Teaching a robot to place objects in dynamic environments
AssigneeFujitsu Limited
InventorSamir Parikh
Life of this patent
Estimating the term…
Abstract
A method may include obtaining first images, each image including an object, and determining a set of one or more visual cues for each. The method may include selecting a common visual cue of the one or more visual cues that is common to each set of one or more visual cues determined for each corresponding image and determining a correlation between a location of the common visual cue in each image of the first images and a location of the object in each image of the first images. The method may include obtaining a second image of an environment and identifying the common visual cue in the second image. The method may include determining a placement location for the object in the environment based on the correlation and a location of the common visual cue in the second image.
Claims20
A method of placing an object in a location, the method comprising:
- obtaining a first plurality of images, each image of the first plurality of images including an object;
- determining a set of one or more visual cues for each image of the first plurality of images;
- selecting a common visual cue of the one or more visual cues that is common to each set of one or more visual cues determined for each corresponding image;
- determining a correlation between a location of the common visual cue in each image of the first plurality of images and a location of the object in each image of the first plurality of images;
- obtaining a second image of an environment;
- identifying the common visual cue in the second image;
- and
- based on the correlation and based on a location of the common visual cue in the second image, determining a placement location for the object in the environment.
A non-transitory computer-readable medium having computer-executable instructions stored thereon that are executable by a processor to perform or control performance of operations comprising:
- obtaining a first plurality of images, each image of the first plurality of images including an object;
- determining a set of one or more visual cues for each image of the first plurality of images;
- selecting a common visual cue of the one or more visual cues that is common to each set of one or more visual cues determined for each corresponding image;
- determining a correlation between a location of the common visual cue in each image of the first plurality of images and a location of the object in each image of the first plurality of images;
- obtaining a second image of an environment;
- identifying the common visual cue in the second image;
- and
- based on the correlation and based on a location of the common visual cue in the second image, determining a placement location for the object in the environment.
A system to place an object in a location, the system comprising:
- a camera;
- a learning module, the learning module configured to:
- obtain a first plurality of images, each image of the first plurality of images including an object;
- determine a set of one or more visual cues for each image of the first plurality of images;
- select a common visual cue of the one or more visual cues that is common to each set of one or more visual cues determined for each corresponding image;
- determine a correlation between a location of the common visual cue in each image of the first plurality of images and a location of the object in each image of the first plurality of images;
- a replay module communicatively coupled to the camera and the learning module, the replay module configured to:
- obtain a second image of an environment from the camera;
- identify the common visual cue in the second image;
- and
- based on the correlation and based on a location of the common visual cue in the second image, determine a placement location for the object in the second image;
- and
- a robotic device communicatively coupled to the replay module and configured to place the object at the placement location in the environment in accordance with instructions received from the replay module.
Dependencies are inferred from the claim text.