Image Data Rate Reduction for an On-Board Real-Time SAR-Processor

Inhalt:
For on-board usage in UAVs a real-time SAR processor has been developed at Laboratorium für Informationstechnologie, Universität Hannover in cooperation with EADS Deutschland GmbH, Friedrichshafen. It consists of six processing nodes providing high processing power for image generation and applying data rate compression techniques on generated images. Two data rate reduction techniques have been investigated, image compression and ROI selection. Image data are compressed by a block based image coding scheme, which reduces data rate by factor of about ten at acceptable image quality for analysis. It is based on JPEG techniques which have been adapted to characteristics of generated SAR image data. An additional data rate reduction can be achieved by restricting processing of raw data to interesting regions (ROI). The resulting compression factor is dependent on size and number of ROIs defined in the scene.