Three-dimensional background removal for vision system
AssigneeMicrosoft Corporation
InventorsMomin Al-Ghosien, Craig Peeper
Life of this patent
Estimating the term…
Abstract
A method for controlling a computer system includes acquiring video of a subject, and obtaining from the video a time-resolved sequence of depth maps. A geometric model of the subject is fit to each depth map in the sequence and tracked into a subsequent depth map in the sequence. From the subsequent depth map, a background section is selected for exclusion. The background section is one that lacks coherent motion and is located more than a threshold distance from the coordinates of the geometric model tracked in. Then, a subsequent geometric model of the subject is fit to the depth map with the background section excluded.
Claims20
A method for controlling a computer system, the method comprising:
- acquiring video of a subject in front of a background;
- obtaining from the video a time-resolved sequence of depth maps, each depth map including an array of pixels;
- fitting a geometric model of the subject to a first depth map in the sequence;
- registering coordinates of the geometric model to a second depth map in the sequence;
- selecting from the second depth map a background section lacking coherent motion and located more than a threshold distance from the coordinates of the geometric model;
- and
- refitting the geometric model of the subject to the second depth map with the background section excluded, said acquiring obtaining, fitting, registering, selecting and refitting enacted within a computer vision system of the computer system.
A method for controlling a computer system, the method comprising:
- acquiring video of a subject in front of a background;
- obtaining from the video a time-resolved sequence of depth maps, each depth map including an array of pixels;
- fitting a first skeleton of the subject to a first depth map in the sequence;
- registering the first skeleton to a second depth map in the sequence;
- for each pixel of the second depth map, incrementing a corresponding exclusion counter if that pixel has been static for a predetermined number of frames of the video and is more than a threshold distance from any skeletal segment of the first skeleton;
- selecting as a background section those pixels for which the corresponding exclusion counter is above a threshold value;
- and
- fitting a second skeleton of the subject to the second depth map with the background section excluded, said acquiring obtaining, fitting, registering, incrementing and selecting enacted within a computer vision system of the computer system.
A game system comprising:
- a vision subsystem configured to obtain from a depth camera a sequence of time-resolved depth maps imaging a player, each depth map including an array of pixels;
- a logic subsystem operatively coupled to the vision subsystem;
- and
- a data subsystem holding instructions executable by the logic subsystem to:
- fit a first skeleton of the player to non-background pixels of a first depth map in the sequence,
- identify as background pixels of a second depth map in the sequence those pixels lacking coherent motion and located outside of a predetermined range of the first skeleton, and
- fit a second skeleton of the player to non-background pixels of the second depth map.
Dependencies are inferred from the claim text.