CN108253976B - Three-stage online map matching algorithm fully relying on vehicle course - Google Patents
Three-stage online map matching algorithm fully relying on vehicle course Download PDFInfo
- Publication number
- CN108253976B CN108253976B CN201810009363.XA CN201810009363A CN108253976B CN 108253976 B CN108253976 B CN 108253976B CN 201810009363 A CN201810009363 A CN 201810009363A CN 108253976 B CN108253976 B CN 108253976B
- Authority
- CN
- China
- Prior art keywords
- gps track
- path
- point
- edge
- vehicle
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
- G01C21/30—Map- or contour-matching
Abstract
The invention discloses an online map matching algorithm fully based on vehicle course, relates to the field of map matching, and particularly relates to map matching based on vehicle course. Map matching is very important for many location information based applications, and online map matching is necessary to support online intelligent transportation applications. Therefore, the three-stage online map matching algorithm fully utilizing the vehicle heading is provided, and the algorithm can obtain a matching result with high accuracy and has low time complexity. Specifically, in the first stage, vehicle heading information is added to a given GPS track point to calculate the k most possible candidate edges of the GPS point in a road network; in the second stage, a path between two adjacent GPS track points is searched, the vehicle navigation direction is used for reducing the search range and serving as an effective search guide, and the search process is accelerated; and finally, selecting the real driving path of the vehicle for a series of GPS track points by using the vehicle course information.
Description
Technical Field
The invention relates to the technical field of map matching, in particular to a map matching algorithm by means of vehicle course.
Background
In recent years, mobile GPS devices, which are mounted in vehicles to collect movement trajectories of users, are widely spread. The GPS track of a vehicle is a data base that supports intelligent traffic services (e.g., route planning, traffic condition detection, etc.) based on location information. To save energy and transmission costs, GPS devices typically record vehicle position information at intervals, which results in sparsity and uncertainty in trajectory data, which can lead to reduced performance of the application. In order to solve these problems, GPS trajectory data is mapped onto a road network in consideration of the travel of vehicles in the road network, and this process is called map matching. In recent years, vehicles have been commonly used as traffic probes to detect traffic conditions, in which case online map matching is necessary.
The map matching work generally has two stages, the first stage is to find the true matching edge in the candidate edge of each GPS track point, and the second stage is to deduce and find the true driving path of the vehicle. Early map matching work was usually done only to find matching edges of GPS track points, but this work could only overcome the measurement error of GPS devices. For each candidate edge, the probability of each candidate edge is different from that of each given GPS track point, the previous work calculates the probability by using direction information, but the used direction information refers to the connecting line direction of the front and rear GPS track points, which proves that the direction information can help map matching, but because the next GPS track point cannot be known during the online matching, the direction information cannot be used in the online matching, so the invention tries to use the vehicle course information collected by a GPS device to complete the online map matching.
Closer map matching work focuses on how to reconstruct the driving path of a vehicle between two adjacent GPS track points. In detail, the matching edges corresponding to two GPS track points are not connected, even far away. This work is primarily directed to resolving the sparsity and uncertainty of the trajectory data. In order to find a path accurately, a map matching algorithm based on hidden markov reasonably uses many additional information such as the topology of a road network (edge-to-edge link relation) or road properties (vehicle speed limit) to increase the accuracy of the matching result. In order to deduce the true driving path, it is necessary to find the driving path of the vehicle between any pair of adjacent GPS track points, and Dijkstra and a-star algorithms are often used in this process of finding the path. However, since the path between each pair of adjacent GPS track points needs to be found, the process of finding the path is extremely time-consuming, and therefore, a heuristic path search algorithm based on the vehicle heading is provided, which can greatly reduce the search time.
Disclosure of Invention
In order to ensure high matching quality and low time complexity of online map matching, the invention provides a three-stage online map matching algorithm fully relying on vehicle course to perform online map matching on GPS track data stream input into the algorithm. For the track data stream, firstly, the track data stream is cut into original GPS track segments with equal length, one track segment is a processing unit, and for each track segment data, a proposed three-stage matching algorithm is applied for matching.
Specifically, the invention provides a specific scheme of a three-stage online map matching algorithm fully taking advantage of the vehicle course, which comprises the following steps:
a three-stage online map matching algorithm that fully relies on vehicle heading includes three stages. The first stage is a given GPS track point, and candidate edges in top-k road networks are determined; the second stage is to find a potential path between two GPS track points; and in the third stage, the real driving path of the vehicle is selected for a series of GPS track points.
Further, the first stage of the three-stage online map matching algorithm fully utilizing the vehicle heading comprises the following 4 steps: step 1: for a given GPS track point, firstly, the GPS track point is taken as the center of a circle, the length r meter is taken as the radius to draw a circle (r is 100 meters), and all edges in a road network covered by the circle are candidate edges of the GPS track point; step 2, calculating the possibility that the candidate edge screened out in the step 1 is the true matching edge of the given GPS track point by using the space probability; step 3, calculating the probability that the candidate edge screened out in the step 1 is the true matching edge of the given GPS track point by using the direction probability; and 4, step 4: and (4) calculating the comprehensive probability of each candidate edge according to the space probability and the direction probability of the candidate edges obtained in the steps (2) and (3), sequencing the probability values of the candidate edges, and selecting top-k candidate edges with the highest probability.
Further, the second stage of the three-stage online map matching algorithm fully utilizing the vehicle heading comprises the following 2 steps: step 1, determining a potential search area by using the heading and the running speed of a vehicle; and 2, establishing a tree structure of nodes in the road network according to the vehicle heading and the road network topological structure, and then finding a path between two GPS track points in a potential search area by utilizing a deep search algorithm (DFS).
Furthermore, the third stage of the three-stage online map matching algorithm fully depends on the vehicle course, and the probability that each candidate path is the matching path is calculated by utilizing the vehicle course information and the topological structure of the road network, and then the path with the maximum probability is selected as the matching path.
Drawings
FIG. 1 is a schematic diagram of the concept necessary in the present invention;
FIG. 2 is a system block diagram of a three-stage online map matching algorithm;
FIG. 3 is a schematic diagram demonstrating the role of vehicle heading in map matching.
FIG. 4 is a path search algorithm by vehicle heading.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
Before introducing the content of the algorithm, 6 necessary concepts of the invention are stated.
The 1 st concept is road network, which is a graph G (N, E) and includes a set of edge sets E and a set of node sets N. Each edge eiIs a directed edge having two nodes, a head node nhAnd tail node nt. Each node n is a coordinate combination of a pair of latitude and longitude, and represents the spatial position of the sampling point.
The 3 rd concept is that the vehicle heading h (0-h-360) refers to the heading direction of the vehicle at the sampling position, and the information can be directly obtained from the track data, as shown in FIG. 1, the vehicle is at the sampling positionGPS sampling point p1Has a course of h1。
The 4 th concept is that the GPS track points record the space and time information of the vehicle, contain longitude and latitude coordinates, timestamp information, instantaneous speed and vehicle course and are recorded as pi=(t1,lati,loni,vi,hi)。
The 5 th concept is that the original GPS track segment is a GPS track point arranged according to a time sequence, and is marked as tau ═ pi,pi+1,…,pi+l> (ii). The parameter l controls the length of the track segment. The original GPS track data stream T contains infinite GPS track segments, and is recorded as T ═ tau1,τ2,…>。
The 6 th concept is that the matched GPS track segment is given to an original GPS track segment, and the matched GPS track segment is the real driving path of the vehicle in the road network and is marked as taum=<ei,ei+1,…,en>. wherein any two adjacent edges are connected. A matched GPS trajectory data stream TmGPS track segment containing infinite matches, denoted Tm=<τm1,τm2,…>。
A three-stage on-line map matching algorithm fully depends on vehicle course to perform on-line map matching on GPS track data stream input into the algorithm. For the track data stream, firstly, the track data stream is cut into original GPS track segments with equal length, one track segment is a processing unit, and for each track segment data, a proposed three-stage matching algorithm is applied for matching. The algorithm comprises three stages, as shown in fig. 2, the first stage is to determine candidate edges in top-k road networks for a given GPS track point; the second stage is to find a potential path between two GPS track points; the third stage is to select the real driving path of the vehicle for a series of GPS track points (i.e. GPS track segments).
1. The first stage comprises the following four steps:
step 1: for a given GPS track point, the GPS track point is taken as the center of a circle, the length r is taken as the radius to draw a circle (r is 100 meters), and all edges in a road network covered by the circle are candidate edges of the GPS track point.
Step 2: and (3) calculating the probability that the candidate edge screened out in the step (1) is the edge really matched with the given GPS track point by using the space probability. Suppose a given GPS trace point piAnd candidate edge ejA difference in distance ofThe distance difference is typically a normal distribution function, which is expressed as follows:distance difference in formulaThe calculation method of (2) is as follows:the function dist being used to calculate the distance between two points, C1And C2Are respectively candidate edges ejTwo nodes of, C3Is GPS track point piTo candidate edge ejVertical point above. Note that if C3Out of the edge ejUp (i.e. at edge e)jOuter), then dist (p)i,C3)→+∞;σ1The standard deviation is calculated, in order to calculate the standard deviation, the matching edge of each GPS track point is calculated by using the existing classical map matching algorithm, then the distance difference set of all the GPS track points and the corresponding matching edges is solved, and finally sigma is easily fitted1The value of (c).
Finding candidate edges of GPS trace points based only on spatial probability may lead to erroneous results. As shown in FIG. 3, the GPS locus point p depends on the distance2Should follow edge e1Match, but this is clearly wrong. Can see the vehicle at the sampling point p2Is closer to e3This is thought to introduce directional probabilities to screen candidate edges for GPS trace points.
And step 3: and (3) calculating the probability that the candidate edge screened out in the step (1) is the true matching edge of the given GPS track point by using the direction probability. Suppose a given GPS trace point piAnd candidate edge ejHas an angle difference ofThe angle difference is also a normal distribution function, and the expression is as follows:each edge in the road network has two end points nhAnd ntSo that they both have two directions d (n)h,nt) And d (n)t,nh),d(nh,nt) The direction of the finger edge is from the end point nhTo nt,d(nt,nh) The direction of the finger edge is from the end point ntTo nh. The calculation method of the angle difference in the formula is as follows:σ2also standard deviation, method of calculation and sigma1And (5) the consistency is achieved.
And 4, step 4: according to spatial probability G1And the direction probability G2Calculating the comprehensive probabilityAnd sequencing the comprehensive probability G of all candidate edges, and selecting top-k candidate edges with the highest probability.
2. The second stage, comprising the following two steps:
step 1: the potential search area is determined by using the vehicle heading and speed information, and is a sector area which is uniquely determined by the following three conditions: firstly, the circle center of the sector is the first track point p of the two GPS track pointsi. Second, the radius of the sector is equal toAndthe velocity of two GPS track points, respectively, Δ t is the sampling time interval of two adjacent GPS track points, and c is a constant (set to 50 meters). And thirdly, the sector is composed of two semicircles, and the diameter of each semicircle is respectively vertical to the vehicle heading hi and hi +1 on the two GPS track points.
Within the potential search area, a potential path between two adjacent GPS track points is sought. Because each GPS track point has k candidate edges, any two adjacent GPS track points may have k in the potential search area2A potential path. Firstly, according to the vehicle course information, which node of two candidate edges corresponding to the GPS track point is the starting point n of the path is judgedsAnd an end point ne. As shown in FIG. 4, a pair of candidate edges e can be easily seen11And e2Node n8And n2Respectively an originating node and a terminating node. After determining the starting point of the path, the path is then searched as follows. Firstly, all the road network nodes n in the potential search areaiFind its child node nc. Sub-node ncThe following two conditions are satisfied: if n iscIs niIs equal to niIn contrast, ncShould be away from neCloser, and further away from ns, as shown in equations (3) and (4); and from niTo ncShould be related to the vehicle heading hiAnd hi+1And (5) in agreement, as shown in equation (5). dist (n)c,ne)<dist(ni,ne)---(3),dist(nc,ns)>dist(ni,ns)---(4),min(|hi-d(ni,nc)|,|hi+1-d(ni,ns)|)<90 ° - - (5) then we refer to the determined parent-child node relationship by nsFor the root node, a tree structure is built with other nodes within the potential search area. If the constructed tree does not contain neThen we consider that there is no path between the two GPS trace points. Otherwise, a potential path between two track points is found by utilizing a depth search algorithm (DFS), and the search is stopped once the path is foundAnd (4) stopping.
3. The third stage comprises the following steps:
for any two adjacent GPS track points, k is at most2A potential path, then for a track segment containing l track points GPS, there will be at most k2(l-1)A potential route. Because the probability that each edge in the route is a true matching edge of the corresponding GPS track point is different, the probability that each route is a true driving route of the vehicle is also different. Calculating the probability of each path being the real driving path of the vehicle by using a formula (6), and finally outputting the path with the maximum probability as a matching pathHerein, theRefer to candidate edge ejIs a point of trace piThe integrated probability values of the edges are truly matched. Note that for a track segment containing l GPS track points < p1,p2,…,plPossible paths of > are denoted asHerein, theRefers to the locus point piThe jth candidate edge.Refer to from the candidate edgeTo the candidate edgeThe path of (2).
Claims (2)
1. A three-stage online map matching algorithm fully relying on vehicle heading is characterized by comprising the following three stages:
(1) in the first stage, for a given GPS track point, determining the most possible candidate edges in top-k road networks by using spatial position information and vehicle heading information;
(2) in the second stage, potential driving paths between two adjacent GPS track points are searched by utilizing vehicle route information;
(3) in the third stage, selecting a real driving path of the vehicle for a series of continuous GPS track point sets;
the first stage comprises the following four steps:
step 1: for a given GPS track point, firstly, drawing a circle by taking the GPS track point as a circle center and taking the length r as a radius, wherein all edges in a road network covered by the circle are candidate edges of the GPS track point;
step 2: calculating the probability that the candidate edge screened out in the step 1 is the true matching edge of the given GPS track point by using the space probability; suppose a given GPS trace point piAnd candidate edge ejA difference in distance ofThe distance difference is typically a normal distribution function, which is expressed as follows:
distance difference in formulaThe calculation method of (2) is as follows:the function dist being used to calculate the distance between two points, C1And C2Are respectively candidate edges ejTwo nodes of, C3Is GPS track point piTo candidate edge ejA vertical point above; note that if C3Out of the edge ejAbove, thatHow dist (p)i,C3)→+∞;σ1The standard deviation is calculated, in order to calculate the standard deviation, the matching edge of each GPS track point is calculated by using the existing classical map matching algorithm, then the distance difference set of all the GPS track points and the corresponding matching edges is solved, and finally sigma is fitted1A value of (d);
and step 3: calculating the probability that the candidate edge screened out in the step 1 is the true matching edge of the given GPS track point by using the direction probability; suppose a given GPS trace point piAnd candidate edge ejHas an angle difference ofThe angle difference is also a normal distribution function, and the expression is as follows:
each edge in the road network has two end points nhAnd ntSo that they both have two directions d (n)h,nt) And d (n)t,nh),d(nh,nt) The direction of the finger edge is from the end point nhTo nt,d(nt,nh) The direction of the finger edge is from the end point ntTo nh(ii) a The calculation method of the angle difference in the formula is as follows:σ2also standard deviation, method of calculation and sigma1The consistency is achieved;
and 4, step 4: obtaining the space probability G of all candidate edges of each GPS track point according to the steps 2 and 31And the direction probability G2First, the comprehensive probability of each edge is calculatedThen, sequencing the comprehensive probability G, and finally selecting top-k candidate edges with the maximum probability;
the second stage comprises the following two steps:
step 1: the potential search area is determined by using the vehicle heading and speed information, and is a sector area which is uniquely determined by the following three conditions: firstly, the circle center of the sector is the first track point p of the two GPS track pointsi(ii) a Second, the radius of the sector is equal to max (v)pi,vpi+1)×Δt+c,vpiAnd vpi+1Respectively the speeds of two GPS track points, delta t is the sampling time interval of two adjacent GPS track points, and c is a constant; thirdly, the sector is composed of two semicircles, and the diameter of each semicircle is respectively vertical to the vehicle heading hi and hi +1 on the two GPS track points;
step 2: searching a potential path between two adjacent GPS track points in a potential search area; because each GPS track point has k candidate edges, any two adjacent GPS track points may have k in the potential search area2A potential path; firstly, according to the vehicle course information, judging which node in two candidate edges corresponding to the GPS track point is the starting point n of the pathsAnd an end point ne(ii) a After determining the starting point of the path, searching the path by using the following steps; firstly, all the road network nodes n in the potential search areaiFind its child node nc(ii) a Sub-node ncThe following two conditions are satisfied: if n iscIs niIs equal to niIn contrast, ncShould be away from neCloser, and further away from ns, as shown in equations (3) and (4); and from niTo ncShould be related to the vehicle heading hiAnd hi+1Consistent, as shown in equation (5);
dist(nc,ne)<dist(ni,ne)---(3)
dist(nc,ns)>dist(ni,ns)---(4)
min(|hi-d(ni,nc)|,|hi+1-d(ni,ns)|)<90°---(5)
then we use n according to the determined parent-child node relationsEstablishing a tree structure for the root node by using other nodes in the potential search area; if the constructed tree does not contain neThen we consider that there is no path between the two GPS trace points; otherwise, a potential path between two track points is found by using a depth search algorithm, and the search is stopped once the path is found.
2. The three-stage online map matching algorithm substantially aided by vehicle heading as claimed in claim 1, wherein the third stage comprises the steps of:
for any two adjacent GPS track points, k is at most2A potential path, then for a track segment containing l track points GPS, there will be at most k2(l-1)A potential route; because the probability that each edge in the path is the real matching edge of the corresponding GPS track point is different, the possibility that each route is the real driving path of the vehicle is also different; calculating the probability of each path being the real driving path of the vehicle by using a formula (6), and finally outputting the path with the maximum probability as a matching path;
herein, theRefer to candidate edge ejIs a point of trace piReally matching the comprehensive probability value of the edges; note that for a track segment containing l GPS track points < p1,p2,…,plPossible paths of > are denoted asHerein, theRefers to the locus point piThe jth candidate edge;refer to from the candidate edgeTo the candidate edgeThe path of (2).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810009363.XA CN108253976B (en) | 2018-01-04 | 2018-01-04 | Three-stage online map matching algorithm fully relying on vehicle course |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810009363.XA CN108253976B (en) | 2018-01-04 | 2018-01-04 | Three-stage online map matching algorithm fully relying on vehicle course |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108253976A CN108253976A (en) | 2018-07-06 |
CN108253976B true CN108253976B (en) | 2021-06-15 |
Family
ID=62724887
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810009363.XA Active CN108253976B (en) | 2018-01-04 | 2018-01-04 | Three-stage online map matching algorithm fully relying on vehicle course |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108253976B (en) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110967990B (en) * | 2018-09-30 | 2021-09-03 | 北京地平线信息技术有限公司 | Track determination method and device and electronic equipment |
CN109640263A (en) * | 2018-11-30 | 2019-04-16 | 重庆大学 | A kind of online trace compression system and method based on mobile edge calculations |
DE102019101569A1 (en) * | 2019-01-23 | 2020-07-23 | Bayerische Motoren Werke Aktiengesellschaft | Method for driverless transfer of a vehicle over a route within a closed area |
CN110232470B (en) * | 2019-04-11 | 2023-01-31 | 创新先进技术有限公司 | Method and device for determining vehicle driving path |
CN111857113B (en) * | 2019-04-12 | 2023-11-03 | 北京地平线机器人技术研发有限公司 | Positioning method and positioning device for movable equipment |
CN110686686B (en) * | 2019-06-04 | 2020-10-02 | 滴图(北京)科技有限公司 | System and method for map matching |
WO2022056770A1 (en) * | 2020-09-17 | 2022-03-24 | 华为技术有限公司 | Path planning method and path planning apparatus |
CN113970333A (en) * | 2021-09-26 | 2022-01-25 | 深圳市跨越新科技有限公司 | Adaptive candidate road searching method, system, terminal device and storage medium |
CN114674335B (en) * | 2022-03-24 | 2023-06-20 | 西南交通大学 | Matching method of optimal target link |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104331422A (en) * | 2014-10-14 | 2015-02-04 | 广州市香港科大霍英东研究院 | Road section type presumption method |
CN105095481A (en) * | 2015-08-13 | 2015-11-25 | 浙江工业大学 | Large-scale taxi OD data visual analysis method |
CN105203116A (en) * | 2015-08-25 | 2015-12-30 | 浙江工业大学 | Map matching method based on conditional random fields and low-sampling-frequency floating car data |
CN105444769A (en) * | 2015-11-26 | 2016-03-30 | 北京百度网讯科技有限公司 | Map matching method and device |
CN105489006A (en) * | 2015-12-15 | 2016-04-13 | 浙江工业大学 | Multi-scale road flow visual analysis method based on taxi GPS data |
CN105526939A (en) * | 2014-09-29 | 2016-04-27 | 高德软件有限公司 | Road coupling method and apparatus thereof |
CN106023587A (en) * | 2016-05-25 | 2016-10-12 | 电子科技大学 | Track data road network precise matching method based on multi-information fusion |
CN106157624A (en) * | 2016-08-04 | 2016-11-23 | 浙江工业大学 | Many granularities road shunting visual analysis methods based on traffic location data |
CN106595680A (en) * | 2016-12-15 | 2017-04-26 | 福州大学 | Vehicle GPS data map matching method based on hidden markov model |
CN106767873A (en) * | 2016-12-30 | 2017-05-31 | 浙江大学 | A kind of map-matching method based on space-time |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10288433B2 (en) * | 2010-02-25 | 2019-05-14 | Microsoft Technology Licensing, Llc | Map-matching for low-sampling-rate GPS trajectories |
-
2018
- 2018-01-04 CN CN201810009363.XA patent/CN108253976B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105526939A (en) * | 2014-09-29 | 2016-04-27 | 高德软件有限公司 | Road coupling method and apparatus thereof |
CN104331422A (en) * | 2014-10-14 | 2015-02-04 | 广州市香港科大霍英东研究院 | Road section type presumption method |
CN105095481A (en) * | 2015-08-13 | 2015-11-25 | 浙江工业大学 | Large-scale taxi OD data visual analysis method |
CN105203116A (en) * | 2015-08-25 | 2015-12-30 | 浙江工业大学 | Map matching method based on conditional random fields and low-sampling-frequency floating car data |
CN105444769A (en) * | 2015-11-26 | 2016-03-30 | 北京百度网讯科技有限公司 | Map matching method and device |
CN105489006A (en) * | 2015-12-15 | 2016-04-13 | 浙江工业大学 | Multi-scale road flow visual analysis method based on taxi GPS data |
CN106023587A (en) * | 2016-05-25 | 2016-10-12 | 电子科技大学 | Track data road network precise matching method based on multi-information fusion |
CN106157624A (en) * | 2016-08-04 | 2016-11-23 | 浙江工业大学 | Many granularities road shunting visual analysis methods based on traffic location data |
CN106595680A (en) * | 2016-12-15 | 2017-04-26 | 福州大学 | Vehicle GPS data map matching method based on hidden markov model |
CN106767873A (en) * | 2016-12-30 | 2017-05-31 | 浙江大学 | A kind of map-matching method based on space-time |
Non-Patent Citations (2)
Title |
---|
An interactive-voting based map matching algorithm;Yuan J,et al.;《2010 Eleventh International Conference on Mobile Data Management》;20101231;43-52 * |
Map-matching for low-sampling-rate GPS trajectories;Lou Y,et al;《17th ACM SIGSPATIAL international conference》;20091231;352-361 * |
Also Published As
Publication number | Publication date |
---|---|
CN108253976A (en) | 2018-07-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108253976B (en) | Three-stage online map matching algorithm fully relying on vehicle course | |
CN106912018B (en) | Map matching method and system based on signaling track | |
Hashemi et al. | A critical review of real-time map-matching algorithms: Current issues and future directions | |
CN109405839B (en) | Traffic network off-line map matching algorithm based on multiple paths | |
EP3136128B1 (en) | Trajectory matching using peripheral signal | |
Mohamed et al. | Accurate real-time map matching for challenging environments | |
Jagadeesh et al. | A map matching method for GPS based real-time vehicle location | |
US9778061B2 (en) | Road density calculation | |
EP3663718B1 (en) | Method and apparatus for estimating a localized position on a map | |
EP2556338B1 (en) | Probe data processing | |
Mohamed et al. | Accurate and efficient map matching for challenging environments | |
CN104900059A (en) | Method for enhancing cell phone base station positioning precision by using Hidden Markov map-matching algorithm | |
US11231282B2 (en) | Method and apparatus for providing node-based map matching | |
TW201131143A (en) | Methods and systems for creating digital transportation networks | |
TW201111744A (en) | Method of verifying or deriving attribute information of a digital transportation network database using interpolation and probe traces | |
CN106781478A (en) | A kind of trace tracking method based on LTE signaling datas | |
US11578982B2 (en) | Method and apparatus for map matching trace points to a digital map | |
EP3093620B1 (en) | System and method for detecting roundabouts from probe data using vector fields | |
EP3678051A1 (en) | Lane count estimation | |
CN106855878B (en) | Historical driving track display method and device based on electronic map | |
US20220338014A1 (en) | Trustworthiness evaluation for gnss-based location estimates | |
Dogramadzi et al. | Accelerated map matching for GPS trajectories | |
Chen et al. | Local path searching based map matching algorithm for floating car data | |
Wu et al. | CLSTERS: A general system for reducing errors of trajectories under challenging localization situations | |
Karimi et al. | Uncertainty in personal navigation services |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |