Multimodal page classification in administrative document image streams

Marçal Rusiñol, Volkmar Frinken, Dimosthenis Karatzas, Andrew D. Bagdanov, Josep Lladós

Research output: Contribution to journalArticleResearchpeer-review

23 Citations (Scopus)


© 2014, Springer-Verlag Berlin Heidelberg. In this paper, we present a page classification application in a banking workflow. The proposed architecture represents administrative document images by merging visual and textual descriptions. The visual description is based on a hierarchical representation of the pixel intensity distribution. The textual description uses latent semantic analysis to represent document content as a mixture of topics. Several off-the-shelf classifiers and different strategies for combining visual and textual cues have been evaluated. A final step uses an n-gram model of the page stream allowing a finer-grained classification of pages. The proposed method has been tested in a real large-scale environment and we report results on a dataset of 70,000 pages.
Original languageEnglish
Pages (from-to)331-341
JournalInternational Journal on Document Analysis and Recognition
Issue number4
Publication statusPublished - 1 Jan 2014


  • Digital mail room
  • Multimodal page classification
  • Visual and textual document description


Dive into the research topics of 'Multimodal page classification in administrative document image streams'. Together they form a unique fingerprint.

Cite this