Biologically inspired solutions for modern-day sensory systems promise to deliver both higher capacity and faster, more efficient processing of information than current computational approaches. Many animals are able to perform remarkable sensing tasks despite only being able to process what would be considered modest data rates and bandwidths. The key biological innovations revolve around dedicated filter designs. By sacrificing some flexibility, specifically matched and hard-wired sensory systems, designed primarily for single roles, provide a blueprint for data and task-specific efficiency. In this paper, we examine several animal visual systems designed to use the polarization of light in spatial imaging. We investigate some implications for artificial optical processing based on models of polarization image processing in fiddler crabs, cuttlefish, octopus, and mantis shrimp.