Automatic Target Recognition (ATR) Technology Components

Period of Performance: 02/27/1998 - 02/27/2000


Phase 2 SBIR

Recipient Firm

Toyon Research Corp.
6800 Cortona Drive Array
Goleta, CA 93117
Principal Investigator

Research Topics


We propose to develop a model-based Automatic Target Recognition (ATR) system suitable for use in a laboratory research environment. The ATR is intended to identify moving ground targets based on returns from an airborne high range resolution moving target indicator (HRRMTI) radar. The ATR will perform the necessary processing of raw radar data, segment targets from the data, and scrore target signatures (range profiles) against synthetically generated template profiles to produce target classifications. The ATR will incorporate numerical optimization to efficiently search through a template database to find the best match for a given measured range profile. Other features of the ATR may include bulk filtering of targets based on range profile length estimation and fusion of multiple HRRMTI measurements if they are found to significantly increase ATR discrimination performance. The ATR will be tested and evaluated using a number of different radar data sets, as they become available. System performance will be quantified by measures such as probability of making a correct target classification, probablility of making a false target classification, and speed with which a target classification is made.