On-Board Characterization Of Hyperspectral Image Exposure And Cloud Coverage By Compression Ratio

Hyperspectral images are useful for remote sensing applications due to the fine spectral resolution relative to regular RGB- and multi-spectral images. The images can be used to study what is observed, rather than just that a given feature is observed. The high spectral resolution comes with a challenge for both small and big satellites: the data volume per observation. For a small earth observation satellite, the time and energy to downlink data is one of the main driving factors behind data latency and imaging capacity.Effective use of satellite time calls for a smart imaging pipeline, both when it comes to planning, execution, and processing of data from observations. The imaging and processing pipeline needs to utilize both on-board and on-ground processing tools and steps. In this paper, a set of the first observations from the HYPSO-1 satellite is analyzed to evaluate how quickly and effectively the compression ratio can be used to identify observations that should be prioritized for downlinking, and which data sets seem to be of less value due to over/under exposure or cloud coverage.