Multi-spectral remote sensing images classification method based on adaptive immune clonal selection culture algorithm

Abstract

In immune clonal selection algorithm for remote sensing images classification problem, only clonal selection mechanism is adopted. It makes the exploitation and exploration of the algorithm limited. To solve above problem, adaptive immune clonal selection culture algorithm is introduced in the paper. It fully uses the dual evolution mechanism of culture algorithm to extract implicit knowledge in belief space. According to the evolution situation noted in topological knowledge, a hybrid selection strategy integrating clonal selection and (μ+λ) selection are proposed in population space. Simulation results indicate that the classification method based on adaptive immune clonal selection culture algorithm can improve the classifier performance better. © 2011 Springer-Verlag.