Advertisement Detection in Digitized Press Images

Asma Ouji (1) , Yann Leydier (1) , Frank Le Bourgeois (1)

1 imagine - Extraction de Caractéristiques et Identification

This paper presents a novel topic consisting in detecting advertisements in digitized documents. The system's mains uses are locating and filtering ads. A color segmentation approach which is strong to digitization noise is introduced. The color separation output is used to carry out layout segmentation in document pages and to compute visual features that control the classification process. Block classification results, given with a variety of magazine/newspaper pages, are presented and discussed.

Domaines: Informatique [cs]

Asma Ouji, Yann Leydier, Frank Le Bourgeois. Advertisement Detection in Digitized Press Images. IEEE International Conference on Multimedia & Expo , Jul 2011, Barcelonne, Spain. pp.1-6, ⟨10.1109/ICME.2011.6011890⟩. ⟨hal-01354387⟩