HELM: Hologram Examination for Life-like Motility(NPO-50634-1)

We seek to develop algorithms to detect moving objects directly in the holograms produced by a Digital Holographic Microscope (DHM), eliminating the majority of computation required for spatial reconstruction. We further will construct a functional prototype system that tracks these particles, analyzes their motion, and trains a machine learning classifier to distinguish life-like motility from non-living organic or inorganic particles. We will demonstrate sufficient speed (limited compute) and memory efficiency (limited memory footprint) to enable laptop-based, in-field, real-time processing of the streaming instrument output leading directly to proposals for an on-board DHM instrument on Europa or Mars.