Improvement of Resolution in Acoustical - Holographic Imaging by Neural Network Processing

Abstract

Recently neural network techniques have been used in various field including signal processing, where this technique is applied to the spectrum estimation, direction estimation of the radio wave and so on. In this paper we propose a neural network processing to improve resolution of the reconstructed images from holograms of long-wavelength, such as acoustical and microwave holograms. Since the wavelengthes are long in such waves, improvement of resolution is necessary, because the apertures of hologram planes are narrow compared with the optical holography, resulting in the low-quality reconstructed image.