Validation and correction of map data using oblique images

• US 8,311,287 B2
• Filed: 09/25/2008
• Issued: 11/13/2012
• Est. Priority Date: 09/25/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Mishra, Pragyana K., Ofek, Eyal, Kimchi, Gur
• Primary Examiner(s): Werner, Brian P
• Application Number:
  US12/237,817
• Publication Number:
  US 20100074538A1
• Time in Patent Office: 1,510 Days
• Field of Search: None
• US Class Current: 382/113
• CPC Class Codes: G06F 18/2451 linear, e.g. hyperplane
• CPC Class Codes: G06T 17/05 Geographic models
• CPC Class Codes: G06T 2207/10032 Satellite or aerial image; Remote sensing
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection

Abstract

Technologies are described herein for validating and correcting map data using oblique images or aerial photographs taken at oblique angles to the earth'"'"'s surface. Pixels within oblique images can be analyzed to detect, validate, and correct other sources of data used in generating maps such as vector data, elevation maps, projection parameters, and three-dimensional model data. Visibility and occlusion information in oblique views may be analyzed to reduce errors in either occluding or occluded entities. Occlusion of road segments due to foliage, z-ordering of freeways, tunnels, bridges, buildings, and other geospatial entities may be determined, validated, and corrected. A learning algorithm can be trained with image-based descriptors that encode visible data consistencies. After training, the algorithm can classify errors and inconsistencies using combinations of different descriptors such as color, texture, image-gradients, and filter responses.

Claims

1. A method for validating a map using an oblique image, the method comprising:

sampling an image region of the oblique image;

computing a descriptor vector of the sampled image region;

applying a classifier function to the descriptor; and

classifying the sampled image region as consistent or inconsistent in response to applying the classifier.

2. The method of claim 1, further comprising correcting map data associated with the oblique image in response to classifying the sampled image region as inconsistent.

3. The method of claim 1, further comprising reducing uncertainty in map data associated with the oblique image in response to classifying the sampled image region.

4. The method of claim 1, wherein the descriptor vector comprises statistical moments of pixel colors within the image region.

5. The method of claim 1, wherein the descriptor vector comprises a normalized gradient of pixels within the image region.

6. The method of claim 1, wherein the descriptor vector comprises responses associated with a steerable filter.

7. The method of claim 1, further comprising constructing the classifier based on hyperplanes that separate training data by a maximal margin.

8. The method of claim 7, wherein training data comprises descriptor vectors associated with known image data.

9. The method of claim 1, wherein classifying the sampled image region comprises detecting an error in vector data associated with the map.

10. The method of claim 1, wherein classifying the sampled image region comprises detecting an error in elevation data associated with the map.

11. A computer storage medium having computer executable instructions for validating a map using an oblique image stored thereon which, when executed by a computer, cause the computer to:

sample an image region of the oblique image;

compute a descriptor vector of the sampled image region;

apply a classifier function to the descriptor; and

classify the sampled image region as consistent or inconsistent in response to applying the classifier.

12. The computer storage medium of claim 11, further causing the computer to correct map data associated with the oblique image in response to classifying the sampled image region as inconsistent.

13. The computer storage medium of claim 11, further causing the computer to reduce uncertainty in map data associated with the oblique image in response to classifying the sampled image region.

14. The computer storage medium of claim 11, wherein the descriptor vector comprises statistical moments of pixel colors within the image region.

15. The computer storage medium of claim 11, wherein the descriptor vector comprises a normalized gradient of pixels within the image region.

16. The computer storage medium of claim 11, wherein the descriptor vector comprises a response of a difference of Gaussian applied to pixels within the image region.

17. The computer storage medium of claim 11, further causing the computer to construct the classifier based on hyperplanes that separate training data by a maximal margin.

18. The computer storage medium of claim 11, wherein classifying the sampled image region comprises detecting an error in vector data associated with the oblique image.

19. The computer storage medium of claim 11, wherein classifying the sampled image region comprises detecting an error in elevation data associated with the oblique image.

20. A method for validating a map using an oblique image, the method comprising:

establishing training data comprising descriptor vectors associated with known image data;

establishing a classifier based on hyperplanes that separate the training data by a maximal margin;

sampling an image region of the oblique image;

computing a descriptor vector of the sampled image region of the oblique image;

applying a classifier function to the descriptor vector of the sampled image region;

classifying the sampled image region as consistent or inconsistent in response to applying the classifier; and

correcting map data associated with the oblique image in response to classifying the sampled image region as inconsistent.