Computer vision system and method using a depth sensor
AssigneeMicrosoft Technology Licensing, LLC
InventorsCharles F. Olivier, III, Jean Sebastien Fouillade, Ashley Feniello, Jordan Correa, Russell Sanchez, Malek Chalabi
Life of this patent
Estimating the term…
Abstract
A robot is provided that includes a processor executing instructions that generate an image. The robot also includes a depth sensor that captures depth data about an environment of the robot. Additionally, the robot includes a software component executed by the processor configured to generate a depth map of the environment based on the depth data. The software component is also configured to generate the image based on the depth map and red-green-blue (RGB) data about the environment.
Claims20
A robot, comprising:
- a processor executing instructions that generate an image, the processor issuing control signals corresponding to the image to be generated;
- a depth sensor that captures depth data about an environment of the robot;
- and
- a software component executed by the processor configured to:
- generate a depth map of the environment based on the depth data, and generate the image based on the depth map and red-green-blue (RGB) data about the environment;
- determine that an illumination level falls below a specified threshold;
- generate the image in response to determining that the illumination level falls below the specified threshold;
- and
- convert depth data into image data by associating color intensity with depth value, wherein depth data is represented as an illumination calculated from depth data as a gray scale image such that natural illumination is simulated according to an inverse-square relationship between the calculated illumination and the depth data.
A mobile device, comprising:
- a processor;
- a depth sensor configured to capture depth data about an environment of the robot;
- a camera configured to capture red-green-blue (RGB) data about the environment;
- and
- a software component executable by the processor, the software component configured to direct the processor to:
- generate a depth map of the environment based on the depth data;
- and
- generate an image based on the depth map and the RGB data
- determine that an illumination level falls below a specified threshold;
- generate the image in response to determining that the illumination level falls below the specified threshold;
- and
- convert depth data into image data by associating color intensity with depth value, wherein depth data is represented as an illumination calculated from depth data as a gray scale image such that natural illumination is simulated according to an inverse-square relationship between the calculated illumination and the depth data.
A method for generating an image, comprising:
- generating a depth map of an environment of the robot;
- capturing red-green-blue (RGB) data about the environment;
- determining that an illumination level falls below a specified threshold;
- generating, in response to determining that the illumination level falls below the specified threshold, an image based on the depth map and the RGB data about the environment;
- and
- converting depth data into image data by associating color intensity with depth value, wherein depth data is represented as an illumination calculated from depth data as a gray scale image such that natural illumination is simulated according to an inverse-square relationship between the calculated illumination and the depth data.
Dependencies are inferred from the claim text.