18340000. GENERATING A MODEL FOR AN OBJECT ENCOUNTERED BY A ROBOT simplified abstract (Google LLC)

From WikiPatents
Jump to navigation Jump to search

GENERATING A MODEL FOR AN OBJECT ENCOUNTERED BY A ROBOT

Organization Name

Google LLC

Inventor(s)

Kurt Konolige of Menlo Park CA (US)

Nareshkumar Rajkumar of Cupertino CA (US)

Stefan Hinterstoisser of Munich (DE)

GENERATING A MODEL FOR AN OBJECT ENCOUNTERED BY A ROBOT - A simplified explanation of the abstract

This abstract first appeared for US patent application 18340000 titled 'GENERATING A MODEL FOR AN OBJECT ENCOUNTERED BY A ROBOT

Simplified Explanation

The abstract describes a method and apparatus for generating a model of an object that a robot encounters in its environment, but cannot recognize using its existing models. The model is created using vision sensor data captured from multiple angles by a vision sensor attached to the robot. The generated model can be used by the robot to detect the object and estimate its position.

  • The patent application focuses on generating a model for an unrecognized object encountered by a robot.
  • The model is created using vision sensor data captured from different perspectives.
  • The vision sensor is connected to the robot and captures the data.
  • The generated model can be used by the robot to detect the object in its environment.
  • The model can also be used to estimate the pose or position of the object.

Potential Applications

  • Robotics: This technology can be applied in various robotic systems where the robot needs to recognize and interact with objects that it has not encountered before.
  • Object Detection: The generated model can be used in object detection systems to identify and classify unknown objects.
  • Pose Estimation: The model can assist in estimating the position and orientation of the object, which can be useful in applications like augmented reality or robotic manipulation.

Problems Solved

  • Recognition of Unknown Objects: The technology addresses the problem of a robot encountering objects that it cannot recognize using its existing models.
  • Limited Model Database: By generating models for unrecognized objects, the robot can expand its database and improve its ability to interact with a wider range of objects.

Benefits

  • Improved Object Recognition: The generated models enable the robot to recognize and detect objects that it previously couldn't identify.
  • Enhanced Interaction: With the ability to generate models for unknown objects, the robot can better interact with its environment and perform tasks involving unfamiliar objects.
  • Accurate Pose Estimation: The generated models can assist in accurately estimating the position and orientation of the object, improving the robot's ability to interact with it effectively.


Original Abstract Submitted

Methods and apparatus related to generating a model for an object encountered by a robot in its environment, where the object is one that the robot is unable to recognize utilizing existing models associated with the robot. The model is generated based on vision sensor data that captures the object from multiple vantages and that is captured by a vision sensor associated with the robot, such as a vision sensor coupled to the robot. The model may be provided for use by the robot in detecting the object and/or for use in estimating the pose of the object.

Retrieved from "http://wikipatents.org/index.php?title=18340000._GENERATING_A_MODEL_FOR_AN_OBJECT_ENCOUNTERED_BY_A_ROBOT_simplified_abstract_(Google_LLC)&oldid=17597"