System, computer program and method for 3D object measurement, modeling and mapping from single imagery

• US 7,995,862 B2
• Filed: 05/09/2008
• Issued: 08/09/2011
• Est. Priority Date: 01/16/2004
• Status: Active Grant

• Current Assignee: Microsoft Corporation
• Original Assignee: Microsoft Corporation
• Inventors: Hu, Yong, Tao, C. Vincent, Jiang, Wanshou, Lenson, Peter, Croitoru, Arie
• Primary Examiner(s): Cunningham; Gregory F
• Application Number:
  US12/118,515
• Publication Number:
  US 20080310756A1
• Time in Patent Office: 1,187 Days
• Field of Search: None
• US Class Current: 382/285
• CPC Class Codes: G06T 17/05 Geographic models
• CPC Class Codes: G06T 7/50 Depth or shape recovery

Abstract

Methods, computer programs, and computer systems facilitate deriving three-dimensional measurement information and/or creating three-dimensional models and maps, from single images of three-dimensional objects. Aspects of the invention include obtaining at least one two-dimensional single image of the object, deriving three-dimensional coordinate information associated with the image, and obtaining three-dimensional measurements based on projection and/or shadow measurements of the object and metadata derived from the single image. In another aspect of the method, the method includes the further step of creating three-dimensional models or maps based on the projection and/or shadow measurements.

Claims

1. A system for deriving three-dimensional coordinate data comprising:

a computer which accepts a two-dimensional single image of an object, wherein the single image includes image data and is associated with an image geometry model (IGM);

an analyzer that derives at least one set of three-dimensional metadata from the single image;

a measurement utility that measures dimensions of the object by performing a projection measurement and a shadow measurement for the object, wherein the projection measurement is based at least in part on one or more of an IGM, height distance of the object, point-to-point distance of the object, line-of-sight distance of the object, volume, and bearing, and further wherein performing said shadow measurement includes determining a length of the shadow of the object based on a height of the object and an altitude of the sun; and

a model generation utility that derives three-dimensional coordinate data based at least in part on the projection, the shadow, and the three-dimensional metadata.

2. The system of claim 1, wherein the IGM is a geometric relationship between a space of the object and a space of the single image.

3. The system of claim 2, wherein the shadow measurement is based at least in part on an IGM, height, point-to-point distance of the objects, the objects may be static or dynamic.

4. The system of claim 2, wherein the IGM utilizes at least one of a physical sensor model, a generalized sensor model, a rational function model (RPM), collinearity equations, direct linear transformations, or any combination thereof.

5. The system of claim 1, wherein the object is a single static object, a compound object, contains multiple sub-objects, stationary, dynamic, real, synthetic, or any combination thereof.

6. The system of claim 1, wherein the projection measurement determines an elevation for the object based at least in part on a digital terrain model (DTM).

7. The system of claim 1, further comprising an interface component that displays content including at least one of the two-dimensional single image, the object, the three-dimensional metadata, the projection, the shadow, the object space, the image space, or any combination thereof, the interface component further accepts an input from a user that permits the user to interact with and manipulate the displayed contents.

8. The system of claim 1, wherein the analyzer compensates for systematic biases occurring in horizontal and vertical directions when capturing the single images.

9. A non-transitory computer software product comprising a computer program that, when executed on a computer, causes the computer to perform a method for creating three-dimensional models, the method comprising:

receiving at least one two-dimensional single image for an object, wherein the single image includes image data, and further wherein the single image is associated with an image geometry model (IGM);

deriving at least one set of three-dimensional metadata from at least one of the single images;

measuring a projection and a shadow measurement for at least one of the objects, wherein said shadow measurement is based, at least in part, on an altitude of the sun; and

generating three-dimensional coordinate data based at least in part on the projection, the shadow and the three-dimensional metadata.

10. The software product of claim 9, wherein the IGM is a geometric relationship between a space of the object and a space of the single image.

11. The software product of claim 9, further comprising measuring a derivative of the projection and a derivative of the shadow.

12. The software product of claim 9, wherein one of the single images is at least one of a satellite image, an aerial image, a ground image, or an image based on a rational function model (RPM).

13. The software product of claim 9, wherein measuring the projection and the shadow includes iterative adjustments of locating an end point of the shadow of the object, updating a base point position, and connecting the base point and the shadow endpoint.

14. The software product of claim 9, further comprising compensating for systematic biases in the projection measurement and shadow measurement.

15. The software product of claim 9, wherein the projection measurement determines an elevation for the object based at least in part on a digital terrain model (DTM).

16. The software product of claim 9, wherein measuring the shadow includes selecting a base point in the image, calculating a set of ground coordinates of the base point, adjusting the height, calculating at least one offset of an estimated shadow endpoint relative to the base point, and casting the shadow endpoint to the image.

17. The software product of claim 9, the three-dimensional metadata includes an object geometry.

18. The software product of claim 9, wherein receiving the at least one two-dimensional single image includes acquiring the image utilizing at least one of a frame sensor, pushbroom sensor, or any combination thereof.

19. A method for deriving three-dimensional measurement information comprising:

means for receiving at least one two-dimensional single image for an object, wherein the at least one two-dimensional single image includes image data and is associated with an image geometry model;

means for deriving at least one set of three-dimensional metadata from at least one of the single images;

means for measuring a projection and a shadow measurement for at least one of the objects, wherein said means for measuring a shadow measurement include a computer that executes a computer program, wherein the computer program causes the computer to determine a length of the shadow of the at least one of the objects based, at least in part, on an altitude of the sun; and

means for generating three-dimensional coordinate data based at least in part on the projection, the shadow, and the three-dimensional metadata.