic engineering method for
accurate and automated road/railway centerline geometry extraction based on
the bearing diagram and fractal behavior.IEEE transactions on intelligent
transportation systems.2012；13(1):115-26.
Document 21.M á ttyus G, Wang S, Fidler S, Urtasun R, editors.Hd maps:Fine-
grained road segmentation by parsing ground and aerial images.Proceedings of
the IEEE Conference on Computer Vision and Pattern Recognition；2016.
Document 22.Yang X, Tang L, Niu L, Zhang X, Li Q.Generating lane-based
intersection maps from crowdsourcing big trace data.Transportation Research
Part C:Emerging Technologies.2018；89:168-87.
Document 23.Zhang T, Arrigoni S, Garozzo M, Yang D-g, Cheli F.A lane-level
road network model with global continuity.Transportation Research Part C:
Emerging Technologies.2016；71:32-50.
Document 24.Zhang T, Yang D, Li T, Li K, Lian X.An improved virtual
intersection model for vehicle navigation at intersections.Transportation
Research Part C:Emerging Technologies.2011；19(3):413-23.
Document 25.Jo K, Lee M, Kim C, Sunwoo M.Construction process of a three-
dimensional roadway geometry map for autonomous driving.Proceedings of the
Institution of Mechanical Engineers,Part D:Journal of Automobile
Engineering.2017；231(10):1414-34.
Document 26.Chen A, Ramanandan A, Farrell JA, editors.High-precision lane-
level road map building for vehicle navigation.Position Location and
Navigation Symposium(PLANS),2010IEEE/ION；2010:IEEE.
Document 27.Karagiorgou S, Pfoser D, Skoutas D.A layered approach for more
robust generation of road network maps from vehicle tracking data.ACM
Transactions on Spatial Algorithms and Systems(TSAS).2017；3(1):3.
Document 28.Xie X, Wong KB-Y, Aghajan H, Veelaert P, Philips W.Road network
inference through multiple track alignment.Transportation Research Part C:
Emerging Technologies.2016；72:93-108.
Document 29.Qiu J, Wang R.Automatic extraction of road networks from GPS
traces.Photogrammetric Engineering&Remote Sensing.2016；82(8):593-604.
Document 30.Du J, Barth MJ.Next-Generation Automated Vehicle Location
Systems:Positioning at the Lane Level.IEEE Transactions on Intelligent
Transportation Systems.2008；9(1):48-57.
Document 31.Chen A, Ramanandan A, Farrell JA, editors.High-precision lane-
level road map building for vehicle navigation.Position Location and
Navigation Symposium(PLANS)；2010.
Document 32.Enhanced Digital Mapping Project Final Report, Technical
report,United States Department of Transportation,Federal Highway
Administration and National Highway Traffic and Safety Administration,http://
Www-nrd.nhtsa.dot.gov/pdf/nrd12/CAMP/EDMap%20Final%20Rep ort/Main%20Report/
FinalRept_111904.pdf,189p.Last accessed Dec.8,2008.
Document 33.Qiu J, Wang R.Road Map Inference:A Segmentation and Grouping
Framework.ISPRS International Journal of Geo-Information.2016；5(8):130.
34. map of navigation electronic frame data exchange format of document
Summary of the invention
In order to solve these problems in background, the fine degree of high-precision road network is further enriched, is not only only focused on
Section network layer and lane network layer in road network, at the same also express in high-precision road network section network layer and lane network layer it
Between corresponding relationship, the invention proposes a kind of high-precision road net model HDRNM.HD road network is divided into section network layer by the model
And road network layer, geometry, topology and attribute information in addition to expressing section, lane in detail also define section network layer
The incidence relation between the network layer of lane.The place that attribute change occurs in section is defined as linear case point by the present invention, this
Linear event location in section is mapped in lane by relative position and is formed using section as linear measurement benchmark by invention
The linear case point in lane carries out partitioned representation to lane in the linear case point by lane.
The technical scheme adopted by the invention is that: a kind of construction method of high-precision road net model, which is characterized in that including
Following steps:
Step 1: defining high-precision road net model；
The model are as follows:
Wherein, W represents road network in formula 1, and C is intersection set, and R is the set in section；In formula 2, with 1,
2 ..., N } indicate section index set, r indicates section, r1, r2..., rNRespectively represent each section in set；In formula 3,
For certain a road section r, SrIt is the shape point in section, SNrIt is section starting point node, ENrIt is road segment end node, QrIt is section
Attribute, RLIt is section connected number, LSCorresponding lane set on section；In formula 4, lane rope is indicated with { 1,2 ..., i }
Draw set, l indicates lane, l1, l2..., liIndicate each lane of certain a road section ShiShimonoseki connection；In formula 5, for a certain lane
L, SlIt is the shape point in lane, SNlIt is the start node in lane, ENlIt is lane goal node, QlIt is the attribute in lane, LLIt is vehicle
Road connected number；In formula 6, Q Dynamic attribute values, expression is or no that t indicates time, the lane in q representation formula 3 and 4
Or the attribute value of the instruction class in section, the corresponding codomain of the value is enumeration type；
Step 2: extracting road net data, including section network layer is extracted and lane network layer is extracted；
Step 3: section network layer and lane network layer incidence relation calculate；
The incidence relation of section r lane l corresponding with the section in formula 3 are as follows:
C=f (M) (7)
In formula 7, C is the corresponding incidence relation in section and lane, and M is the set of linear case point, passes through linear event
Point indicates the position of attribute change on certain section；
The then functional relation C=f (M) in the corresponding lane in section is defined as:
Wherein, i indicates the sequence of the lane morphological data collection perpendicular to track direction；J indicates linearity range along section direction
Sequence；TotalLaneShapeFunction indicates the lane morphological data collection that section is parallel under current road segment；
Total points of totalMNum expression current road segment lower linear time point M point；InIt is unit matrix；L indicates lane, l1,jIndicate section
Lower i-th expression formula being parallel in j-th of the lane linearity range of road；LS indicates lane morphological data collection LSFunction's
It writes a Chinese character in simplified form；LSi,jIndicate that i-th is parallel to the lane morphological data collection LSFunction in section in j-th of linearity range under section
Expression formula；xjIt indicates in jth -1 to j-th of linearity range, the coordinate value range along section direction；XiI-th is indicated to be parallel to
Coordinate value range of the lane morphological data collection in section on road direction；
Step 4: building high-precision road net model.
Compared with the existing technology, the beneficial effects of the present invention are: proposing a kind of HDRNM model (A high
Definition road network model), which is divided into section network layer and road network layer for HD road network, in addition to
Section, the geometry in lane, topology and attribute information are expressed in detail, also defines incidence relation between section and lane.
The place that attribute change occurs in section is defined as linear case point by the present invention, and the present invention is using section as the linearity
Benchmark is measured, the linear event location in section is mapped to the linear case point for forming lane in lane by relative position,
Partitioned representation is carried out to lane in the linear case point by lane, and indicates that attribute becomes on certain section by linear case point
The position of change.The mapping relations adequately consider unpiloted demand, enrich the fine degree of high-precision road network, meet
The unmanned demand to the fining of high-precision road network, and this method can also be applied to the production knot of high-precision lane grade road network
Fruit is checked automatically.HDRNM from data content to data model on road net model all than section grade it is abundant.
Detailed description of the invention
Fig. 1 is the model construction functional block diagram of the embodiment of the present invention.
Specific embodiment
Understand for the ease of those of ordinary skill in the art and implement the present invention, with reference to the accompanying drawings and embodiments to this hair
It is bright to be described in further detail, it should be understood that implementation example described herein is merely to illustrate and explain the present invention, not
For limiting the present invention.
Referring to Fig.1, the present invention provides a kind of construction methods of high-precision road net model, comprising the following steps:
Step 1: defining high-precision road net model；
The model are as follows:
Wherein, W represents road network in formula 1, and C is intersection set, and R is the set in section；In formula 2, with 1,
2 ..., N } indicate section index set, r indicates section, r1, r2..., rNRespectively represent each section in set；In formula 3,
For certain a road section r, SrIt is the shape point in section, SNrIt is section starting point node, ENrIt is road segment end node, QrIt is section
Attribute (specific object is referring to national standard (document 34)), RLIt is section connected number, LSCorresponding lane set on section；Formula 4
In, lane index set is indicated with { 1,2 ..., i }, and l indicates lane, l1, l1..., liIndicate each of certain a road section ShiShimonoseki connection
Lane；In formula 5, for a certain lane l, SlIt is the shape point in lane, SNlIt is the start node in lane, ENlIt is lane terminal
Node, QlIt is attribute (the including but not limited to length in lane, the width in lane, the gradient in lane, the curvature in lane half in lane
Diameter etc., the value can carry out expansion of content with unpiloted application demand), LLIt is lane connected number；In formula 6, Q
Dynamic attribute values, expression are or no that t indicates the time, the attribute of the instruction class in lane or section in q representation formula 3 and 4
Value, the corresponding codomain of the value is enumeration type；
Step 2: extracting road net data, including section network layer is extracted and lane network layer is extracted；
Section network layer is extracted, and the direction in section is extracted using PCA Algorithm；Principal
Component analysis (PCA) is a kind of data analysing method (36) for commonly being used to describe feature distribution a little.PCA
Purpose be by n dimension initial data approximately indicated with k dimension, and after approximate representation to the loss of data as far as possible
It is small, and be to say that unit vector is found during data are transformed into new coordinate system by PCA from original coordinate system makes total
It can be maximized according to the variance of projection in this direction.PCA algorithm calculates eigenmatrix, and eigenmatrix is to represent spy mostly
Levy the sample distribution in dimension, the correlation put for metric point and around it.Feature vector is calculated by eigenmatrix.
Characteristic value X1, X2 are obtained by calculating feature vector.K=MAX { X1, X1 }/{ X1+X2 } value represents linear degree.Wherein K > 0.9
It is straight line (document 33) that cluster point, which could be represented, and therefore, the present embodiment is judged by linear degree in some point range
Point set whether belonged to linear relationship.
The present embodiment defines a search radius searchR first, carries out gauss projection to all coordinate points.Second
Step is normalized the point in the point search radius and PCA is projected, two-dimensional coordinate is projected to one-dimensional from arbitrary point
In space.Linear degree point K value is calculated by characteristic value.Third step, the present embodiment filter out the point set of all K > 0.9.It will
There is the point set greater than 0.9 of intersection to merge, form maximum linear point set, which respectively corresponds each section
In all lane center points set.4th step carries out PCA projection to maximum linear point set, obtains the main side in each section
To projecting direction, which is the direction in section.
Lane network layer is extracted, and is to establish the lane network based on multi-direction constraint；Specific implementation process is to mention section
It also needs to carry out coordinate points further classification after taking out to extract, finds lane point set different in section.In this mistake
Cheng Zhong, the present embodiment, which has been used, realizes that lane is extracted using the principal direction of PCA and angle threshold σ as constraint direction.Angle threshold σ table
Show the difference of next point of current coordinate point and the coordinate points and the projecting direction value of principal direction.Experimental result is come
It sees, the range of angle threshold σ is between [0,30 °].If section shape is typically chosen 15 ° of empirical value closer to straight line, if road
The shape on road is more bent, and angle threshold is selected in 30 °.
Specific extraction process is as follows:
By last process as a result, the present embodiment is obtained with the point sequence being ranked up according to direction under section
And the angle of principal direction.Firstly, the present embodiment carries out by first point tracking in principal direction every section, it is excellent according to principal direction
First principle is traversed, and finding current point and traversing point angle is the point that the point within the scope of σ degree is considered as on the same lane.Weight
Multiple circulation tracking, until having traversed all points.Length computation is carried out according to section direction to all lanes, is found on section
The linearity magnitude of linear case point.
Step 3: section network layer and lane network layer incidence relation calculate；
The incidence relation of section r lane l corresponding with the section in formula 3 are as follows:
C=f (M) (7)
In formula 7, C is the corresponding incidence relation in section and lane, and M is the set of linear case point, passes through linear event
Point indicates the position of attribute change on certain section；
The then functional relation C=f (M) in the corresponding lane in section is defined as:
Wherein, i indicates the sequence of the lane morphological data collection perpendicular to track direction；J indicates linearity range along section direction
Sequence；TotalLaneShapeFunction indicates the lane morphological data collection that section is parallel under current road segment；
Total points of totalMNum expression current road segment lower linear time point M point；InIt is unit matrix；L indicates lane, l1,jIndicate section
Lower i-th expression formula being parallel in j-th of the lane linearity range of road；LS indicates lane morphological data collection LSFunction's
It writes a Chinese character in simplified form；LSi,jIndicate that i-th is parallel to the lane morphological data collection LSFunction in section in j-th of linearity range under section
Expression formula；xjIt indicates in jth -1 to j-th of linearity range, the coordinate value range along section direction；XiI-th is indicated to be parallel to
Coordinate value range of the lane morphological data collection in section on road direction；
The place that attribute change occurs in section is defined as linear case point by the present embodiment, using section as linear measurement base
Linear event location in section is mapped to by relative position and is linearly sat by the lane of linear measurement benchmark of lane by standard
Under mark system, form the linear case point in lane in the linear coordinate system in lane, then by the linear case point in lane to lane into
Row partitioned representation；
Wherein partitioned representation realizes that process is that section is mapped to lane using linear reference system；Linear reference system is from section
To the mapping method in lane are as follows: the present embodiment is mapped the linear case point on lane with section direction selection linear reference system
Onto corresponding lane, then using lane as the object of dimension amount, change to attributes corresponding linear position in lane will be selected to make
For the foundation divided to the lane in same a road section.According to the road actual conditions of China, linear case point after mapping
It should be less than 10 meters with the linearity magnitude error of former M in track direction.
Step 4: building high-precision road net model.
The present invention is indicated with high-precision road net model HDRNM (high definition road network model)
The corresponding road net model of HDmap, HDRNM from data content to data model on road net model all than section grade it is abundant.
It should be understood that the part that this specification does not elaborate belongs to the prior art.
It should be understood that the above-mentioned description for preferred embodiment is more detailed, can not therefore be considered to this
The limitation of invention patent protection range, those skilled in the art under the inspiration of the present invention, are not departing from power of the present invention
Benefit requires to make replacement or deformation under protected ambit, fall within the scope of protection of the present invention, this hair
It is bright range is claimed to be determined by the appended claims.
Claims (4)
1. a kind of construction method of high-precision road net model, which comprises the following steps:
Step 1: defining high-precision road net model；
The model are as follows:
Wherein, W represents road network in formula 1, and C is intersection set, and R is the set in section；In formula 2, with 1,2 ...,
N } indicate section index set, r indicates section, r1, r2..., rNRespectively represent each section in set；In formula 3, for
Certain a road section r, SrIt is the shape point in section, SNrIt is section starting point node, ENrIt is road segment end node, QrIt is the attribute in section,
RLIt is section connected number, LSCorresponding lane set on section；In formula 4, lane indexed set is indicated with { 1,2 ..., i }
It closes, l indicates lane, l1, l2..., liIndicate each lane of certain a road section ShiShimonoseki connection；In formula 5, for a certain lane l, Sl
It is the shape point in lane, SNlIt is the start node in lane, ENlIt is lane goal node, QlIt is the attribute in lane, LLIt is that lane connects
Continuous number；In formula 6, Q Dynamic attribute values, expression is or no that t indicates the time, lane in q representation formula 3 and 4 or
The attribute value of the instruction class in section, the corresponding codomain of the value is enumeration type；
Step 2: extracting road net data, including section network layer is extracted and lane network layer is extracted；
Step 3: section network layer and lane network layer incidence relation calculate；
The incidence relation of section r lane l corresponding with the section in formula 3 are as follows:
C=f (M) (7)
In formula 7, C is the corresponding incidence relation in section and lane, and M is the set of linear case point, passes through linear case point table
Show the position of attribute change on certain section；
The then functional relation C=f (M) in the corresponding lane in section is defined as:
Wherein, i indicates the sequence of the lane morphological data collection perpendicular to track direction；J indicates linearity range along the sequence in section direction
Column；TotalLaneShapeFunction indicates the lane morphological data collection that section is parallel under current road segment；
Total points of totalMNum expression current road segment lower linear time point M point；InIt is unit matrix；L indicates lane, l1,jIndicate section
Lower i-th expression formula being parallel in j-th of the lane linearity range of road；LS indicates lane morphological data collection LSFunction's
It writes a Chinese character in simplified form；LSi,jIndicate that i-th is parallel to the lane morphological data collection LSFunction in section in j-th of linearity range under section
Expression formula；xjIt indicates in jth -1 to j-th of linearity range, the coordinate value range along section direction；XiI-th is indicated to be parallel to
Coordinate value range of the lane morphological data collection in section on road direction；
Step 4: building high-precision road net model.
2. high-precision road net model according to claim 1, which is characterized in that section network layer described in step 2 is extracted,
It is the direction in the track point set and section that section is extracted by PCA；Specific implementation process includes following sub-step:
Step 2.1.1: defining a search radius searchR, carries out gauss projection to all coordinate points；
Step 2.1.2: from arbitrary point, the point in the point search radius is normalized and PCA is projected, by two-dimensional coordinate
It projects in the one-dimensional space, linear degree point K value is calculated by characteristic value；
Step 2.1.3: the point set of all K > 0.9 is filtered out；There to be the point set greater than 0.9 of intersection to merge, is formed maximum
Linear point set, the maximum linear point set respectively correspond the set of all lane center points in each section；
Step 2.1.4: PCA projection is carried out to maximum linear point set, obtains the projecting direction of the principal direction in each section, the projection
Direction is the direction in section.
3. high-precision road net model according to claim 1, which is characterized in that lane network layer described in step 2 is extracted,
It is to establish the lane network based on multi-direction constraint；Specific implementation process includes following sub-step:
Step 2.2.1: realize that lane is extracted using the principal direction of PCA and angle threshold σ as constraint direction；
Step 2.2.2: every section is carried out by first point tracking in principal direction, according to the progress time of principal direction preferential principle
It goes through, finding current point and traversing point angle is the point that the point within the scope of σ degree is considered as on the same lane；
Step 2.2.3: repetitive cycling tracking, until having traversed all points；
Step 2.2.4: length computation is carried out according to road direction to all lanes, finds the linear of the online property case point in section
Metric.
4. high-precision road net model according to claim 1 to 3, it is characterised in that:, will be in section in step 2
The place that attribute change occurs is defined as linear case point, using section as linear measurement benchmark, by the linear event in section
Position is mapped to by relative position using lane as under the linear coordinate system in the lane of linear measurement benchmark, in the linear coordinate system in lane
The middle linear case point for forming lane, then partitioned representation is carried out to lane by the linear case point in lane；
Wherein partitioned representation realizes that process is that section is mapped to lane using linear reference system；Linear reference system is from section to vehicle
The mapping method in road are as follows: with section direction selection linear reference system, the linear case point on lane is mapped to corresponding lane
On, then using lane as the object of dimension amount, change to attributes corresponding linear position in lane will be selected to make as to same a road section
On the foundation that is divided of lane.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810986942.XA CN109101743B (en) | 2018-08-28 | 2018-08-28 | Method for constructing high-precision road network model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810986942.XA CN109101743B (en) | 2018-08-28 | 2018-08-28 | Method for constructing high-precision road network model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109101743A true CN109101743A (en) | 2018-12-28 |
| CN109101743B CN109101743B (en) | 2023-01-17 |
Family
ID=64851592
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810986942.XA Active CN109101743B (en) | 2018-08-28 | 2018-08-28 | Method for constructing high-precision road network model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109101743B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109614966A (en) * | 2019-03-07 | 2019-04-12 | 纽劢科技(上海)有限公司 | It is a kind of based on information fusion Lidar sensor efficient road surface and curb detection method |
| CN109976332A (en) * | 2018-12-29 | 2019-07-05 | 惠州市德赛西威汽车电子股份有限公司 | One kind being used for unpiloted accurately graph model and autonomous navigation system |
| CN110275883A (en) * | 2019-05-17 | 2019-09-24 | 浙江吉利控股集团有限公司 | A kind of high-precision map storage system and method |
| CN110415314A (en) * | 2019-04-29 | 2019-11-05 | 当家移动绿色互联网技术集团有限公司 | Construct method, apparatus, storage medium and the electronic equipment of intersection road network |
| CN111383450A (en) * | 2018-12-29 | 2020-07-07 | 阿里巴巴集团控股有限公司 | Traffic network description method and device |
| CN115100880A (en) * | 2022-06-17 | 2022-09-23 | 南京莱斯信息技术股份有限公司 | Bus signal priority control method for realizing bus balanced distribution |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140114556A1 (en) * | 2012-10-23 | 2014-04-24 | University Of Southern California | Traffic prediction using real-world transportation data |
| CN108010316A (en) * | 2017-11-15 | 2018-05-08 | 上海电科智能系统股份有限公司 | A kind of road traffic multisource data fusion processing method based on road net model |
| CN108009666A (en) * | 2016-10-28 | 2018-05-08 | 武汉大学 | The preferential optimal path computation method of level based on dynamic road network |
-
2018
- 2018-08-28 CN CN201810986942.XA patent/CN109101743B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140114556A1 (en) * | 2012-10-23 | 2014-04-24 | University Of Southern California | Traffic prediction using real-world transportation data |
| CN108009666A (en) * | 2016-10-28 | 2018-05-08 | 武汉大学 | The preferential optimal path computation method of level based on dynamic road network |
| CN108010316A (en) * | 2017-11-15 | 2018-05-08 | 上海电科智能系统股份有限公司 | A kind of road traffic multisource data fusion processing method based on road net model |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109976332A (en) * | 2018-12-29 | 2019-07-05 | 惠州市德赛西威汽车电子股份有限公司 | One kind being used for unpiloted accurately graph model and autonomous navigation system |
| CN111383450A (en) * | 2018-12-29 | 2020-07-07 | 阿里巴巴集团控股有限公司 | Traffic network description method and device |
| CN111383450B (en) * | 2018-12-29 | 2022-06-03 | 阿里巴巴集团控股有限公司 | Traffic network description method and device |
| CN109614966A (en) * | 2019-03-07 | 2019-04-12 | 纽劢科技(上海)有限公司 | It is a kind of based on information fusion Lidar sensor efficient road surface and curb detection method |
| CN109614966B (en) * | 2019-03-07 | 2019-07-26 | 纽劢科技(上海)有限公司 | It is a kind of based on information fusion Lidar sensor efficient road surface and curb detection method |
| CN110415314A (en) * | 2019-04-29 | 2019-11-05 | 当家移动绿色互联网技术集团有限公司 | Construct method, apparatus, storage medium and the electronic equipment of intersection road network |
| CN110415314B (en) * | 2019-04-29 | 2020-04-03 | 当家移动绿色互联网技术集团有限公司 | Method and device for constructing road network at road intersection, storage medium and electronic equipment |
| CN110275883A (en) * | 2019-05-17 | 2019-09-24 | 浙江吉利控股集团有限公司 | A kind of high-precision map storage system and method |
| CN115100880A (en) * | 2022-06-17 | 2022-09-23 | 南京莱斯信息技术股份有限公司 | Bus signal priority control method for realizing bus balanced distribution |
| CN115100880B (en) * | 2022-06-17 | 2024-05-14 | 南京莱斯信息技术股份有限公司 | Bus signal priority control method for realizing balanced distribution of buses |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109101743B (en) | 2023-01-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210172756A1 (en) | Lane line creation for high definition maps for autonomous vehicles | |
| CN109101743A (en) | A kind of construction method of high-precision road net model | |
| Wong et al. | Mapping for autonomous driving: Opportunities and challenges | |
| CN109256028A (en) | A method of it is automatically generated for unpiloted high-precision road network | |
| WO2020029601A1 (en) | Method and system for constructing transverse topological relationship of lanes in map, and memory | |
| US20230018923A1 (en) | Image-based keypoint generation | |
| Yang et al. | Generating lane-based intersection maps from crowdsourcing big trace data | |
| CN106980657A (en) | A kind of track level electronic map construction method based on information fusion | |
| CN112212874B (en) | Vehicle track prediction method and device, electronic equipment and computer readable medium | |
| CN110715671B (en) | Three-dimensional map generation method and device, vehicle navigation equipment and unmanned vehicle | |
| CN106441319A (en) | System and method for generating lane-level navigation map of unmanned vehicle | |
| CN106296814B (en) | Highway maintenance detection and virtual interactive interface method and system | |
| Mi et al. | Automated 3D road boundary extraction and vectorization using MLS point clouds | |
| WO2021003487A1 (en) | Training data generation for dynamic objects using high definition map data | |
| CN113358125B (en) | Navigation method and system based on environment target detection and environment target map | |
| CN115393745A (en) | Automatic bridge image progress identification method based on unmanned aerial vehicle and deep learning | |
| Flade et al. | Lane detection based camera to map alignment using open-source map data | |
| Xia et al. | DuARUS: Automatic Geo-object Change Detection with Street-view Imagery for Updating Road Database at Baidu Maps | |
| Chiang et al. | Bending the curve of HD maps production for autonomous vehicle applications in Taiwan | |
| CN116105717A (en) | Lane-level high-precision map construction method and system | |
| Wang et al. | Fast and reliable map matching from large-scale noisy positioning records | |
| Wang et al. | Lightweight High-precision Map for Specific Scenes | |
| Yang et al. | Review and Challenge: High Definition Map Technology for Intelligent Connected Vehicle | |
| Renault et al. | GPS/GIS localization for management of vision referenced navigation in urban environments | |
| Pang et al. | FLAME: Feature-likelihood based mapping and localization for autonomous vehicles |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Wang Xin Inventor before: Zheng Ling Inventor before: Li Bijun Inventor before: Wang Xin |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230607 Address after: 430072 Hubei Province, Wuhan city Wuchang District of Wuhan University Luojiashan Patentee after: WUHAN University Address before: 430223 7 floors, Block B, R&D Building No. 1, Huaengineering Science Park, Donghu New Technology Development Zone, Wuhan City, Hubei Province Patentee before: WUHAN ZHONGXIANG TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right |