A PSO based approach to image reconstruction from projections

Abstract

Image reconstruction from projection is used to find out the density distribution within the physical object, which is NP-Hard problem if the number of projections for reconstruction is greater than two. Exact solution of this type of the problem is not possible. Evolutionary techniques can be used for such type of problems to find out some approximate solutions. Particle Swarm Optimization (PSO), a relatively recent category of stochastic global optimization technique, is used here to find out the approximation solution in polynomial time. PSO is easy to implement as compare to the other evolutionary technique such as genetic algorithm. © 2011 by IJTS (CESER Publications).