Image Classification Using an Ant Colony Optimization Approach
Automatic semantic clustering of image databases is a very challenging research problem. Clustering is the unsupervised classification of patterns (data items or feature vectors) into groups (clusters). Clustering algorithms usually employ a similarity measure in order to partition the database such that data points in the same partition are more similar than points in different partitions. In this paper an Ant Colony Optimization (ACO) and its learning mechanism is integrated with the K-means approach to solve image classification problems. Our simulation results show that the proposed method makes K-Means less dependent on the initial parameters such as randomly chosen initial cluster centers. Selected results from experiments of the proposed method using two different image databases are presented.
KeywordsAnt Colony Optimization (ACO) K-Means Image Classification and Clustering
Unable to display preview. Download preview PDF.