Searching for logo and trademark images on the web

  • Authors:
  • Euripides G. M. Petrakis;Epimenides Voutsakis;Evangelos E. Milios

  • Affiliations:
  • Technical University of Crete (TUC), Chania, Greece;Technical University of Crete (TUC), Chania, Greece;Dalhousie University, Halifax, Nova Scotia, Canada

  • Venue:
  • Proceedings of the 6th ACM international conference on Image and video retrieval
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

This work shows that it is possible to exploit text and image content characteristics of logo and trademark images in Web pages for enhancing the performance of retrievals on the Web. Searching for important (authoritative) Web pages and images is a desirable feature of many Web search engines and is also taken into account. State-of-the-art methods for assigning higher ranking to important Web pages, over other Web pages satisfying the query selection criteria, are considered and evaluated. PicASHOW exploits this idea in retrieval of important images on the Web using link information alone. WPicASHOW (Weighted PicASHOW), is a weighted scheme for co-citation analysis incorporating within the link analysis method of PicASHOW the text and image content of the queries and of the Web pages. The experimental results demonstrate that Web search methods utilizing content information (or combination of content and link information) perform significantly better than methods using link information alone.