and 
No. 
lote 
ing 
SR- 
ure 
  
THE ARTS-IP RESEARCH PROGRAMME: 
A EUROPEAN LONG-TERM APPROACH TO REMOTE SENSING 
E. CHIARUCCI, A. FARINA and-F. MARCOZ 
ARTS-IP WORKING GROUP, presently with 
SELENIA S.p.A. - Via Tiburtina Km. 12.400 - I-00131 ROME 
SUMMARY 
This paper describes a cooperative research project, named 
"Adaptive Real Time Strategies for Image Processing: A case 
study for satellite data" (ARTS-IP), sponsored by the European 
"Esprit" initiative, in the area of Advanced Information Pro- 
cessing. The aim of the ARTS-IP project is to explore  innova- 
tive procedures and computer architectures to process, on à 
real-time basis, the extremely large amount of informations 
provided by satellite borne SARs for the remote sensing of the 
earth. 
The research programme, involving 150 man/year for R & D 
jn a five year period,.is in the first vear of feasibility 
study. The paper illustrates the rationale of the system con- 
cept being investigated, the objectives of the research and 
advantages over known techniques. Additionally, it describes 
the organization of the research, the activities undertaken 
and the plan of future work. 
1. . INTRODUCTION 
Roughly speaking, the remote sensing architecture to be 
investigated is built around a so called "dual look" synthetic 
aperture radar (SAR) system concept. More precisely, the SAR 
in the "first look" mode gathers and processes data from a 
swath with a limited information content capacity (and thus 
limited processing), such as low resolution. The coarse image 
perceived is then analyzed to ascertain whether areas of -inte- 
rest are present, to be further acquired and processed by the 
SAR with a "second look" modality, yielding detailed  informa- 
tions at a much higher resolution. 
In the framework of the above concept the following  advan- 
tages over conventional remote sensing techniques are worth 
mentioning: (i) the multimode operation; (ii) the adaptive 
capability of the system to manage efficiently system resour- 
ces such as radar transmitted power and on-board computer re- 
sources where and when they are needed; (iii) the opportunity 
of sharing the processing load between on-board and ground 
facilities, and (iv) the identification on a real-time basis 
(the first decision being made on-board with a short reaction 
time) of few and sparse areas in a uniform background (e.g. 
forest disease, oil pollution and spills in the ocean, etc.). 
The correct operation of such an image (data). driven pro- 
cessing scheme relies on the implementation of appropriate 
switching procedures between the first and second look SAR 
modes. The ARTS-IP project will explore and test. such" 'pro- 
345