Image reconstruction of an image sensor based optical tomography

Abstract

Process tomography is a technique that has the potential for enabling great improvements in efficiency and safety in process industries, while minimizing waste and pollution in a range of applications. It provides several methods of obtaining the concentration profile of a process. It has become one of the vast growing technologies nowadays and it can be applied to many types of processes and unit operation. The main target of this project is to design a tomography system which is able to reconstruct an image from four-projection of light beams through the CMOS linear image sensors. The accuracy of the image obtained dependants on the number of sensors used, the projection technique applied and the number of projections used. For this project, the image reconstruction is based on Linear Back Projection (LBP) algorithm. The objective of this project is to investigate the process tomography, and how can it be used to construct an image inside the pipe. The characteristics of the light source and sensor are investigated in order to choose the appropriate sensor. Image reconstruction is done using MATLAB.