CMOS-integrated organic neuromorphic imagers for high-resolution dual-modal imaging

Area of Science:

• Artificial Vision Systems
• Neuromorphic Engineering
• Organic Electronics

Background:

• Current artificial visual systems face challenges in integrating static and dynamic information processing.
• Multimodal operation in visual sensors is crucial for comprehensive scene understanding.

Purpose of the Study:

• To develop a CMOS-integrated organic neuromorphic imager with dual imaging modes.
• To enable efficient spatiotemporal information extraction and in-sensor memory functionality.

Main Methods:

• Fabrication of a 640×512-pixel CMOS-integrated organic neuromorphic imager.
• Implementation of standard (frame-based) and synaptic (neuromorphic) imaging modes.
• Development of a CMOS-compatible photolithography method for device patterning.

Main Results:

• The neuromorphic imager successfully extracts high-resolution spatiotemporal maps (light distribution, motion trajectories).
• Adjustable multi-level memory behavior was achieved through modulation of charge recombination-trapping dynamics.
• In-sensor memorization exceeding 18 minutes and real-world spatiotemporal imaging were demonstrated.

Conclusions:

• The developed organic neuromorphic imager offers efficient spatiotemporal data acquisition and in-sensor processing.
• The device shows significant potential for applications in industrial monitoring and motion detection.
• The CMOS-compatible fabrication method supports high-resolution, non-destructive patterning of organic neuromorphic devices.