EUSIPCO'2002 banner

Paper data
Image Similarity For Automatic Video Summarization

Yahiaoui Itheri, Eurecom
Merialdo Bernard, Eurecom
Huet Benoit, Eurecom

Page numbers in the proceedings:
Volume III pp 353-356

Content based Audio and Video Indexing (1/2)

Paper abstract
Image similarity is a key issue for many multimedia applications. Video summarization is no exception. We have recently proposed a number of methodologies for creating visually significant summaries of videos. Our approach relies heavily on the metric which decides on whether two video key-frames are similar or not. In this paper, we compare a number of histogram representations and possible distance measures with the objective of improving the quality of video summaries.

A PDF version is available here