SARAI

SARAI – SAR signature simulation of targets

The objective of the project is to develop a simulated operating environment IMINT (Imagery Intelligence). This will be done through the formation of a multilevel deep learning network
in order to “understand” the SAR images and time series by learning the functionalities without technical supervision and to carry out specific detection, classification and / or recognition
tasks by specializing the technologies developed for the identified scenario. The objective of the activity is to support the execution of the tasks planned for Phase 1 of the SARAI project.

In detail, the activities are divided into 5 tasks:

• Development, calibration and test of a SAR signature simulator.
• Identification, selection and collection of 3D geometric models of the targets of interest both open source and from commercial suppliers.
• Management, adaptation and conversion of 3D geometric models as input from the SAR simulator.
• Generation of SAR signatures.
• Contribution to the project documentation relating to the SAR signature simulator.