Probalistic reverse geocoding

Information

  • Patent Grant
  • 8838379
  • Patent Number
    8,838,379
  • Date Filed
    Friday, March 8, 2013
    11 years ago
  • Date Issued
    Tuesday, September 16, 2014
    10 years ago
Abstract
The accuracy of a reverse geocode for a GPS fix in a mobile phone navigation device is improved significantly by considering together the user's last known location on an original route, the elapsed time between the last known location on the original route and the latest GPS fix, and the likelihood that the user could have travelled from the last known location on the original route to one of numerous candidate locations (generated from the reverse geocode) in the elapsed time. The navigation server relies only on a GPS fix, timestamps, and a back-end routing database to provide improved or superior accuracy of reverse geocoding a GPS fix (selection of a point on a road) when a relevant user has deviated from the intended route. The reverse geocoding allows for accurate and functional re-routing capabilities.
Description
BACKGROUND OF THE INVENTION

1. Field of the Invention


This invention relates generally to wireless communication. More particularly it relates to route guidance (navigation) on a mobile wireless device including global positioning satellite (GPS) capability.


2. Background of Related Art


Navigation devices are typically GPS-based electronic systems used in a vehicle to provide a real-time map of the vehicle's current location, together with step-by-step directions to a requested destination.


All navigation devices must handle the condition in which the user deviates (leaves) their intended route. When this condition occurs, the application typically determines the user's current location by “reverse geocoding” the last GPS fix from the device. This new current location is then used to generate a new route to the original destination.


Reverse geocoding is a process of returning an estimated position on a street as it relates to a given latitude/longitude coordinate. Mobile applications available today use a simple reverse geocoding algorithm to locate the closest point on any road segment near the GPS coordinate received from the device.


However, due to the inherent inaccuracy of a GPS fix, this simplistic approach may locate the user on an incorrect road that is near the user's true location. Since this occurs while the user is driving and after they have deviated from the intended route, the cost of an incorrect reverse geocode is quite high since the new route will also be incorrect.


SUMMARY OF THE INVENTION

In accordance with the principles of the present invention, a method of providing an accurate location to a portable navigation device in a congested area when in an off-route condition comprises obtaining a latest location fix of the user's wireless navigation device, the latest location fix being off-route. A last known on-route location of a user's wireless navigation device is obtained. An elapsed time between the last known on-route location and the latest location fix is determined. A reverse geocode is performed on the latest location fix resulting in a plurality of potential location candidates. An estimated time of travel from the last known on-route location to each of the potential location candidates is determined, and a current location of the user's wireless navigation device is mapped as the potential location candidate having a travel time closest to the elapsed time.


A method of providing navigation server-based support for a wireless navigation device in accordance with another aspect of the invention comprises considering together a user wireless device's last known on-route location, a latest GPS fix of the user's wireless device, and an elapsed time between the last known location on the original route and the latest GPS fix. A likelihood that the user's wireless device could have travelled from the last known on-route location on the original route to one of a plurality of candidate locations generated from a reverse geocode, in the elapsed time, is determined. An estimated time of travel from the last known on-route location to each of the plurality of candidate locations is determined, and mapping information is provided based on a current location of the user's wireless navigation device being determined as being the potential location candidate that has an estimated time of travel closest to the elapsed time.





BRIEF DESCRIPTION OF THE DRAWINGS

Features and advantages of the present invention will become apparent to those skilled in the art from the following description with reference to the drawings:



FIG. 1 shows an exemplary mobile phone including a user interface providing route guidance navigation, as a user goes off route, in accordance with the principles of the present invention.



FIG. 2 shows an exemplary method providing probabilistic reverse geocoding for a user of a mobile phone with an operating route guidance navigation interface, in accordance with the principles of the present invention.



FIG. 3 shows a graphical depiction of a list of candidate reverse geocodes generated by a current GPS fix, each being a point on a road segment near the actual latitude/longitude, in accordance with the principles of the present invention.



FIGS. 4A to 4G show a route generated between the last known location and each respective candidate reverse geocode result shown in FIG. 3.





DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

Car navigation systems typically employ superior GPS technology, gyros and actual vehicle speed to determine a vehicle's current position more accurately. But more recently navigation applications have become available on wireless phones. However, mobile phones don't have access to a directly measured speed of a vehicle as do installed car navigation systems. Moreover, the antenna of a mobile phone may not be in an ideal position as are the antennas for an installed car navigation system. While very good, inaccuracies in the measurement of a current location may lead to an erroneous mapping onto a road in a mobile phone navigation application, particularly in areas where many roads intersect. In a worse case, such inaccuracies can result in mapping of a vehicle on a wrong road.


The present inventors have discovered that the accuracy of a reverse geocode can be improved significantly by considering together the user's last known location on the original route, the elapsed time between the last known location on the original route and the latest GPS fix, and the likelihood that the user could have travelled from the last known location on the original route to one of numerous candidate locations (generated from the reverse geocode) in the elapsed time.


The present invention provides an alternative approach to reverse geocoding that relies only on a GPS fix, timestamps, and a back-end routing database. The present invention provides improved or superior accuracy of reverse geocoding a GPS fix (selection of a point on a road) when a relevant user has deviated from the intended route. The reverse geocoding allows for accurate and functional re-routing capabilities.



FIG. 1 shows an exemplary mobile phone including a user interface providing visual route guidance navigation, as a user goes off route, in accordance with the principles of the present invention.


In particular, FIG. 1 depicts a mobile phone 120 providing visual route guidance navigation to a user via a visual user interface including a mobile phone display. In the disclosed embodiments, the user is presumed to be in an automobile, though the present invention relates equally to travel by way of any vehicle, including bicycles, bus, as well as travel by foot.


Though shown and described with respect to a visual user interface to guide a user of a mobile phone 120, the present invention relates equally to audio route guidance and/or visual route guidance.


On the display of the mobile phone 120, a visual user interface displays a map with relevant paths (e.g., roads in the case of vehicle travel). In the described situation, the user of the mobile phone 120 is carrying it with them while traveling in a car or bus as they approach an area congested with roadways. The invention has particular applicability to points at which one road passes over another, where a navigation application can easily get confused. In the given situation the user is traveling in a vehicle in an area near to quite a few roadways.


The triangular icon 402 depicts the vehicle in which the user is traveling, at its last known position of the mobile phone 120 as it traveled along a route 406. In the shown example, the user was presented with, and was expected to travel along, a previously calculated route 408. However, in the given scenario, the latest GPS fix puts the user at a point that would otherwise be considered to be along a different roadway 410. The triangular icon 404 depicts the latest GPS position fix of a user as reported by the GPS device of the mobile phone 120 carried by a passenger of a vehicle that was previously traveling along a preplanned route 408.



FIG. 2 shows an exemplary method providing probabilistic reverse geocoding for a user of a mobile phone with an operating route guidance navigation interface, in accordance with the principles of the present invention.


In particular, as shown in FIG. 2, a mobile device 120 performs a navigation method supported by route guidance navigation information provided by a remote navigation server 140 within a phone network. The mobile device 120 communicates with the remote navigation server 140 via an otherwise conventional wireless phone protocol.


The method for probabilistic geocoding in accordance with the principles of the present invention begin with the following preferable input:


(a) Last known location on a road segment (map matched by the application) (While the user travels along a route, the application ‘map matches’ each GPS fix to the nearest point on the route. When the distance to the nearest point exceeds some threshold, the user is considered off route.


(b) Timestamp of last known location.


(c) Current GPS fix (lat/lon) from the device.


(d) Timestamp of current GPS fix.


An exemplary method for probabilistic geocoding in accordance with the principles of the present invention is as follows:


In the mobile device 120, steps 202-208 are performed in the exemplary embodiment, whereas steps 302-316 are performed at a suitable navigation server 140 in the exemplary embodiment.


In step 202 of the mobile device 120, the mobile phone 120 determines or otherwise obtains information informing it that it is ‘off-route’.


In step 204, a new route request is created.


In step 206, the new route request is sent to the navigation server 140.


In step 208, after the navigation server 140 provides a new route back to the mobile phone 120, guidance is started on the new route received back from the navigation server 140.


The navigation server 140 receives a new route request from the mobile phone 120 in step 302.


In step 304, the navigation server 140 determines candidate locations. The navigation server 140 also examines the road network proximate to the latest GPS fix position. The navigation server 140 also selects points on nearest road links within a given error radius of the GPS fix.


In step 306, the navigation server 140 determines if there are more than one candidate location, or road presence point, within a given error radius of the latest GPS fix. A list of candidate reverse geocode location results is generated from the current GPS fix 404. Each candidate location is a point on a road segment near the actual latitude/longitude.



FIG. 3 shows a graphical depiction of a list of candidate reverse geocodes generated by a current GPS fix, each being a point on a road segment near the actual latitude/longitude, in accordance with the principles of the present invention.


In particular, as shown in FIG. 3, a latest GPS fix is depicted by the triangular icon 404. However, in the given example, at least 6 candidate locations 502-512 are identified by the navigation server 140.


Referring back to FIG. 2, in step 308, for each candidate location, a route is generated for each candidate location, from the last known location 406, to the respective candidate location.



FIGS. 4A to 4G show a route generated between the last known location and each respective candidate reverse geocode result shown in FIG. 3.


In particular, FIG. 4A depicts a route to get to location candidate 510 from the last known position 402. FIG. 4B depicts a route to get to location candidate 512 from the last known position 402. FIG. 4C depicts a route to get to location candidate 502 from the last known position 402. FIG. 4D depicts a route to get to location candidate 508 from the last known position 402. FIG. 4E depicts a route to get to location candidate 506 from the last known position 402. FIG. 4F depicts a route to get to location candidate 504 from the last known position 402. FIG. 4G depicts a route to get to location candidate 514 from the last know position 402.


For each candidate reverse geocode result, a route is generated between the last known location 402 and the respective candidate location 502-514. This route generation step calculates estimated elapsed time based on the speed limit of each road segment traveled. The present invention also contemplates that other influences such as traffic impacts may be taken into account, e.g., the presence and duty cycle of traffic lights along the way, stop signs, etc.


Referring back to FIG. 2, in step 310, for each route to respective candidate locations calculated in step 308, the travel time is determined, and compared to an elapsed time between the time stamp of the last known location 406 and the time stamp of the latest GPS fix 404.


The location candidates 502-514 are ranked based on the elapsed time between the two GPS fixes 402 and 404, and the estimated time of travel between the known and current location.


The travel time for each route is generally the driving distance on each road segment×the speed limit on each road segment. For simplicity purposes of the given embodiment shown in FIGS. 4A to 4G, it is presumed that the speed limit is the same on all shown roads.


In step 312, a best fit candidate location is selected as a candidate location having a time closest to an expected time since the last known location was fixed. Thus, the location candidate with the smallest time difference is selected and returned as the reverse geocode result. Given the presumption above about a same speed limit on all shown roads in this example, the candidate location shown in FIG. 4G is a best fit.


Referring back to FIG. 2, in step 314, a new route is calculated from the latest or current location fix 404, to the original destination.


Finally, in step 316, the new route is communicated to the client mobile phone 120.


The present invention may be generalized further to find a best-fit route between a series of 2 or more points, in accordance with the principles of the present invention. In this case, there is not a last known location but pairs of candidates from which the most likely route may be selected. This would be more computationally intensive but can be used to improve the accuracy of reverse geocoding (of a moving device) when there is no last known location.


The invention may also be generalized by maintaining an ongoing list of likely reverse geocode results. For each subsequent GPS fix, the previous list may be used to generate the new list. This may be employed even when not on a route.


The present invention has particular application to navigation application providers.


While the invention has been described with reference to the exemplary embodiments thereof, those skilled in the art will be able to make various modifications to the described embodiments of the invention without departing from the true spirit and scope of the invention.

Claims
  • 1. A method of providing an accurate location to a user's wireless navigation device, comprising: receiving a route request from a user's wireless navigation device;obtaining a last known on-route location of said navigation device;obtaining a latest location fix of said navigation device;performing a reverse geocode on said latest location fix to determine a plurality of candidate locations within a given error radius of said latest location fix associated with said navigation device;determining an estimated time of travel between each of said plurality of candidate locations and said last known on-route location; andreturning a most likely candidate location, from said plurality of candidate locations, based on a closest match between said respective estimated times, and an elapsed time between said last known on-route location and said latest location fix.
  • 2. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said latest location fix is a Global Positioning System (GPS) fix.
  • 3. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said navigation device calculates a new route from a current location of said navigation device and said returned most likely candidate location.
  • 4. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said navigation device is a portable navigation device.
  • 5. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said navigation device is a wireless phone.
  • 6. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: generation of said plurality of candidate locations is based on a speed limit of each road segment traveled.
  • 7. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said plurality of candidate locations is based on traffic.
  • 8. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: said plurality of candidate locations is based on a duty cycle of a traffic light.
  • 9. The method of providing an accurate location to a user's wireless navigation device according to claim 1, wherein: a same speed limit is used for determining said estimated time of travel between each of said plurality of candidate locations and said last known on-route location.
Parent Case Info

This application is a continuation of U.S. application Ser. No. 12/588,143, entitled “Probalistic Reverse Geocoding,” filed on Oct. 6, 2009, now U.S. Pat. No. 8,396,658; which claims priority from U.S. Provisional Application No. 61/136,805, entitled “Probabilistic Reverse Geocoding” to Barcklay, filed Oct. 6, 2008, the entirety of both are explicitly incorporated herein by reference.

US Referenced Citations (164)
Number Name Date Kind
5263136 DeAguiar Nov 1993 A
5359529 Snider Oct 1994 A
5781200 Lu Jul 1998 A
5832408 Tamai Nov 1998 A
5973700 Taylor Oct 1999 A
6026398 Brown Feb 2000 A
6064941 Nimura May 2000 A
6104416 McGuiness Aug 2000 A
6108650 Musk Aug 2000 A
6119013 Maloney Sep 2000 A
6144338 Davies Nov 2000 A
6262741 Davies Jul 2001 B1
6317686 Ran Nov 2001 B1
6377945 Risvik Apr 2002 B1
6401034 Kaplan Jun 2002 B1
6424912 Correia Jul 2002 B1
6434482 Oshida Aug 2002 B1
6470189 Hill Oct 2002 B1
6480783 Myr Nov 2002 B1
6487495 Gale Nov 2002 B1
6507785 Stefan Jan 2003 B1
6529143 Mikkola Mar 2003 B2
6571169 Miyaki May 2003 B2
6587782 Nocek Jul 2003 B1
6636803 Hartz Oct 2003 B1
6671424 Skoll Dec 2003 B1
6714205 Miyahita Mar 2004 B1
6734867 Munshi May 2004 B1
6820092 Nakano Nov 2004 B2
6940407 Miranda-Knapp Sep 2005 B2
6954697 Smith Oct 2005 B1
7054743 Smith May 2006 B1
7093286 King Aug 2006 B1
7155339 Tu Dec 2006 B2
7190839 Feather Mar 2007 B1
7373246 O'Clair May 2008 B2
7379729 Holland May 2008 B2
7385600 Marion Jun 2008 B2
7444326 Jagadish Oct 2008 B1
7480566 Laverty Jan 2009 B2
7546202 Oh Jun 2009 B2
7565239 de Silva et al. Jul 2009 B2
7627656 Anand et al. Dec 2009 B1
7706977 Soehren Apr 2010 B2
7853403 Tanaka Dec 2010 B2
7949642 Yang May 2011 B2
RE42927 Want Nov 2011 E
8095434 Puttick Jan 2012 B1
8265864 Kaufman Sep 2012 B1
8296062 Yamane Oct 2012 B2
20010021894 Sakamoto Sep 2001 A1
20010038626 Dynarski Nov 2001 A1
20010047241 Khavakh Nov 2001 A1
20020021231 Schlager Feb 2002 A1
20020037716 McKenna Mar 2002 A1
20020042819 Reichert Apr 2002 A1
20020067353 Kenyon Jun 2002 A1
20020082774 Bloebaum Jun 2002 A1
20020083062 Neal Jun 2002 A1
20020098851 Walczak Jul 2002 A1
20020130953 Riconda Sep 2002 A1
20020190861 Wentworth Dec 2002 A1
20020198694 Yang Dec 2002 A1
20030011623 Dermer Jan 2003 A1
20030033083 Nakashima Feb 2003 A1
20030034936 Ernst Feb 2003 A1
20030055555 Knockeart Mar 2003 A1
20030071728 McDonald Apr 2003 A1
20030095525 Lavin May 2003 A1
20030128211 Watanabe Jul 2003 A1
20030182052 DeLorme Sep 2003 A1
20040003125 Ichimura Jan 2004 A1
20040023645 Olsen Feb 2004 A1
20040027258 Pechatnikov Feb 2004 A1
20040030493 Pechatnikov Feb 2004 A1
20040135784 Cohen Jul 2004 A1
20040158829 Beresin Aug 2004 A1
20040185870 Matsuda Sep 2004 A1
20040203603 Pierce Oct 2004 A1
20040203873 Gray Oct 2004 A1
20040215641 Kothuri Oct 2004 A1
20040217980 Radburn Nov 2004 A1
20040220918 Scriffignano Nov 2004 A1
20040224702 Chaskar Nov 2004 A1
20040225437 Endo Nov 2004 A1
20040249568 Endo Dec 2004 A1
20050148342 Sylvain Jul 2005 A1
20050149253 Nambata Jul 2005 A1
20050188333 Hunleth Aug 2005 A1
20050228780 Diab Oct 2005 A1
20050251331 Kreft Nov 2005 A1
20050270311 Rasmussen Dec 2005 A1
20050288033 McNew Dec 2005 A1
20060005114 Williamson Jan 2006 A1
20060015513 Poyhonen et al. Jan 2006 A1
20060023626 Krausz Feb 2006 A1
20060055693 Sylthe Mar 2006 A1
20060064235 Ishikawa Mar 2006 A1
20060089788 Laverty Apr 2006 A1
20060089792 Manber Apr 2006 A1
20060105782 Brock May 2006 A1
20060116818 Chao et al. Jun 2006 A1
20060135178 Allyn Jun 2006 A1
20060155679 Kothuri Jul 2006 A1
20060167616 Yamane Jul 2006 A1
20060174209 Barros Aug 2006 A1
20060200304 Oh Sep 2006 A1
20060200308 Arutunian Sep 2006 A1
20060206586 Ling Sep 2006 A1
20060223518 Haney Oct 2006 A1
20060229802 Vertelney Oct 2006 A1
20060246922 Gasbarro Nov 2006 A1
20060253247 De Silva et al. Nov 2006 A1
20070015518 Winter Jan 2007 A1
20070036318 Gits et al. Feb 2007 A1
20070072620 Levitan Mar 2007 A1
20070083557 Leiserowitz Apr 2007 A1
20070083649 Zuzga Apr 2007 A1
20070118520 Bliss May 2007 A1
20070124216 Lucas May 2007 A1
20070153983 Bloebaum Jul 2007 A1
20070156334 Vu Jul 2007 A1
20070174259 Amjadi Jul 2007 A1
20070213043 Son Sep 2007 A1
20070219706 Sheynblat Sep 2007 A1
20070219708 Brasche Sep 2007 A1
20070233817 Johnson Oct 2007 A1
20070238472 Wanless Oct 2007 A1
20070239752 Beitman Oct 2007 A1
20070253642 Berrill Nov 2007 A1
20070281690 Altman Dec 2007 A1
20070288613 Sudame et al. Dec 2007 A1
20070298812 Singh Dec 2007 A1
20080004043 Wilson Jan 2008 A1
20080071465 Chapman Mar 2008 A1
20080082262 Silva et al. Apr 2008 A1
20080104227 Birnie May 2008 A1
20080112551 Forbes et al. May 2008 A1
20080130597 Kalhan Jun 2008 A1
20080139114 Ranganathan Jun 2008 A1
20080140307 Chen Jun 2008 A1
20080177839 Chang Jul 2008 A1
20080218407 Norda Sep 2008 A1
20080249983 Meisels Oct 2008 A1
20080261560 Ruckart Oct 2008 A1
20080268822 Johnson Oct 2008 A1
20080270366 Frank Oct 2008 A1
20080275637 Kim Nov 2008 A1
20080280599 Cheng Nov 2008 A1
20080307445 Garg Dec 2008 A1
20080319990 Taranenko Dec 2008 A1
20090009397 Taylor Jan 2009 A1
20090018840 Lutz Jan 2009 A1
20090029693 Liwell Jan 2009 A1
20090055087 Beacher Feb 2009 A1
20090061852 Feher Mar 2009 A1
20090144247 Wistrand et al. Jun 2009 A1
20090150349 Cartin Jun 2009 A1
20090171955 Merz Jul 2009 A1
20090192702 Bourne Jul 2009 A1
20090237307 Tsai Sep 2009 A1
20090325615 Mckay Dec 2009 A1
20100094550 Tsurutome et al. Apr 2010 A1
20110035141 Barker Feb 2011 A1
Foreign Referenced Citations (3)
Number Date Country
WO2006125291 Nov 2006 CA
2008 039698 Feb 2008 JP
WO2006071271 Jul 2006 WO
Non-Patent Literature Citations (8)
Entry
Ahern et al., “World Explorer: Visualizing Aggregate Data from Unstructured Text in Geo-Referened Collections,” Jun. 18-23, 2007, ACM, JCDL'07.
International Search Report in PCT/US2009/05486 dated Jan. 21, 2010.
International Search Report in PCT/US2008/10543 dated Aug. 19, 2010.
International Search Report in PCT/US2008/10542 dated Aug. 19, 2010.
European Search Report received in European Appl. No. 09819546.4 dated Oct. 15, 2012.
John Krumm et al, “Map Matching with Travel Time Constraints”, SAE 2007 World Congress, Apr. 19, 2007.
International Search Report in PCT/US2009/05486 dated Jan. 12, 2011.
U.S. Appl. No. 11/251,766, filed Oct. 2005, Barcklay.
Related Publications (1)
Number Date Country
20130184986 A1 Jul 2013 US
Provisional Applications (1)
Number Date Country
61136805 Oct 2008 US
Continuations (1)
Number Date Country
Parent 12588143 Oct 2009 US
Child 13790197 US