The Butterfly Effect in Primary Visual Cortex

Exploring and establishing artificial neural networks with electrophysiological characteristics high computational efficiency is a popular topic in the field of computer vision. Inspired by working mechanism primary visual cortex, pulse-coupled network (PCNN) can exhibit synchronous oscillation, refractory period, exponential decay. However, evidence shows that neurons highly complex non-linear dynamics when stimulated external periodic signals. This chaos phenomenon, also known as " butterfly effect", cannot be explained all PCNN models. In this work, we analyze main obstacle preventing models from imitating real cortex. We consider neuronal excitation stochastic process. then propose novel network, called continuous-coupled (CCNN). Theoretical analysis indicates dynamic behavior CCNN distinct PCNN. Numerical results show model exhibits under DC stimulus, chaotic AC which consistent neurons. Furthermore, image video processing mechanisms are analyzed. Experimental on segmentation indicate has better performance than state-of-the-art cortex