Method for Determining Scattered Disparity Fields in Stereo Vision

• US 20090207235A1
• Filed: 11/30/2005
• Published: 08/20/2009
• Est. Priority Date: 11/30/2005
• Status: Active Grant

• Current Assignee: Telecom Italia S.p.A.
• Original Assignee: Telecom Italia S.p.A.
• Inventors: Presti, Giovanni, Francini, Gianluca, Lepsoy, Skjalg
• Granted Patent: US 8,275,195 B2
• Time in Patent Office: Days
• US Class Current: 348/46
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 7/33 using feature-based methods
• CPC Class Codes: G06V 10/147 Details of sensors, e.g. sensor lenses fingerprint or palmprint sensors G06V40/13; vascular sensors G06V40/145; eye sensors G06V40/19

Abstract

In a system for stereo vision including two cameras shooting the same scene, a method is performed for determining scattered disparity fields when the epipolar geometry is known, which includes the steps of: capturing, through the two cameras, first and second images of the scene from two different positions; selecting at least one pixel in the first image, the pixel being associated with a point of the scene and the second image containing a point also associated with the above point of the scene; and computing the displacement from the pixel to the point in the second image minimising a cost function, such cost function including a term which depends on the difference between the first and the second image and a term which depends on the distance of the above point in the second image from a epipolar straight line, and a following check whether it belongs to an allowability area around a subset to the epipolar straight line in which the presence of the point is allowed, in order to take into account errors or uncertainties in calibrating the cameras.

Claims

1-11. -11. (canceled)

12. A method for determining scattered disparity fields in stereo vision, comprising the steps of:

capturing a first and a second image of a same scene from two different positions, so that at least one set of pixels of said first image is associated with a corresponding set of points on said second image;

selecting at least one pixel in said first image; and

determining a point of said second image associated with said pixel, wherein determining said point comprises minimising a cost function having as a variable a disparity vector between said pixel and said point and depending on a difference between the first and the second image in a window centered on said pixel and in a corresponding window centered on the point to be determined,

said cost function comprising a term which depends in a monotonously increasing way on the distance of said point to be determined from an epipolar straight line in said second image associated with said pixel.

13. The method according to claim 12, wherein said first and second images are captured through a first camera and a second camera, respectively, and said term depends on calibration parameters of said first and second cameras.

14. The method according to claim 13, wherein said term is weighed depending on an uncertainity of camera calibration data.

15. The method according to claim 12, wherein said term is proportional to the square of the distance between said point and the epipolar line.

16. The method according to claim 12, wherein the step of determining the point of said second image associated with said pixel comprises delimiting in said second image an allowability area of said point around said epipolar straight line.

17. The method according to claim 16, wherein said allowability area is contained in a half plane delimited by a straight line perpendicular to said epipolar straight line.

18. The method according to claim 12, wherein said cost function comprises parameters comprised of brightness gradients in the pixels of said window centered on said pixel in said second image, and of brightness differences between pixels of said window centered on said pixel and pixels of said second image in the same window.

19. The method according to claim 12, wherein the step of determining the point of said second image associated with said pixel comprises associating a relative weight to all pixels contained in said window centered on said pixel.

20. The method according to claim 19, wherein the weight associated with pixels in a central area of said window centered on said pixel is greater than the weight associated with pixels in an external area of said window.

21. The method according to claim 12, wherein the step of determining said point comprises said step of minimising the cost function.

22. A system for stereo vision, comprising a first and a second shooting cameras adapted to shoot a same scene and a processing unit adapted to receive from said first and second cameras respective simultaneous images of said scene, said processing unit capable of being configured for processing said respective images according to claim 12.