Biovision-Inspired Perovskite Intelligent Camera for Panchromatic and Metameric Sensing

Area of Science:

• Optics and Photonics
• Biomimetic Engineering
• Materials Science

Background:

• Biological vision systems face trade-offs between information acquisition and processing.
• Mantis shrimp have advanced spectral sensing but limited neural processing for color perception.

Purpose of the Study:

• To develop a multispectral imaging system inspired by mantis shrimp vision.
• To integrate spectral detection with advanced visual processing for enhanced color recognition.
• To overcome limitations of current imaging systems, such as metamerism.

Main Methods:

• Fabrication of perovskite narrowband photodetectors using aerosol-liquid-solid spraying.
• Development of a computational multispectral imaging system capturing seven spectral images.
• Design of a deep-learning-based color fusion network for RGB representation.

Main Results:

• High-quality perovskite photodetectors covering ultraviolet to visible spectrum were created.
• The system achieved single-shot, seven-spectral-image capture.
• The deep-learning network effectively translated multispectral data into enhanced RGB color recognition.

Conclusions:

• The developed perovskite intelligent camera mimics biological vision strengths.
• This novel approach advances multispectral imaging capabilities.
• Potential applications include machine vision, remote sensing, and medical imaging.