The DARPA project, 'Develop and Demonstrate Real-Time Wavelet Based Automatic Target Recognition Using Sonar and Synthetic Aperture Radar (SAR) data' was initiated March 30,1998 with a kick-off meeting attended by personnel from Rice University CML, Northrop Grumman ESSD and DARPA DSO. The overall goal is the transition of DARPA sponsored university research into a fielded systems demonstration using a liaison between a university and an industrial partner familiar with the specific defense application. In the first year of this contract Northrop Grumman obtained and applied Rice CML's wavelet-based compression/expansion algorithms to sonar and SAR data collected for automatic target recognition. Also Northrop Grumman sent Rice CML sonar data collected for automatic target recognition. Also Northrop Grumman sent Rice CML's sonar data sets and references to MSTAR data sets for use in the development of wavelet- based detection techniques. In the second year Northrop Grumman retained its ATR algorithms for use on compressed data and also evaluated denoising and segmentation algorithms based on Hidden Markov Trees (HMT) defined on wavelet coefficients. These HMT technologies were developed by Rice CML during the first year of the contract. In the third year Northrop Grumman made ready a demonstration of wavelet based compression/expansion on Global Hawk SAR data running in real time using the Real Time ATR (RT-ATR) Lab at the DARPA technology Integration Center (TIC) in Arlington, VA. Also available for demonstration is the wavelet-based Rice developed denoising algorithms.
Develop and Demonstrate Real-Time Wavelet Based Automatic Target Recognition Using Sonar and Synthetic Aperture Radar (SAR) Data
2001
9 pages
Report
Keine Angabe
Englisch
Acoustic Detection , Radiofrequency Detection , Detection & Countermeasures , Target recognition , Synthetic aperture radar , Wavelet transforms , Computer programs , Data bases , Algorithms , Real time , Expansion , Data compression , Acoustic data , Sonar images , Synthetic aperture sonar , Denoising
Automatic Target Recognition using Multipolar Bistatic Synthetic Aperture Radar Images
Online Contents | 2010
|