Method and system for autonomously developing or augmenting geographical databases by mining uncoordinated probe data

• US 6,385,539 B1
• Filed: 08/13/1999
• Issued: 05/07/2002
• Est. Priority Date: 08/13/1999
• Status: Expired due to Term

• Current Assignee: 21ST CENTURY GARAGE LLC
• Original Assignee: Daimler Chrysler Company LLC (Stellantis NV)
• Inventors: Rogers, Seth Olds, Wilson, Christopher Kenneth Hoover, Langley, Patrick Wyatt
• Primary Examiner(s): Nguyen, Tan
• Assistant Examiner(s): Tran, Dalena
• Application Number:
  US09/373,559
• Time in Patent Office: 998 Days
• Field of Search: 701/117, 701/118, 701/119, 701/208, 701/213, 701/200, 701/207, 701/212, 701/214, 701/215, 342/454, 342/456, 342/357.13, 342/357.17, 702/5, 340/995, 340/905, 340/989, 340/910, 340/913, 340/2.7, 340/286.14, 340/933, 340/934, 340/936, 340/988
• US Class Current: 701/213
• CPC Class Codes: G01C 21/30 Map- or contour-matching
• CPC Class Codes: G01C 21/3822 Road feature data, e.g. slope data
• CPC Class Codes: G01C 21/3841 Data obtained from two or more sources, e.g. probe vehicles
• CPC Class Codes: G08G 1/0112 from the vehicle, e.g. floating car data [FCD]
• CPC Class Codes: G08G 1/0129 for creating historical data or processing based on historical data
• CPC Class Codes: G09B 19/10 Modelling

Abstract

A method and system are provided for developing geospatial information concerning a particular area using a plurality of uncoordinated probe vehicles moving in the area. The probe vehicles are equipped with a position determining system and gather probe vehicle specific geospatial information as the vehicles move in the area. This probe vehicle specific information is combined over time into a data set. A central processor analyzes the data set to determine geospatial information having a higher quality than that gathered by the individual probe vehicles for the particular area.

Claims

1. A method for autonomously developing substantially static geospatial information concerning a particular area using a plurality of uncoordinated probe systems moving in said area, the method comprising the acts of:

obtaining probe system specific geospatial information of a lower quality from only the plurality of probe systems moving in said area;

combining said probe system specific information obtained over time into a data set;

analyzing said data set to determine substantially static geospatial information of a higher quality;

wherein said probe system specific geospatial information is position information;

wherein the probe systems are probe vehicles, and further comprising the act of developing map information on the particular area based on the higher quality geospatial information;

wherein the act of developing the map information comprises the act of deriving at least one of roadway and vehicle lane geometry information; and

wherein the act of deriving at least one of roadway and vehicle lane geometry information comprises the acts of;

obtaining a baseline trace of the position information of an initial probe vehicle to determine an initial virtual roadway centerline; and

refining the initial virtual roadway centerline based on additional traces of position information from subsequent probe vehicles traveling the roadway to derive the location of the roadway.

2. The method according to claim 1, wherein the map information is a road map, and further comprising the act of deriving traffic control information for the road map based on the higher quality geospatial information.

3. The method according to claim 1, further comprising the act of evaluating said data set to infer environmental driving information for the particular area.

4. The method according to claim 1, wherein the act of refining comprises the acts of iteratively performing a weighted averaging of a precedingly determined virtual roadway centerline with each additional position trace to determine a refined virtual roadway centerline.

5. The method according to claim 4, further comprising the acts of:

calculating offsets from the virtual roadway centerline for each position in a position trace for a plurality of said position traces; and

clustering the offsets to determine the vehicle lane geometry information for the roadway.

6. The method according to claim 5, wherein the act of clustering is performed using a hierarchical agglomerative clustering-type algorithm.

7. The method according to claim 1, wherein the probe systems are probe vehicles, and further comprising the act of deriving traffic control information based on the higher quality geospatial information.

8. The method according to claim 1, wherein said probe specific geospatial information comprises both a position of the probe system and time information on the position.

9. The method according to claim 1, wherein the act of obtaining probe specific geospatial information comprises the act of equipping each probe system with a global positioning system receiver.

10. The method according to claim 9, wherein said probe specific geospatial information comprises signal interruption information from the global positioning system receiver, and further comprising the act of developing map information of obstructions above the probe systems in the particular area based on the higher quality geospatial information of the signal interruption information.

11. The method according to claim 1, wherein said analyzing act averages the data set to determine the higher quality geospatial information.

12. A system for developing substantially static geospatial information concerning a particular area, comprising:

a plurality of probe vehicles operable to move in the area in an uncoordinated manner, each probe vehicle including at least a position sensor to detect its position with a defined accuracy as it moves in the area, and a communications device to communicate the detected positions;

a central processor which receives the detected positions from the plurality of probe vehicles, the central processor comprising means for combining only the detected positions into a data set and means for analyzing the data set to derive the substantially static geospatial information having a higher level of accuracy than that detected by the individual probe vehicles;

wherein the geospatial information at a higher level of accuracy is roadway lane information, said means for analyzing the data set to determine the roadway lane information comprising;

means for determining a virtual centerline of the roadway based on position traces formed by the detected positions from each probe vehicle;

means for calculating offsets from the virtual centerline for each position in a position trace for each of the position traces; and

means for clustering the offsets to determine the vehicle lane information for the roadway.

13. The system according to claim 12, wherein the means for analyzing averages the detected positions.

14. The system according to claim 12, wherein the geospatial information includes at least one of road geometry and traffic control information.

15. The system according to claim 12, wherein the means for obtaining a virtual centerline comprises:

means for obtaining an initial roadway centerline from an initial position trace from a probe vehicle; and

means for refining the initial roadway centerline based on additional position traces from the plurality of probe vehicles.

16. The system according to claim 15, wherein the means for refining the initial roadway centerline comprises means for iteratively performing a weighted averaging of a precedingly determined roadway centerline with each additional position trace to determine the virtual roadway centerline.

17. A method for autonomously developing vehicle lane information for a roadway map using a plurality of probe vehicles supplying position traces when traveling, the method comprising the acts of:

obtaining a baseline roadway centerline from an initial position trace from one of the probe vehicles;

refining the baseline roadway centerline to obtain a virtual centerline by iteratively performing a weighted averaging of the baseline roadway centerline with a number of position traces from the plurality of probe vehicles;

calculating an offset from the virtual centerline for each position in the position traces supplied by the plurality of probe vehicles; and

clustering the offsets into vehicle lanes and averaging the clustered offsets to determine each lane centerline.