Integration of Bioinspired Vision Principles Towards the Design of Autonomous Guidance, Navigation, and Control Systems

In this study, key-characteristics of different bioinspired vision systems are presented, and their importance in the design of autonomous navigation systems is highlighted. Two paradigms reflecting authors' research efforts in the design of vision systems, mimicking the human and animal retina, are presented. The first paradigm relates to the development of novel polymer nanostructures and nanoantennas aimed at integrating polarization principles with CMOS designs, mimicking bioinspired retina systems. The second paradigm relates to the development and testing of a novel neuromorphic spike event-based polarimetric Dynamic Vision Sensor (DVS), for dynamic imaging of a scene, with enhanced temporal contrast and high object discrimination. The outcome of this study indicates that integration of several bioinspired vision principles, such as polarization of light and neuromorphic detection, play an important role in the design of innovative bioinspired nanomaterials and neuromorphic vision architectures.

I. Introduction

Arthropods, certain marine invertebrate species, and human cognitive vision systems exhibit unique vision capabilities, complementary to each other; if selectively combined it would lead to the development of highly efficient robotic, autonomous driving systems with enhanced vision characteristics [1]–[10]. These superb vision capabilities stem from the geometrical, structural, functional and biochemical parameters of the biological retina, the decentralized sensomotor system design of the arthropods, and the cognition-assisted human vision paired with efficient neural network processing engines, both at retina and brain.