Wire Detection in mmW Radar Using Convolutional Neural Networks

Award Information
Agency: Department of Defense
Branch: Army
Contract: W56KGU-17-C-0034
Agency Tracking Number: A2-6559
Amount: $491,069.36
Phase: Phase II
Program: SBIR
Solicitation Topic Code: A152-093
Solicitation Number: 2015.2
Solicitation Year: 2015
Award Year: 2017
Award Start Date (Proposal Award Date): 2017-09-23
Award End Date (Contract End Date): 2018-09-22
Small Business Information
1495 Chain Bridge Rd, McLean, VA, 22101
DUNS: 966276920
HUBZone Owned: N
Woman Owned: N
Socially and Economically Disadvantaged: N
Principal Investigator
 Kenneth D Morton, Ph.D.
 Chief Scientist
 (919) 439-6362
Business Contact
 Mark Hibbard
Phone: (703) 872-7280
Email: mark@covar.com
Research Institution
This is a proposal for Phase II of SBIR topic A152-093. During Phase I, CoVar Applied Technologies, Inc., developed novel methods for the detection of power-lines in mmW radar data by leveraging deep learning approaches and Bayesian tracking. CoVars novel algorithms were demonstrated to outperform state of the art algorithms from the relevant research literature. These algorithms will be enhanced in the Phase II effort, expanding the conditions under which they can reliably perform, and the types of wires and powerlines that they are capable of detecting. The algorithms and their modifications can be easily transferred to additional mmW radar sensors not considered in Phase I. Furthermore, during Phase II, the mmW algorithm(s) will be fused with other sensors systems, such as IR imaging and LIDAR, to enable tracking of detected powerlines prior to, and during, degraded visual conditions. The resulting detection and tracking algorithms will be implemented in a real-time architecture and integrated with visualization systems to be presented in the cockpit. These new capabilities will improve the tactical capabilities of rotorcraft and help save lives lost during takeoff and landing operations.

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

US Flag An Official Website of the United States Government