Computer Vision Techniques Applied to Space Object Detect, Track, ID, and Characterize

Abstract

Space-based object detection and tracking represents a fundamental step necessary for detailed analysis of space objects. Initial observations of a resident space object (RSO) may result from careful sensor tasking to observe an object with well understood dynamics, or measurements-of-opportunity on an object with poorly understood dynamics. Dim and eccentric objects present a particular challenge which requires more dynamic use of imaging systems. As a result of more stressing data acquisition strategies, advanced techniques for the accurate processing of both point and streaking sources are needed. This paper will focus on two key methods in computer vision used to determine interest points within imagery. The Harris Corner method and the method of Phase Congruency can be used to effectively extract static and streaking point sources, and to indicate when apparent motion is present within an observation. The geometric inferences which can be made from the resulting detections will be discussed, including a method to evaluate the localization uncertainty of the extracted detections which is based on the computation of the Hessian of the detector response. Finally, a technique which exploits the additional information found in detected streak endpoints to provide a better centroid in the presence of curved star streaks is explained and additional applications for the presented techniques are discussed.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2014
Accession Number
ADA616753

Entities

People

  • Brad Sease
  • Brien Flewelling

Organizations

  • Air Force Research Laboratory

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes
  • Sensors

DTIC Thesaurus Topics

  • Acquisition
  • Air Force Research Laboratories
  • Algorithms
  • Artificial Satellites
  • Computations
  • Computer Vision
  • Computers
  • Detection
  • Detectors
  • Eigenvalues
  • Identification
  • Image Processing
  • Observation
  • Orbits
  • Space Objects
  • Space Surveillance
  • Uncertainty

Readers

  • Computer Vision.
  • Sensor Fusion and Tracking Systems.
  • Space Exploration and Orbital Mechanics.

Technology Areas

  • AI & ML
  • Space
  • Space - Space Objects