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.