Fuzzy K-means Application to Semantic Clustering for Image Retrieval

摘要

Several approaches have been used in defining the semantic features of images. This paper considers the most efficient approach by comparing the high points of kmean algorithm and fuzzy k mean algorithms. We observe that both approaches are efficient however based on the experimental set up, result shows that hidden features of images such as color texture and shape are more captured using fuzzy based k mean approach. The paper concludes by recommending larger experimental setup for further testing.