Journal of Multimedia, Vol 5, No 2 (2010), 123-141, Apr 2010
doi:10.4304/jmm.5.2.123-141

Text Readability within Video Retrieval Applications: A Study On CCTV Analysis

Neil Newbold, Lee Gillam

Abstract


 The indexing and retrieval of video footage requires appropriate annotation of the video for search queries to be able to provide useful results. This paper discusses an approach to automating video annotation based on an expanded consideration of readability that covers both text factors and cognitive factors. The eventual aim is the selection of ontological elements that support wider ranges of user queries through limited sets of annotations derived automatically through the analysis of expert annotations of prior content. We describe how considerations of readability influence the approach taken to ontology extraction components of the system in development, and the automatic population of a CCTV ontology from analysis of expert transcripts of video footage. Considerations are made of the semantic content of the expert transcripts through theories on readability analysis and terminology extraction to provide knowledgebased video retrieval. Using readability studies to improve the text, we suggest that the semantic content can be made more accessible and improves the terminology extraction process which highlights the key concepts. This information can be used to determine relationships in the text, as a proxy for relationships between video objects with strong potential for interlinkage.



Keywords


video retrieval, inter-annotation, readability, automatic annotation

References



Full Text: PDF


Journal of Multimedia (JMM, ISSN 1796-2048)

Copyright @ 2006-2014 by ACADEMY PUBLISHER – All rights reserved.