FPGA Based Satellite Adaptive Image Compression System

Space-borne imaging sensors onboard remote sensing satellites capture images of specific interest to the mission. These images contain tremendous volumes of data. Storage capacity onboard the satellite is limited and downlink channels have limited bandwidth capacity. Moreover, most remote sensing satellites are visible to their ground stations for only short periods of time during the day. Therefore, there is significant delay between image capture and terrestrial analysis. Hence, onboard image compression is essential to deliver captured data to the ground in a timely fashion. Smart and efficient onboard image compression using reconfigurable computing technology such as field programmable gate arrays (FPGAs) represents a unique solution that significantly increases the mission effectiveness. The High Performance Computing (HPC-I) payload on the Australian scientific mission satellite FedSat is a demonstration of reconfigurable computing technology in space for a variety of applications including image compression. In this paper, we present the design and implementation on HPC-I of the Satellite Adaptive Image Compression System (SAICS) for space missions. The system concept is based on an FPGA implementation of the adaptive JPEG-LS algorithm. The SAICS is a low complexity, high speed adaptive compression system that uses lossless and near lossless techniques depending on local image statistics. Performance evaluation indicates that satisfactory compression results can only be achieved through a careful trade-off between design complexity and efficiency.