Comparative analysis of parallel OPIR compression on space processors

Abstract

Requirements for higher video quality in space applications continuously calls for increased resolution in imaging sensors, higher bit-depth codecs, more creative solutions for preprocessing and compression techniques, and faster, yet resilient, space-grade platforms. Understanding how these variables interact and affect each other on different platforms is crucial in system development when trying to meet requirements and constraints, such as compression speed, compression ratio (CR), image quality, bandwidth, etc. To analyze this interaction, we present a comparative analysis between compression speed and compression ratio using serial and parallel compression codes on different platforms and architectures, focusing upon video data from overhead-persistent infrared (OPIR) sensors on spacecraft. Previous research allowed us to compare CR and image quality with new preprocessing techniques, but it did not evaluate and address the challenges of compression speed on space-grade processors. Performance is critical, since of course the preprocessing and compression codes plus downlink of compressed data must require less total time than downlink of the raw data, in order for compression to be fully effective.