Plasmon-Enhanced Optoelectronic Graded Neurons for Dual-Waveband Image Fusion and Motion Perception

Area of Science:

• Materials Science
• Optoelectronics
• Artificial Intelligence

Background:

• Vision systems require synchronous temporal-spatial information processing across wide wavebands.
• Existing motion recognition systems face limitations in accuracy and spectral sensitivity.

Purpose of the Study:

• To develop dual-waveband sensitive optoelectronic synapses for high-accuracy motion recognition.
• To create a dynamic vision system capable of precise motion perception.

Main Methods:

• Fabrication of wedge-shaped nanostructures on molybdenum disulfide (MoS₂) monolayers for plasmon-enhanced wideband absorption.
• Utilizing optoelectronic graded neurons exhibiting photo-induced conductance plasticity at 633 nm and 980 nm wavelengths.
• Development of a 20 × 20 optoelectronic neuron dynamic vision system and neural network computing systems.

Main Results:

• Demonstrated remarkable photo-induced conductance plasticity in optoelectronic graded neurons at dual wavelengths.
• Achieved precise detection and perception of various motions using the dynamic vision system.
• Significantly enhanced motion trajectory recognition accuracy from <80% to >99% using dual-wavelength fused images.

Conclusions:

• Dual-waveband optoelectronic synapses offer a promising approach for advanced motion recognition and perception.
• The developed system significantly surpasses previous accuracy levels for identifying motion trajectories.
• This technology paves the way for more sophisticated artificial vision systems.