Patent · US Active

Deep learning-based localization of UAVs with respect to nearby pipes

US11584525B2 · kind B2 · utility

1Cited by
3References
22Claims
0Family size

Assignees

Inventors

Key dates

Filing dateNov 30, 2020
Grant dateFeb 21, 2023
Priority date
Expiry dateAug 20, 2041

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06T2207/20084
  • WIPO fieldTransport
  • WIPO sectorMechanical engineering

Abstract

A system and methodology for launching, flying and perching on a cylindrically curved surface in an environment without human intervention. The system and methodology include an environment awareness sensor device suite having a depth camera arranged to capture and output image data and 3D point cloud data of a field of view; an asset targeting unit arranged to set an asset as a destination location for a landing; a trajectory path determiner arranged to calculate a trajectory path to the destination location; a flight controller arranged to launch and fly the autonomous aerial vehicle to the destination location according to the trajectory path; a situational status determiner arranged to, in real-time, predict a location of an object with respect to the autonomous aerial vehicle based on 3D point cloud data for the object, determine the object is the asset based on a confidence score and autonomously land on the asset.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.