Depth mapping based on pattern matching and stereoscopic information

Patent metadata

Field	Value
Publication number	US-9582889-B2
Application number	US-84486410-A
Country	US
Kind code	B2
Filing date	Jul 28, 2010
Priority date	Jul 30, 2009
Publication date	Feb 28, 2017
Grant date	Feb 28, 2017

Abstract

A method for depth mapping includes projecting a pattern of optical radiation onto an object. A first image of the pattern on the object is captured using a first image sensor, and this image is processed to generate pattern-based depth data with respect to the object. A second image of the object is captured using a second image sensor, and the second image is processed together with another image to generate stereoscopic depth data with respect to the object. The pattern-based depth data is combined with the stereoscopic depth data to create a depth map of the object.

First claim

The invention claimed is:

1. A method for depth mapping, comprising:
projecting a pattern of optical radiation onto an object;
capturing a first image of the pattern on the object using a first image sensor, and processing the first image alone to generate pattern-based depth data with respect to the object;
capturing a second image of the object using a second image sensor, wherein the projected pattern does not appear in the second image, and processing the second image together with the first image to generate stereoscopic depth data with respect to the object; and
combining the pattern-based depth data with the stereoscopic depth data to create a depth map of the object,
wherein combining the pattern-based depth data with the stereoscopic depth data comprises computing respective measures of confidence associated with the pattern-based depth data and stereoscopic depth data, and selecting depth coordinates from among the pattern-based and stereoscopic depth data responsively to the respective measures of confidence.

2. A method for depth mapping, comprising:
projecting a pattern of optical radiation onto an object;
capturing a first image of the pattern on the object using a first image sensor, and processing the first image alone to generate pattern-based depth data with respect to the object;
capturing a second image of the object using a second image sensor, wherein the projected pattern does not appear in the second image, and processing the second image together with the first image to generate stereoscopic depth data with respect to the object; and
combining the pattern-based depth data with the stereoscopic depth data to create a depth map of the object,
wherein combining the pattern-based depth data with the stereoscopic depth data comprises defining multiple candidate depth coordinates for each of a plurality of pixels in the depth map, and selecting one of the candidate depth coordinates at each pixel for inclusion in the depth map.

3. The method according to claim 2 , wherein the multiple candidate depth coordinates comprise, for at least some of the pixels, a null coordinate indicating that no valid depth coordinate was found.

4. The method according to claim 2 , wherein selecting the one of the candidate depth coordinates comprises applying weighted tensor voting among the pixels in order to select the one of the candidate depth coordinates based on the candidate depth coordinates at neighboring pixels.

5. A method for depth mapping, comprising:
receiving at least one image of an object, captured by an image sensor, the image comprising multiple pixels;
processing the at least one image to generate depth data comprising multiple candidate depth coordinates and respective measures of confidence associated with the candidate depth coordinates for each of a plurality of the pixels;
applying a weighted voting process to the depth data, wherein votes for the candidate depth coordinates are weighted responsively to the respective measures of confidence, in order to select one of the candidate depth coordinates at each pixel; and
outputting a depth map of the object comprising the selected one of the candidate depth coordinates at each pixel.

6. The method according to claim 5 , wherein the multiple candidate depth coordinates comprise, for at least some of the pixels, a null coordinate indicating that no valid depth coordinate was found.

7. The method according to claim 5 , wherein applying the weighted voting process comprises applying weighted tensor voting among the pixels in order to select the one of the candidate depth coordinates based on the candidate depth coordinates at neighboring pixels.

8. The method according to claim 7 , wherein applying the weighted tensor voting comprises computing a weighted sum of covariance matrices over the neighboring pixels, and selecting the one of the candidate depth coordinates based on a difference between eigenvalues of the summed covariance matrices.

9. Apparatus for depth mapping, comprising:
an illumination subassembly, which is configured to project a pattern of optical radiation onto an object;
a first image sensor, which is configured to capture a first image of the pattern on the object;
at least a second image sensor, which is configured to capture at least a second image of the object, wherein the projected pattern does not appear in the second image and
a processor, which is configured to process the first image alone to generate pattern-based depth data with respect to the object, to process a pair of images including the first image and the second image to generate stereoscopic depth data with respect to the object, and to combine the pattern-based depth data with the stereoscopic depth data to create a depth map of the object,
wherein the processor is configured to associate respective measures of confidence with the pattern-based depth data and stereoscopic depth data, and to select depth coordinates from among the pattern-based and stereoscopic depth data responsively to the respective measures of confidence.

10. Apparatus for depth mapping, comprising:
an illumination subassembly, which is configured to project a pattern of optical radiation onto an object;
a first image sensor, which is configured to capture a first image of the pattern on the object;
at least a second image sensor, which is configured to capture at least a second image of the object, wherein the projected pattern does not appear in the second image; and
a processor, which is configured to process the first image alone to generate pattern-based depth data with respect to the object, to process a pair of images including the first image and the second image to generate stereoscopic depth data with respect to the object, and to combine the pattern-based depth data with the stereoscopic depth data to create a depth map of the object,
wherein the processor is configured to define multiple candidate depth coordinates for each of a plurality of pixels in the depth map, and to select one of the candidate depth coordinates at each pixel for inclusion in the depth map.

11. The apparatus according to claim 10 , wherein the multiple candidate depth coordinates comprise, for at least some of the pixels, a null coordinate indicating that no valid depth coordinate was found.

12. The apparatus according to claim 10 , wherein the processor is configured to apply weighted tensor voting among the pixels in order to select the one of the candidate depth coordinates based on the candidate depth coordinates at neighboring pixels.

13. Apparatus for depth mapping, comprising:
at least one image sensor, which is configured to capture at least one image of an object, the image comprising multiple pixels; and
a processor, which is configured to process the at least one image to generate depth data comprising multiple candidate depth coordinates and respective measures of confidence associated with the candidate depth coordinates for each of a plurality of the pixels, to apply a weighted voting process to the depth data, wherein votes for the candidate depth coordinates are weighted responsively to the respective measures of confidence, in order to select one of the candidate depth coordinates at each pixel, and to output a depth map of the object comprising the selected one of the candidate depth coordinates at each pixel.

14. The apparatus according to claim 13 , wherein the multiple candidate depth coordinates comprise, for at least some of the pixels, a null coordinate indicating that no valid depth coordinate was found.

15. The apparatus according to claim 13 , wherein the proc

Assignees

• Shpunt Alexander
• Medioni Gerard
• Cohen Daniel
• Sali Erez
• Deitch Ronen
• Apple Inc

Inventors

• Shpunt Alexander
• Medioni Gerard
• Cohen Daniel
• Sali Erez
• Deitch Ronen

Classifications

• G06T2207/10024

  Color image · CPC title

• H04N13/254

  in combination with electromagnetic radiation sources for illuminating objects · CPC title

• H04N13/271

  wherein the generated image signals comprise depth maps or disparity maps · CPC title

• G06T7/521 Primary

  from laser ranging, e.g. using interferometry; from the projection of structured light · CPC title

• H04N13/239

  using two two-dimensional [2D] image sensors having a relative position equal to or related to the interocular distance (H04N13/243 takes precedence) · CPC title