Classification of Remote Sensing Images Based on Ant Colony Optimization
LIU Xiao-ping,LI Xia,HE Jin-qiang,AI Bin,PENG Xiao-juan
DOI: https://doi.org/10.11834/jrs.20080233
2008-01-01
Journal of Remote Sensing
Abstract:This paper presents a bottom-up approach to improve the classification performance for remote sensing applications.Top-down approaches,such as statistical classifiers,have inherited limitations in dealing with complicated relationships in classification.For example,data correlation between bands of remote sensing imagery has caused problems in generating satisfactory classification with statistical methods.In this paper,ant colony optimization(ACO) based on swarm intelligence is used to improve classification performance.Actually,ACO is a complex multi-agent system,in which agents with simple intelligence can complete complex tasks through cooperation such as classification problems.Ants guide their route selection based on pheromone,which is accumulated from the collective movements of individual ants.In this way,an ant learns from the past experience of others.Ant-Miner is different from decision tree approaches.The entropy measure is a local heuristic measure,which considers only one attribute at a time,and so it is sensitive to attribute correlation problems.Whereas in Ant-Miner,pheromone updating tends to cope better with attribute correlation,since pheromone updating is directly based on the performance of the rule as a whole.Thus,Ant-Miner should have great potential in improving remote sensing classification.In this study,an Ant-Miner program for discovering classification rules is developed for the classification of remote sensing images.In the Ant-Miner program,the route search by an ant colony is to find the best links between attribute nodes and class nodes.An attribute node corresponds to a band value of remote sensing images.An attribute node can only be selected once and must be associated with a class node.Each route corresponds to a classification rule,and discovering a classification rule can be regarded as searching for an optimal route.To enable ACO to effectively classify remote sensing imagery of very large data sets,original band values are sliced into a number of intervals by using a discretization technique.The ACO method is more explicit and comprehensible than mathematical equations.Our study in Guangzhou city indicates that the ant colony-based classifier yields better accuracy than conventional maximum likelihood classifiers and decision tree classifiers.The overall accuracy of the ACO method is 88.6%,with a Kappa coefficient of 0.861.The decision tree method has an accuracy of 85.4% and a Kappa coefficient of 0.822.The maximum likelihood method has an accuracy of 83.3% and a Kappa coefficient of 0.796.The results clearly support the conclusion that the method explored in this paper can be more effective than conventional classification methods.