Correction and Reconstruction of Experimental Data Using the Step-by-Step and Combined Shift Methods

Abstract

The numerical step-by-step and combined shift methods for correction and reconstruction of experimental data presented as a set of superposed pulse responses of the measuring system are considered. These methods are effective for different types of superposition functions and make it possible not only to correct the data with various distortions in the course of measurements, but also to reconstruct the system response from a shorter initiating pulse, which means an improvement in the resolution of the measuring system. Comparative analysis of the proposed methods is presented and the inaccuracies of reconstruction with different types of superposition function, as well as different volumes of experimental data and noise levels in them. The possibility is considered of using these methods for renewal of multidimensional images or objects in them smudged by shifts or distorted by the Gaussian function. Examples of experimental-data reconstruction and image renewal are presented.