ACO algorithm processing multisensor data for urban land cover

Abstract

A novel ant colony optimization (ACO) algorithm takes inspiration from the coordinated behavior of ant swarms finding the shortest way from their nests and the food source, which has been applied on many research areas for solving optimization problems, but it has seldom been used in remote sensing data processing. ACO algorithm has many potential advantages in remote sensing data processing, such as it does not assume an implicit assumption for processing dataset, it can take into account of contextual information, it has strong robustness, and it can combine different sources of data. This paper represents an application of the combination of Landsat TM data and Envisat ASAR data based on ACO algorithm for land cover classification. The classification results based on ACO algorithm were compared with MLC and C4.5, the experimentation results and analysis indicate that the ACO algorithm can provide a new efficient approach for land cover classification using multi-source of remote sensing data.