Hardware‐Oriented Visual Information Edge Detection, Based on Proton Migration‐Driven Lateral Inhibition

Abstract Lateral inhibition is a phenomenon observed in retinal neurons that enhances edge perception and increases the contrast of visual images. Here, a novel proton migration‐driven multi‐channel device is developed, which is designed for edge detection in real images by employing the principle of lateral inhibition. The device comprises 24 indium zinc oxide (IZO) channels interconnected via proton migration within a common Nafion electrolyte, enabling high interactivity among the IZO channels. The conductance of the respective channels is determined by the proton concentration in the channels as a result of intercalation and de‐intercalation of protons with voltage pulse applications. The device clearly mimics various retinal functions, such as center‐surround antagonism and optical illusions such as the Mach band and Hermann grid, in which perceived contrast is enhanced at the edges between dark and bright areas. Based on these characteristics, the multi‐channel device is used to extract the edge information from real images by utilizing the intrinsic properties of the ionic materials. These results suggest that proton‐based highly interactive devices have the potential in use for applications related to the development of artificial hardware‐oriented vision systems.