You are here

Structure from Motion

Award Information
Agency: Department of Defense
Branch: Air Force
Contract: FA8651-04-C-0233
Agency Tracking Number: F041-170-1026
Amount: $99,944.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: AF04-170
Solicitation Number: 2004.1
Timeline
Solicitation Year: 2004
Award Year: 2004
Award Start Date (Proposal Award Date): 2004-06-15
Award End Date (Contract End Date): 2004-12-15
Small Business Information
11675 Jollyville Road, Suite 300
Austin, TX 78759
United States
DUNS: 838828960
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Thayne Coffman
 Principal Scientist
 (512) 342-0010
 tcoffman@21technologies.com
Business Contact
 Darrin Taylor
Title: Vice President
Phone: (512) 342-0010
Email: dtaylorz@21technologies.com
Research Institution
N/A
Abstract

State-of-the-art Structure From Motion (SFM) techniques are combined with novel methods to yield a real-time SFM capability for long calibrated image sequences. Unlike most approaches, the work is tailored to long continuous sequences of calibrated imagery generated from the near-planar small-baseline motions of a loitering aerial platform. This environment enables novel approaches for optimizing the processing order of the image frames, and for seeding calculations for optical flow, stereo correspondence, and camera parameter estimate refinement. It allows the 3D scene model to be built incrementally and transmitted to the ground incrementally, reducing network bandwidth requirements. Combined, these approaches reduce the computational requirements and error in the system to a point that accurate real-time operation on a loitering air platform is feasible. This capability will improve automatic target recognition, situational awareness in the combat environment, operations against aboveground and underground structures, and battlefield damage assessment.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government