Pigeon-inspired optimization approach to information granulation-based fuzzy RBF neural networks for image fusion

Nowadays, bio-mimetic method comes more and more popular as being applied in various fields including image fusion. A number of population-based algorithms are proposed to solve this kind of problem for example the ant colony optimization and the artificial bee colony optimization. In this paper, an original approach with bionic solutions named Pigeon-Inspired Optimization (PIO) is proposed, which possesses high efficiency especially for discrete optimization problem. It is combined with the fuzzy radial basis function neural network to optimize the problem of image fusion. A population-based searching process is provided to the optimization model, which is inspired by the movements of pigeons that can accurately guide a group of pigeons to their final destination. A series of experiments are set up in this paper to verify the feasibility and effectiveness of the method.