DEPTH SENSING USING AN RGB CAMERA
First Claim
Patent Images
1. A computer-implemented method of image processing comprising:
- receiving, at a processor, at least one color image from an RGB camera, the image being of a scene comprising a plurality of surfaces;
applying the at least one color image to a trained machine learning component; and
receiving as output from the trained machine learning component, a depth map comprising, for individual image elements of the color image, distances from the RGB camera to the surfaces in the scene depicted by the individual image elements.
3 Assignments
0 Petitions
Accused Products
Abstract
A method of sensing depth using an RGB camera. In an example method, a color image of a scene is received from an RGB camera. The color image is applied to a trained machine learning component which uses features of the image elements to assign all or some of the image elements a depth value which represents the distance between the surface depicted by the image element and the RGB camera. In various examples, the machine learning component comprises one or more entangled geodesic random decision forests.
67 Citations
20 Claims
-
1. A computer-implemented method of image processing comprising:
-
receiving, at a processor, at least one color image from an RGB camera, the image being of a scene comprising a plurality of surfaces; applying the at least one color image to a trained machine learning component; and receiving as output from the trained machine learning component, a depth map comprising, for individual image elements of the color image, distances from the RGB camera to the surfaces in the scene depicted by the individual image elements. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A system to process an image, the system comprising a computing-based device configured to:
-
receive at least one color image from an RGB camera, the image being of a scene comprising a plurality of surfaces; apply the at least one color image to a trained machine learning component; and receive as output from the trained machine learning component, a depth map comprising, for individual image elements of the color image, distances from the RGB camera to the surfaces in the scene depicted by the individual image elements. - View Dependent Claims (19)
-
-
20. A computer-implemented method of image processing comprising:
-
receiving, at a processor, at least one color image from an RGB camera, the image being of a scene comprising a plurality of surfaces; applying the at least one color image to a trained machine learning component, the trained machine learning component comprising an entangled geodesic random decision forest; and receiving as output from the trained machine learning component, a depth map comprising, for individual image elements of the color image, distances from the RGB camera to the surfaces in the scene depicted by the individual image elements.
-
Specification