An Unsupervised Classification Method of Remote Sensing Images Based on Ant Colony Optimization Algorithm
Remote sensing images classification method can be divided into supervised classification and unsupervised classification according to whether there is prior knowledge. Supervised classification is a machine learning procedure for deducing a function from training data; unsupervised classification is a kind of classification which no training sample is available and subdivision of the feature space is achieved by identifying natural groupings present in the images values. As a branch of swarm intelligence, ant colony optimization algorithm has self-organization, adaptation, positive feedback and other intelligent advantages. In this paper, ant colony optimization algorithm is tentatively introduced into unsupervised classification of remote sensing images. A series of experiments are performed with remote sensing data. Comparing with the K-mean and the ISODATA clustering algorithm, the experiment result proves that artificial ant colony optimization algorithm provides a more effective approach to remote sensing images classification.
Keywordsunsupervised classification pheromone data discretization ant colony optimization algorithm
Unable to display preview. Download preview PDF.
- 2.Shuang, L., Shengyan, D., Shuming, X.: Comparion and research on remote sensing classificision methods. J. Henan University Trans. 32, 70–73 (2002)Google Scholar
- 7.Dorigo, M., Dicaro, G.: Ant colony Optimization: A New Meta-heuristic. In: Proc. of 1999 IEEE Congress on Evolutionary Computation Proceedings (CEC 1999), pp. 1470–1477. IEEE Press, Washington (2001)Google Scholar
- 8.Zhenglong, W., Rujing, W., Minggui, T., Meisheng, X.: Mining Classification Rule Based on Colony Algorithm. J. Computer Engineering and Application 20, 30–33 (2004)Google Scholar
- 9.Shugen, W., Yun, Y., Ying, L., Chonghua, C.: Automatic Classification of Remotely Sensed Images Based on Artificial Ant Colony Algorithm. J. Computer Engineering and Application 29, 77–80 (2005)Google Scholar