Map aggregation

• US 8,266,132 B2
• Filed: 03/03/2009
• Issued: 09/11/2012
• Est. Priority Date: 03/03/2009
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Ofek, Eyal, Epshtein, Boris
• Primary Examiner(s): Leroux, Etienne
• Assistant Examiner(s): Nguyen, Cindy
• Application Number:
  US12/396,496
• Publication Number:
  US 20100225665A1
• Time in Patent Office: 1,288 Days
• Field of Search: 707/706, 707/709, 345/629
• US Class Current: 707/706
• CPC Class Codes: G06V 30/422 Technical drawings; Geographical maps

Abstract

A system described herein includes a text extractor component that extracts text from a digital image and a determiner component that automatically determines whether or not the digital image is a map of a geographic region based at least in part upon the extracted text. The system additionally includes a correlator component that generates correlation data that causes the digital image to be correlated with a portion of a reference map that pertains to the geographic region if the determiner component determines that the digital image is a map of the geographic region.

Claims

1. A method comprising the following computer-executable acts:

automatically locating a digital image, the digital image being a candidate map of a geographic region;

automatically determining whether or not the digital image is a map of a geographic region, wherein automatically determining whether or not the digital image is a map of the geographic region comprises;

extracting text from the digital image;

comparing the extracted text with a database of known points of interest for the geographic region to locate matches between the extracted text and the known points of interest for the geographic region; and

determining whether or not there is spatial correspondence between the extracted text in the digital image and matched known points of interest, wherein the digital image is determined to be a map of the geographic region if and only if there is spatial correspondence between the extracted text in the digital image and the matched known points of interest; and

if the digital image is determined to be a map of the geographic region, automatically generating correlation data that correlates the digital image with a reference map that includes the geographic region.

2. The method of claim 1, wherein automatically generating correlation data that correlates the digital image with a reference map that includes the geographic region comprises correlating matches between the extracted text and entries in the database of known points of interest.

3. The method of claim 2, wherein determining if matches between the extracted text and entries in the database of known points of interest have a spatial correspondence comprises back-projecting the digital image onto the reference map.

4. The method of claim 1, further comprising:

receiving a plurality of digital images from a web crawler; and

performing the automatically determining and the automatically generating for each of the plurality of images.

5. The method of claim 1, further comprising:

receiving a user query for a map of the geographic region;

displaying on a computer screen the reference map of the geographic region and an indication that the digital image is also a map of the geographic region.

6. The method of claim 5, wherein the indication is a link that, upon selection, causes the digital image to be displayed on the computer screen.

7. The method of claim 6, wherein at least a portion of the digital image is displayed concurrently with the reference map.

8. The method of claim 7, wherein the at least the portion of the digital image overlays at least a portion of the reference map.

9. The method of claim 5, wherein the indication is a concurrent display of the digital image with the reference map.

10. The method of claim 1, further comprising:

modifying the digital image such that a coordinate system of the digital image and a coordinate system of the reference map correspond.

11. The method of claim 1, wherein determining whether or not there is spatial correspondence between the extracted text in the digital image and matched known points of interest comprises determining whether the matched known points of interest located in the digital image translate to corresponding points of interest in the reference map within a threshold radius of error.

12. The method of claim 1, wherein determining whether or not there is spatial correspondence between the extracted text in the digital image and matched known points of interest comprises:

constructing a Voronoi diagram on the reference map based at least in part upon the matched known points of interest in the reference map, the Voronoi diagram comprising a plurality of cells;

back-projecting the digital image onto the reference map; and

determining whether the matched points of interest on the digital image lie within corresponding cells of the Voronoi diagram on the reference map.

18. The system of claim 1, further comprising an image locator component that crawls the World Wide Web to locate the digital image.

13. A system comprising the following computer-executable components:

a text extractor component that extracts text from a digital image;

a determiner component that automatically determines whether or not the digital image is a map of a geographic region, wherein the determiner component determines whether or not the digital image is a map of a geographic region by comparing the text extracted from the digital image with entries in a database of known points of interest in the geographic region to identify matches, and subsequently determining whether the matched text extracted from the digital image has a spatial correspondence with the known points of interest on a reference map; and

a correlator component that generates correlation data that causes the digital image to be correlated with a portion of the reference map that pertains to the geographic region if the determiner component determines that the digital image is a map of the geographic region;

wherein the text extractor component, the determiner component, and the correlator component are implemented on one or more processors.

14. The system of claim 13, further comprising a display component that displays an indication on a computer display screen that the digital image is a map of the geographic region.

15. The system of claim 14, wherein the display component displays the indication on the computer screen upon receipt of a query from a user pertaining to viewing a map of the geographic region.

16. The system of claim 15, wherein the display component causes at least a portion of the digital image to be displayed concurrently with at least a portion of the reference map.

17. The system of claim 13, wherein the indication is a link to the digital image.

20. The system of claim 13, wherein the determiner component comprises a spatial comparer component that constructs a Voronoi diagram on the reference map based upon locations of the known points of interest on the reference map and back-projects the image onto the Voronoi diagram, wherein the Voronoi diagram comprises a plurality of cells, and wherein the determiner component determines whether or not the digital image is a map of the geographic region based upon spatial correspondence between the known points of interest on the digital image and respective cells of the Voronoi diagram.

19. A computer-readable storage device comprising instructions that, when executed by a processor, cause the processor to perform acts comprising:

receiving a digital image;

extracting text from the digital image;

comparing the extracted text with a database of known points of interest;

if a sufficient match exists between the extracted text and contents of the database of known points of interest, analyzing a spatial correspondence between the extracted text and entries in the database of known points of interest;

determining whether or not the digital image is a map of a geographic region based at least in part upon the analyzing of the spatial correspondence between the extracted text and the entries in the database of known points of interest; and

if the digital image is a map of the geographic region, correlating a portion of the reference map that corresponds to the geographic region with the digital image.