Spatial resolution of hard-field tomography systems with severely limited data

Stylianos-Alexios Tsekenis, Nataša Terzija, Paul Wright and Hugh McCann

School of Electrical & Electronic Engineering, University of Manchester, UK

ABSTRACT

The term ‘spatial resolution’ frequently appears in scientific literature from numerous research areas, where it is used to characterise the performance of an imaging system or a tomographic modality. In this paper we will first investigate the current definitions and understanding of the term in different research areas and provide the reader with an interdisciplinary definition. Subsequently we will focus on the importance of spatial resolution in the case of an in-cylinder near-infra red (Near-IR) absorption chemical species tomography (CST) system to visualise the distribution of hydrocarbons in a compression ignition test engine. The functional stages of tomographic imaging are presented and examples of factors that contribute to the reduction of spatial resolution are given from the medical and industrial imaging research community. An experimental method is described in detail to consistently measure the spatial resolution of the Near-IR absorption CST system that uses a sparse and irregular beam array. The method is based on tomographic reconstruction of an edge feature using a novel angular translation technique and the division of the reconstructed images into a matrix of sectors. Analysis of the collected data for each sector proceeds with the plotting of the edge spread function (ESF), followed by the line spread function (LSF) and finally the 1-dimensional modulation transfer function (MTF). An amplitude threshold is set and the highest occurring frequency is translated back into the spatial domain to yield a spatial resolution figure for each sector. We believe that the proposed experimental approach can be adapted and applied to other systems that are characterised by a non- isotropic spatial resolution and possibly different tomographic modality.

Keywords Spatial resolution, hard-field, tomography, sparse array, combustion engine