A remote sensing images feature selection approach based on Ant Colony Algorithm

Some initial investigations are conducted to apply Ant Colony Algorithm (ACO) for feature selection of remote sensing images. As a novel branch of computational intelligence, ACO has strong capabilities of Self-organization adaptation; hence it is natural to view ACO as a powerful information processing and problem-solving method in both the scientific and engineering fields. Ant Colony Algorithm posses nonlinear classification properties along with the biological properties, being parallel operation and insensitiveness to initial condition of images. Preliminary Results indicate effectiveness and application of our method proposed and efficiently avoid some drawbacks of traditional methods. In addition our work also pushes research on this problem further.