Image Representation and Retrieval Using Support Vector Machine and Fuzzy C-means Clustering Based Semantical Spaces

摘要: This paper presents a learning based framework for content-based image retrieval to bridge the gap between low-level features and high-level semantic information presented in images on semantically organized collections. Both supervised (probabilistic multi-class support vector machine) unsupervised (fuzzy c-means clustering) techniques are investigated associate global MPEG-7 color edge with their semantical and/or visual categories. It represents successive level of abstraction confidence or membership scores obtained from algorithms. A fusion-based similarity matching function is employed these new representations rank retrieve most similar compared query image. Experimental results generic database manually assigned categories medical different modalities examined body parts demonstrate effectiveness proposed approach commonly used Euclidean distance measure descriptors.