SBIR/STTR Research Award
ANALYSIS OF ATR PROBLEM COMPLEXITY AND SCALABILITY
Award amount
$99,994Program and phase
STTR · Phase IAward year
1998Award details
- Awarding agency
- Department of Defense
- Branch
- Air Force
- Recipient company
- Alaphatech, Inc.
- Proposal award date
- Not reported
- Contract end date
- Not reported
- UEI
- Not reported
- Solicitation number
- Not reported
- Topic code
- Not reported
Research abstract
An important problem in the analysis of ATR systems is quantifying the growth of algorithm requirements, such as storage and processing cost, as the input problem l situation in classical complexity analysis, where the notion of input size is well-defined, there does not currenlty exist a clear notion of input size for the ATR problem. In this effort, we develop a measure of ATR problem size that is akin to e provides a numerical assessment of ATR problem complexity that logically orders ATR problems by intrinsic difficulty and is tractable to compute. We then establish the utility of the measure for estimating an ATR's ability to scale to larger mission probelms for which it has not been trained. Finally, as a byproduct of the effort, we provide approaches to constructing recognition algorithms which are in a certain sense minimal.
Technology keywords
Not reported