A hybrid tomographic reconstruction algorithm for high speed X-ray tomography

Abstract

This paper compares two techniques for reconstructing images from time-resolved X-ray tomography data: Simultaneous Algebraic Reconstruction Technique (SART) and Adaptive Genetic Algorithm (AGA). The advantages and disadvantages of both are discussed, and we combine them into a hybrid algorithm. The hybrid method uses results from AGA as the starting values for SART. We demonstrate that this leads to improved reconstruction results: less noise than SART and better reconstruction of shapes than AGA.