Content-based image retrival in the World Wide Web

In general the image search engines on the World Wide Web rely purely on the keywords around the images and the filenames, which produces a lot of irrelevant images in their search results. Alternative to this there are other methods based on content based image retrieval which requires user interaction to submit a query image, and gets images that are similar in content. In this paper we presented a novel approach that combines the two above methods for retrieval of relevant images. In this method we first retrieves the results of a keyword query from an existing image search engine, then clusters the results based on extracted image features such as color and texture and returns the cluster that is inferred to be the most relevant to the search query.