Automatic cataloguing of advertisement in magazines

Ricardo Toledo, Ramon Baldrich, Xavier Orriols, Juan Sanchez, Xavier Binefa

    Research output: Contribution to journalArticleResearchpeer-review


    Advertising activities are one of the main aspects of companies. Its global evaluation is an important index, not only within economic sectors, but also for public economic policies. The evaluation of the adverts is carried out by specialised companies, which deliver advertisement expenditure information. Their work consist in an exhaustive collection of adverts (e.g., all adverts from any media), and deliver information (integrated from every media) relevant for their clients. The delivered information can range from statistical data to bear witness of an advertisement campaign. Important quality indicators for the delivered information are the exhaustiveness, the delay of delivering and a correct identification of different versions of an advert. In this article, we present an advanced computer-aided system for collecting and delivering advertising information from magazines and daily press. The system provides computer-aided data validation and exploitation resources. Using computerised document image analysis and image database indexing and retrieval, the system is able to locate an advert in a page, extract relevant quality indicators and search the advert (or similar ones) in a database. This tool is configured as an intranet and offers resources for image data acquisition, storage/retrieval and advert quality indicators extraction, however, the key of the system is the underlying idea of incorporating computer-aided visual information management. © Springer Science+Business Media, LLC 2006.
    Original languageEnglish
    Pages (from-to)119-144
    JournalMultimedia Tools and Applications
    Issue number2
    Publication statusPublished - 1 Nov 2006


    • Computer-aided visual inspection
    • Content-based image retrieval
    • Document image analysis
    • Multimedia system


    Dive into the research topics of 'Automatic cataloguing of advertisement in magazines'. Together they form a unique fingerprint.

    Cite this