Efficient representation of zooming information in videos using multi resolution mosaics

  • Authors:
  • Asif Masood;Wajeeha Kanwal

  • Affiliations:
  • Department of Computer Science, Military College of Signals, National University of Sciences and Technology, Pakistan;Department of Computer Science, Military College of Signals, National University of Sciences and Technology, Pakistan

  • Venue:
  • ICIC'10 Proceedings of the Advanced intelligent computing theories and applications, and 6th international conference on Intelligent computing
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Video sequence often includes zoom camera motion. As a result the same scene is represented by frames at different resolutions. Present video mosaicing techniques construct mosaic images by mapping video frames onto a single resolution reference image plane. In such a mosaic the registration of zoomed frames requires diminishing in size to fit in the resolution of the final mosaic image. Consequently, extra details from video zoom are lost. To address this problem of handling variations in resolution we propose a novel idea of multi resolution mosaics. There are two contributions in this work; first a featureless registration method with a minimal weighted fusion procedure. Secondly, a multi resolution mosaic structure which captures zooming information in videos.