Title of Paper: Real-Time Data Processing Onboard Remote Sensor Platforms

Principal Author: Sid Cook

Organization: Lockheed Martin Space Systems Astronautics

Abstract: With EO-1 Hyperion and MightySat in orbit NASA and the DoD are showing their continued commitment to hyperspectral imaging (HSI). As HSI sensor technology continues to mature, the ever-increasing amounts of sensor data generated will result in a need for more costly communication and data handling systems. Lockheed Martin with experience in spacecraft design and expertise in developing special purpose onboard processors has teamed with Applied Signal & Image Technology (ASIT) who has an extensive heritage in HSI to develop a real-time and intelligent onboard processing (OBP) system that reduces HSI sensor downlink requirements. Our goal is to reduce the downlink requirement by a factor > 100, while retaining the necessary spectral fidelity of the sensor data needed to satisfy the many science, military, and intelligence goals of these systems. Our initial spectral compression experiments leverage commercial-off-the-shelf (COTS) spectral exploitation algorithms for segmentation, material identification and spectral compression that ASIT has developed. ASIT will also support the modification and integration of this COTS software into the OBP. Other commercially available COTS software for spatial compression will also be employed as part of the overall compression processing sequence. Demonstrations of our initial capability will be presented at the August meeting. Over the next two years elements of a high-performance re-configurable OBP will be brass-boarded to implement proven preprocessing steps that distill the HSI data stream in both spectral and spatial dimensions. The system will intelligently reduce the volume of data that must be stored, transmitted to the ground, and processed while minimizing the loss of information.