Image reconstruction using the reconfiguration technique

摘要

Image reconstruction has been under study for decades and plays an important role in science, technology and security. In this paper, we propose an identification method for coded images. This method is different from conventional object identification frameworks, not in the sense of the extracted geometrical features but in the sense of the reconfiguration concept used. The goal of reconfiguration in this study is to reach a point that will make object identification possible. There have been research projects previously conducted on this topic and most of them are discretization-based. Our method, based on the reconfiguration concept, is a technology-based method that allows object identification. Our experimentation shows that the object can be identified within 8 iterations of the algorithm, beyond which no additional accuracy is achieved. This identification method, based on reconfiguration, extends the existing ones and enhances state-of-the-art object identification methods.