Generating a graphic model of a geographic object and systems thereof

• US 8,374,390 B2
• Filed: 06/24/2009
• Issued: 02/12/2013
• Est. Priority Date: 06/24/2009
• Status: Active Grant

• Current Assignee: HERE Global B.V. (Nokia Corporation)
• Original Assignee: Navteq B.V.
• Inventors: Stroila, Matei, Perelmutova, Yelena V., Kohlmeyer, Bradley
• Primary Examiner(s): STREGE, JOHN B
• Application Number:
  US12/490,676
• Publication Number:
  US 20100328316A1
• Time in Patent Office: 1,329 Days
• Field of Search: 382/103, 382/190
• US Class Current: 382/103
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06T 11/60 Editing figures and text; Combining figures or text
• CPC Class Codes: G06T 2200/08 involving all processing steps from image acquisition to 3D model generation
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06V 10/774 Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 20/582 of traffic signs
• CPC Class Codes: G06V 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G06V 30/10 Character recognition

Abstract

Systems, devices, features, and methods for constructing a graphic model of a geographic object, such as a road sign or features thereof, from an image, such as, for example, to develop a navigation database are disclosed. For example, one method comprises receiving a plurality of images of regions, such as of roads or paths. An image of the plurality of images is identified. A process to determine scale-invariant components of a geographic object in the identified image may be performed. Other processes, such as generating a dictionary to facilitate optical character recognition may be performed. A graphic model of the geographic object is generated based on the scale-invariant components and/or the optical character recognition. The generated graphic model may be associated with a navigation or map database.

Claims

1. A method of automatically constructing a graphic model of a road sign from an image to enhance a navigation database, the method comprising:

receiving a plurality of images of regions about roads;

identifying an image of the plurality of images that includes a road sign;

performing a scale-invariant feature transform on the road sign in the image;

generating a graphic model of the road sign based on results of the scale-invariant feature transform, wherein generating the graphic model includes;

matching scale-invariant components determined from performing the scale-invariant feature transform with reference scale-invariant components in a data library;

identifying a road sign symbol of the road sign in the image based on the matching; and

retrieving a graphic road sign symbol icon from a template database based on identifying the road sign symbol;

generating a local dictionary by retrieving text data of geographic features within a configurable area around a location associated with the identified image, the text data of the geographic features retrieved from the navigation database; and

associating the graphic model of the road sign with the navigation database.

2. The method of claim 1, wherein the road sign in the image comprises a traffic guide sign.

3. The method of claim 1, wherein the road sign in the image comprises an expressway sign.

4. The method of claim 1, wherein the plurality of images comprise raster images.

5. The method of claim 1, wherein the graphic model of the road sign is in a scalable vector graphics format.

6. The method of claim 1, further comprising:

performing optical character recognition on the road sign in the image based on the generated local dictionary.

7. The method of claim 6, wherein generating the graphic model is also based on results of the optical character recognition.

8. The method of claim 6, wherein the configurable area is configured based on selecting a radius about the location.

9. The method of claim 8, wherein the location comprises a latitude coordinate and a longitude coordinate.

10. The method of claim 1, further comprising:

prior to performing the scale-invariant feature transform, performing image segmentation on the road sign in the image.

11. The method of claim 10, wherein generating the graphic model includes laying out the graphic road sign symbol icon as a function of the image segmentation.

12. The method of claim 1, further comprising:

detecting an arrow of the road sign in the image based on binary matching.

25. The method of claim 1, wherein the graphic model includes the graphic road sign symbol icon positioned next to an image representation of a roadway.

26. The method of claim 1, further comprising:

receiving the location associated with the identified image, wherein the location is determined via a positioning system.

13. A computer-readable medium configured to store a computer program that performs a method of sign reconstruction from a path image, the method comprising:

receiving a plurality of images of paths;

identifying an image of the plurality of images that includes a path sign;

selecting an image representation of the path sign from the identified image;

performing image segmentation on the image representation of the path sign, the image segmentation including identifying different regions of the path sign;

generating a dictionary by retrieving words associated with geographic features from a configurable area around a location associated with the identified image;

performing optical character recognition on at least one of the different regions in the image representation of the path sign as a function of the dictionary to identify text of the path sign;

performing a scale-invariant feature transform on at least one of the different regions in the image representation of the path sign to identify a path sign feature;

generating a graphic model of the path sign based on the image segmentation, the identified text of the path sign, and the identified path sign feature; and

associating the graphic model with the master navigation database.

14. The method of claim 13, wherein generating the graphic model comprises retrieving graphic template characters corresponding to the identified text.

15. The method of claim 13, wherein generating the graphic model comprises retrieving a graphic template symbol corresponding to the identified path sign feature.

16. The method of claim 13, wherein associating the graphic model with the master navigation database includes storing the graphic model as a geographic data attribute in a road segment data record.

17. The method of claim 13, wherein associating the graphic model with the master navigation database includes linking or indexing the graphic model with a road segment data record.

18. A method for constructing a graphic model of a geographic object from an image, the method comprising:

receiving an image from a plurality of collected images of geographic areas, the image including a desired geographic object;

determining scale-invariant components of the desired geographic object in the image;

matching the determined scale-invariant components with reference scale-invariant components in a map data library;

identifying a symbol of the desired geographic object in the image based on the matching;

generating a graphic model of the desired geographic object based on the identified symbol and based on a roadway associated with the desired geographic object.

19. The method of claim 18, wherein generating the graphic model comprises selecting a symbol icon based on the identified symbol.

20. The method of claim 18, wherein the identified symbol comprises a shield symbol of a road sign.

21. The method of claim 18, wherein the map data library comprises a local library generated based on a location associated with the image.

22. A graphic model of a path object displayed on a navigation user interface, the graphic model generated from a method of reconstructing a geographic object from an image, the method comprising:

receiving a plurality of images corresponding to paths;

identifying an image of the plurality of images that includes a path object;

determining scale-invariant features of the path object in the image to identify a symbol of the path object; and

generating a graphic model of the path object based on the identified symbol of the path object in the image, wherein generating the graphic model includes retrieving a graphical icon from a template database based on the identified symbol, wherein the graphic model includes an image representation of at least one path and the path object.

23. The method of claim 22, wherein the determined scale-invariant features comprise gradient vectors.

24. The method of claim 22, wherein the graphic model includes the image representation of the at least one path and the path object.