Robust Method for IKONOS Imagery Fusion in Frequency Domain

AuthorsAlireza Safdarinezhad, Arash Rahmanizadeh, Mohammad Javad Valadan Zoej, Mehdi Mokhtarzade, Mojtaba Jannati
Conference TitleInternational Conference on Advancements in Electronics and Power Engineering (ICAEPE'2011)
Holding Date of Conference2011-12-1
Event PlaceBangkok
Page number103-108
PresentationSPEECH
Conference LevelInternational Conferences

Abstract

In this paper, the process of fusion multi-spectral (MS) and Panchromatic (PAN) images of IKONOS sensor using Fourier transform was performed. Results using high-frequency spatial information from PAN Image and spectral information of MS images show that the method has appropriate capability for fusion process. Then, with developing a new indicator using statistical and geometric analysis in feature space to assessment fusion results, the process of fusion in frequency domain analyzed and structure of optimal frequency for fusion process with assist of proposed indicators was identified. Finally, technique to speed up identification of structures of Frequency fusion was presented that significantly reduces the computational cost. Results showed that performance of proposed optimality indicator is independent of fusion technique and the indicator can also be used in other fusion methods.

Paper URL

tags: feature space, Fourier transform, images fusion, optimization, PAN-Sharpening.