Study On The Classification Of Remotely Sensed Images Based On Ant Colony Optimization

The classification of remotely sensed images is an important mean of obtaining information.How to improve the accuracy of classification is an important content of remote sensing research. Adding features and researching new classification methods are the ways to improve accuracy of classification. Ant colony algorithm applied in remote sensing image classification is a new classification technology based on preliminary swarm intelligence. Studying the applicability of ant colony algorithm based on more features and exploring the advantages and performance of ant colony algorithm are provided with very important significance.This paper takes a case study on the land use classification of the outskirts of Fuzhou study area in Fujian province. We build the multi-source database which contains spectral, topography and textural characters.Classification rules based different characters are discovered from the samples through ant colony algorithm. Classification experiments are performed based on these rules.The traditional maximum likelihood method, C4.5 algorithm and rough sets classifications are also performed to check the accuracies. The results have suggested that the accuracy of classification based on more characters is higher than based on less characters through ant colony algorithm. And the accuracy of classification based on the ant colony algorithm is higher than other methods.In addition, the land use changes in Fuzhou during the last 9 years is studied by using remote sensing technology based on ant colony algorithm. The results have suggested that the land use changed dramatically during the last 9 years.We have analyzed the causes of changes and proposed some suggestions to the development of this region.