| The algorithm of image fusion based on multiscale geometric analysis and particle swarm optimization(PSO) is mainly discussed in the paper.Firstly, the concept of multiscale geometric analysis is analyzed in theory and wavelet transform, curvelet transform, contourlet transform and NSCT are analyzed in detail. Then the theoretical basis of PSO algorithm is introduced, and the formulas of the standard PSO are given. Secondly, the basic levels of image fusion and evaluating methods for quality of fused images are introduced. Finally, the algorithm of fusion of multi-focus images, medical images and remote sensing image based on multiscale geometric analysis and particle swarm optimization is proposed. This method takes multiple fused images received from fusion algorithm based on multiscale transform as initial particle swarm and takes two evaluation indicators as objectives for optimization.When different types of source images are processed for optimization with the fusion method the paper proposes, according to different imaging mechanism and fusion purposes, different evaluation indicators are selected as the objective functions of the multi-objective optimization. In the case of multi-focus image, average gradient and information entropy are selected as the objective; in the case of medicine image, average gradient and root mean square error are selected as the objective; in the case of remote sensing image, spectral correlation coefficient and average gradient are selected as the objective. Considering the practical application of PSO in image fusion, a new formula of inertia weight and the reasonable selection of optimal solution are designed. Results of different input are studied in image fusion of multi-focus images.Finally, based on the fusion method the paper proposed, experiments of three types of images are done; Simulation results and quantitative evaluating parameters consistently show the algorithm has better fusion performance. |
PDF Full Text Request