Pixel sensor integrated neuromorphic VLSI system for real-time applications

A pixel sensor integrated VLSI system, based on a bio-inspired neuromorphic model, is reported. The design employs photodiode sensors where the analog input states in related 2D cell array are programmed directly with visual patterns. The system functionality and specific image processing tasks are demonstrated with measurements, where real-time input patterns are introduced. The static power consumption estimations for the array are given, as larger arrays are feasible due to compact design. The estimated and measured numerical results are compared. The obtained convenient sensor integration can lead to real-time high-resolution applications of this system.