Full Program »
Semantic-Based Image Retrieval Using Balanced Clustering Tree
In this paper, we perform a retrieval for similar images by content to classify images on the basis of the clustering balanced tree, C-Tree, and the k-NN algorithm (k-Nearest Neighboor), in which the input data is a query image and the output is a set of similar images and semantic classification for the query image. This structure is constructed on the basis of separating the nodes from the leaf node and growing towards the root to create a balanced tree. Then, for an input query image, a set of similar images are searched on the C-Tree to serve as the basis for the semantic classification of the query image based on the k-NN algorithm. On the base of these semantic classifications, the SPARQL query is automatically generated to query on constructed ontology for the image. We propose a model for semantic-based image retrieval on the C-Tree and ontology to analyze the semantics of an image and extract a similar set of images. To evaluate the effectiveness of the proposed method, we experimented with image datasets such as COREL (1000 images), Wang (10,800 images), ImageCLEF (20,000 images). The results are compared and evaluated with the relevant projects published recently on the same set of input data.