Super-Resolution Spike Event-based Polarimetric Dynamic Vision Sensor p(DVS) Cognitive Imaging

In this study, a novel experimental design of a cognitive imaging system aiming at enhancing the spatial and temporal resolution of neuromorphic vision sensors, is presented. Super resolution images of moving or semi-obstructed targets, by means of the “Polarimetric Dynamic Vision Sensor (pDVS)” system, were obtained. This system improves the contrast sensitivity and proves to be an efficient strategy for rapid scene analysis by making use of deep learning. The outcome of this study has a tremendous impact to contribute to the design of innovative bioinspired-based vision systems for rapid classification, identification, motion detection, and tracking.

I. Introduction

Integration of vision characteristics associated with certain marine invertebrate species and arthropods, with human cognitive vision would lead to the development of highly efficient autonomous navigation systems and robotic systems. [1]–[14]. However, integrating bioinspired vision systems with artificial electronics is a challenging task, especially due to size and power limitations.