A Spatiotemporal Helix Approach to Geospatial Exploitation of Motion Imagery

Abstract

The transition from static to motion imagery enabled by UAVs, and distributed sensor networks introduces significant challenges monitoring multiple video streams, indexing large amounts of video, and retrieving video segments with significant spatiotemporal events. Current analyst workflow systems support spatiotemporal analysis minimally: analysts roam through large image mosaics, revisit the same area, and make use of accompanying geographic information systems (GIS) databases to examine the evolution of a scene. However, the analyst still needs to decide whether the trajectory of a vehicle appears suspicious, whether a moving convoy resembles a formation of military vehicles, or whether the trajectory of a truck today resembles yesterday s pattern of movement of another vehicle. As data influx is increasing at substantially high rates, agencies are trying hard to keep up with tremendous amounts of incoming data, further exasperated during crises (e.g. military operations), where the rate of incoming information explode (e.g. as areas of interest are under heavy surveillance by fleets of UAVs).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2004
Accession Number
ADA432580

Entities

People

  • Alper Caglayan
  • Anthony Stefanidis
  • Peggy Agouris

Tags

Communities of Interest

  • Air Platforms
  • Autonomy

DTIC Thesaurus Topics

  • Abstracts
  • Contracts
  • Detectors
  • Engineering
  • Geographic Information Systems
  • Information Operations
  • Information Systems
  • Military Operations
  • Military Vehicles
  • Monitoring
  • Networks
  • Sensor Networks
  • Trajectories
  • Transitions
  • Vehicles
  • Wireless Sensor Networks

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Geospatial Intelligence and Artificial Intelligence Analytics
  • Systems Analysis and Design