A Biomimetic Self-Adaptive Neurovision Eye With an Integrated Gel Iris and Retinamorphic Architecture

Area of Science:

• Neuromorphic Engineering
• Biomimetic Systems
• Materials Science

Background:

• Biological vision systems exhibit remarkable self-adaptive regulation for stable perception.
• Artificial systems require integrated light control and in-sensor processing for similar functionality.

Purpose of the Study:

• To develop a biomimetic self-adaptive neurovision system.
• To integrate dynamic light control with advanced visual processing capabilities.

Main Methods:

• A novel system combining a photothermally responsive hydrogel iris with a hemispherical ionogel retina.
• The hydrogel iris modulates opacity (2.95-89.85%) via a thermochromic effect.
• The ionogel retina features a micropillar array with a photosynaptic ionotronic heterostructure for broadband photoperception (360-680 nm) and learning.

Main Results:

• The system demonstrated autonomous opacity modulation and broadband light perception.
• Achieved visual learning capabilities, including paired-pulse facilitation (118%).
• Maintained high-fidelity image recognition (97.2% accuracy) even under intense light exposure.

Conclusions:

• The synergistic integration of adaptive optics and sensory processing in a neuromorphic system is feasible.
• This technology offers a pathway towards zero-powered soft visual prosthetics and enhanced machine vision applications.