Binary Tomography Reconstruction by Particle Aggregation

This paper presents a novel reconstruction algorithm for binary tomography based on the movement of particles. Particle Aggregate Reconstruction Technique (PART) supposes that pixel values are particles, and that the particles can diffuse through the image, sticking together in regions of uniform pixel value known as aggregates. The algorithm is tested on four phantoms of varying sizes and numbers of forward projections and compared to a random search algorithm and to SART, a standard algebraic reconstruction method. PART, in this small study, is shown to be capable of zero error reconstruction and compares favourably with SART and random search.