Clustering visually similar images to improve image search engines
Thomas Deselaers, Daniel Keysers · RWTH Publications (RWTH Aachen) · 2003
At the moment Google image search is probably the only widely known way to search the world wide web for images.Google's search engine works based on text retrieval: The images are not indexed by their appearance but by text which can be found in the context of the image.To achieve enhancements for the user we propose to reorder the images using a combination of methods from computer vision and data mining.We use features invariant against translation and rotation to represent the image content and the k-means and LBG cluster algorithms to present the images in groups in a more convenient way to the user.To test this method we created a new database from Google image search results.