Web Image Re-Ranking Using Query-Specific Semantic Signatures
- S . Ramya, Mr . K . Karthik
Traditional methods for image retrieval used metadata associated with images, commonly known as keywords. These methods empowered many World Wide Web (WWW) search engines and achieved reasonable amount of accuracy. A data base shape, color, texture of content based image retrieval (CBIR) and classification algorithm is based on the K-means clustering is proposed in this paper. The algorithm is found the content based image capture with intra cluster and cluster will be calculated a similarity between the shape and texture of image grouping with minimum euclidean distance considerations in taking a picture of the relevant semantics of a database. The image classification using k-means clustering has been applied successfully in shape, colour, texture data base image. By using this algorithm we can classified any type of colour data base image in different field. Performance evaluation methods are now done with precision and recal for different databases. Response time to fine the most signatures of sixth is 85 %.