SBIR-STTR Award

Autonomous Target Recognition Using Neural Networks
Award last edited on: 9/4/2002

Sponsored Program
SBIR
Awarding Agency
DOD : Navy
Total Award Amount
$549,981
Award Phase
2
Solicitation Topic Code
N87-251
Principal Investigator
Steven Engle

Company Information

Expert-EASE Systems Inc

1301 Shoreway Road Suite 420
Belmont, CA 94002
   (415) 593-3200
   N/A
   N/A
Location: Single
Congr. District: 14
County: San Mateo

Phase I

Contract Number: N00163-88-C-0024
Start Date: 1/5/1988    Completed: 00/00/00
Phase I year
1988
Phase I Amount
$49,981
An autonomous target recognition system, incorporating neural nets is proposed. A prototype will be developed and evaluated in a variety of circumstances to assess the systems reliability, tolerance to signal noise, and ability to identify objects given only orthogonal views of the training objects. This approach offers a number of advantages over current machine vision technologies and is generalizable to a wide variety of passive sensor systems. It is also implementable in VLSI circuits or optical computers.

Phase II

Contract Number: N00163-88-C-0024
Start Date: 1/5/1988    Completed: 00/00/00
Phase II year
1989
Phase II Amount
$500,000
The Phase I research has demonstrated the capabilities of neural networks in vision related applications, and the technology has considerable promise in a variety of other application areas as well. However, there is still a considerable amount of research to be performed before the full potential of applications is understood. In the light of this situation, the Phase II project will focus on the further development of the volts system, expanding its capabilities as a tool for use in general neural net re search by implementing the Portable Neural Environment (PNE). The pne will include support for full portability across a variety of computer architectures, including parallel processors, and will form the basis for a high-performance, cost-effective neural net workstation. The system will then be used to implement a full capability neural net-based navigation system using aerial imagery.