Automatic moving object verification

Patent metadata

Field	Value
Publication number	US-10043104-B2
Application number	US-201514962269-A
Country	US
Kind code	B2
Filing date	Dec 8, 2015
Priority date	Jan 5, 2015
Publication date	Aug 7, 2018
Grant date	Aug 7, 2018

Abstract

A method for determining a likelihood that a first object captured in a first image and a second object captured in a second image are the same object includes capturing the first image from a first viewpoint and a second image from a second viewpoint, wherein the first object is in the first image, and the second object is in the second image. The method also includes determining a first likelihood that a first visual feature on the first object and a second visual feature on the second object are the same visual feature, and determining a second likelihood that a dimension of the first object and a corresponding dimension of the second object are the same. The method then includes determining a final likelihood that the first object and the second object are the same object based at least partially upon the first likelihood and the second likelihood.

First claim

What is claimed is:

1. A method for analyzing images, the method comprising:
capturing, using a first camera, a first image from a first viewpoint, wherein a first object at a first geographic location is in the first image;
capturing, using a second camera, a second image from a second viewpoint, wherein a second object at a second geographic location is in the second image;
determining, using a processor, a first percentage that a first visual feature on the first object and a second visual feature on the second object are the same visual feature;
determining, using the processor, a second percentage that a dimension of the first object and a corresponding dimension of the second object are the same;
determining, using the processor, a final percentage that the first object and the second object are the same object based at least partially upon the first percentage and the second percentage; and
storing the final percentage in a memory device.

2. The method of claim 1 , wherein the first percentage is based at least partially upon a similarity of an appearance of the first visual feature on the first object and an appearance of the second visual feature on the second object.

3. The method of claim 2 , wherein the first percentage is also based at least partially upon a location of the first visual feature on the first object and a location of the second visual feature on the second object.

4. The method of claim 1 , further comprising determining, using the processor, a third percentage that the first object is in the same geographic location as the second object at a time that the second image is captured.

5. The method of claim 1 , wherein the first image is captured by the first camera on a satellite, and the second image is captured by the second camera on an unmanned aircraft system.

6. The method of claim 1 , further comprising:
generating a model of the second object based at least partially upon the second image; and
wherein determining the second percentage comprises:
determining the second percentage that the dimension of the first object and the corresponding dimension of the second object are the same based at least partially on the model of the second object.

7. The method of claim 1 , wherein determining the final percentage comprises calculating a Bayesian probability using the first percentage and the second percentage.

8. The method of claim 7 , wherein the first object and the second object are determined to be the same object when the Bayesian probability is greater than or equal to a predetermined percentage.

9. The method of claim 1 , wherein the first object is a first ship, and the second object is a second ship.

10. The method of claim 1 , further comprising transmitting an alert to a user indicating the final percentage.

11. The method of claim 10 , further comprising providing an application to the user for installation on a computer system, wherein the alert is transmitted over a wireless communication channel to a wireless device, and wherein the alert causes the computer system to auto-launch the application when the wireless device is connected to the computer system.

12. The method of claim 11 , wherein the computer system is offline when the alert is transmitted.

13. The method of claim 12 , wherein the alert includes a uniform resource locator that specifies a data source where the first and second images are stored, and wherein the alert enables connection via the uniform resource locator to the data source over the Internet when the wireless device is connected to the computer system and the computer system comes online.

14. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a computer system, cause the computer system to perform operations, the operations comprising:
providing an application to a user for installation on a second computer system;
capturing a first image, using a first camera, from a first viewpoint, wherein a first object at a first geographic location is in the first image;
capturing a second image, using the first camera, from a second viewpoint, wherein a second object at a second geographic location is in the second image;
determining, using a processor, a first percentage that a first visual feature on the first object and a second visual feature on the second object are the same visual feature;
determining, using the processor, a second percentage that a dimension of the first object and a corresponding dimension of the second object are the same;
determining, using the processor, a third percentage that the first object is in the same geographic location as the second object at a time that the second image is captured;
determining, using the processor, a final percentage that the first object and the second object are the same object based at least partially upon the first percentage, the second percentage, and the third percentage;
storing the final percentage in a memory device; and
transmitting an alert, which indicates the final percentage, over a wireless communication channel to a wireless device when the second computer system is offline, wherein the alert causes the second computer system to auto-launch the application when the wireless device is connected to the second computer system.

15. A system for analyzing images, the system comprising:
a cross-view transform and matching device generating a first percentage that a first visual feature on a first object in a first image and a second visual feature on a second object in a second image are the same visual feature, wherein the first object is at a first geographic location when captured by a first camera, and the second object is at a second geographic location when captured by a second camera;
a geometric verification device generating a second percentage that a location of the first visual feature on the first object and a location of the second visual feature on the second object are the same location;
an object model generation device generating a third percentage that a length of a portion of the first object and a length of a corresponding portion of the second object are the same length;
a geographic verification device generating a fourth percentage that the first object is in the same geographic location as the second object at a time that the second image is captured; and
a decision device determining a final percentage that the first and second objects are the same object based upon the first percentage from the cross-view transform and matching engine, the second percentage from the geometric verification engine, the third percentage from the object model generation unit, and the fourth percentage from the geographic verification engine.

16. The system of claim 15 , wherein the cross-view transform and matching device is trained using a plurality of visual features from a plurality of images using kernel tensor analysis.

17. The system of claim 15 , wherein the second percentage of the geometric verification device is computed based on fitting the first visual feature to a model of the second object.

18. The system of claim 15 , wherein the object model generation device generates a three dimensional model of the second object.

19. The system of claim 18 , wherein the length of the portion of the second object is recovered from the model.

20. The system of claim 15 , wherein the fourth percentage of the geographic verification device is based upon a distance and time between a last known geographic lo

Assignees

• Avigilon Fortress Corp

Inventors

• Qian Gang
• Rasheed Zeeshan

Classifications

• G06V20/17

  taken from planes or by drones · CPC title

• G06V10/761

  Proximity, similarity or dissimilarity measures · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06V20/176 Primary

  Urban or other man-made structures · CPC title

• G06K9/00637

  Physics · mapped topic