A novel quantum behaved Particle Swarm optimization algorithm with chaotic search for image alignment

In an attempt to improve existing evolutionary metaheuristics quantum computing principles have been used. While some of them focus on the representation scheme adopted others deal with the behavior of the underlying algorithm. In this paper, we propose a search strategy that combines the ideas of use of a chaotic search with a selection operation within a quantum behaved Particle Swarm optimization algorithm. This search strategy is developed in order to achieve image alignment through maximization of an entropic measure: mutual information. The proposed framework is general as it handles any kind of transformation. Experimental results show the effectiveness of the algorithm to achieve good quality alignment for both mono modality and multimodality images. The proposed combination of the two features has lead to better solutions compared to those obtained by using each feature alone.