Spatio-temporal confidence maps

Patent metadata

Field	Value
Publication number	US-9269153-B2
Application number	US-201314057917-A
Country	US
Kind code	B2
Filing date	Oct 18, 2013
Priority date	Oct 18, 2012
Publication date	Feb 23, 2016
Grant date	Feb 23, 2016

Abstract

A method and an apparatus for generating a confidence map for a disparity map associated to a set of two or more images are described. Motion between at least two subsequent sets of two or more images is determined. Based on the determined motion information static and dynamic regions in the images of the sets of two or more images are detected and separated. A disparity change between a disparity value determined for a static region of a current image and a motion compensated disparity value of a previous image is determined. The result of the determination is taken into account for generating or refining a confidence map.

First claim

The invention claimed is:

1. A method for generating a confidence map for a disparity map associated to a sequence of sets of two or more images, the method comprising:
determining motion between at least two subsequent sets of two or more images;
detecting and separating static and dynamic regions in the images of the sets of two or more images based on the determined motion information;
determining a disparity value for a static region in a current image;
determining a motion compensated disparity value for said static region in a previous image;
determining a disparity change between the disparity value and the motion compensated disparity value; and
generating or refining a confidence map by increasing a confidence value for said static region if the determined disparity change is below a threshold or by applying a filter to confidence values of the confidence map, wherein output confidence values of the filter are controller by input confidence values and the determined disparity change.

2. The method according to claim 1 , wherein the motion between at least two subsequent sets of two or more images is described by optical flow maps.

3. The method according to claim 1 , wherein the filter primarily modifies those confidence values whose figures are located in a central range of a value range of the confidence values.

4. The method according to claim 1 , wherein the filter modifies all confidence values equally without preference.

5. An apparatus for generating a confidence map for a disparity map associated to a sequence of sets of two or more images, comprising:
a processor and an associated memory configured to:
determine motion between at least two subsequent sets of two or more images;
detect and separate static and dynamic regions in the images of the sets of two or more images based on the determined motion information;

determine a disparity value for a static region in a current image;
determine a motion compensated disparity value for said static region in a previous image;
determine a disparity change between the disparity value and the motion compensated disparity value; and
generate or refine a confidence map by increasing a confidence value for said static region if the determined disparity change is below a threshold or by applying a filter to confidence values of the confidence map, wherein output confidence values of the filter are controller by input confidence values and the determined disparity change.

6. The apparatus according to claim 5 , wherein the motion between at least two subsequent sets of two or more images is described by optical flow maps.

7. The apparatus according to claim 5 , wherein the filter is adapted to primarily modify those confidence values whose figures are located in a central range of a value range of the confidence values.

8. The apparatus according to claim 5 , wherein the filter is adapted to modify all confidence values equally without preference.

Assignees

• Thomson Licensing

Inventors

• Gandolph Dirk
• Jachalsky Joern
• Putzke-Roeming Wolfram
• Schlosser Markus

Classifications

• G06T2207/10021

  Stereoscopic video; Stereoscopic image sequence · CPC title

• G06T7/2006

  Physics · mapped topic

• G06T7/0075 Primary

  Physics · mapped topic

• G06T7/593

  from stereo images · CPC title

• G06T7/215

  Motion-based segmentation · CPC title