Assessing optimal image fusion methods for very high spatial resolution satellite images to support coastal monitoring

Byungyun Yang, Minho Kim, Marguerite Madden

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

This study examines best image fusion approaches for generating pansharpened very high resolution (VHR) multispectral images to be utilized for monitoring coastal barrier island development. Selected fusion techniques assessed in this research come from the three categories of spectral substitution (e.g., Brovey transform and multiplicative merging), arithmetic merging (e.g., modified intensity-hue-saturation and principal component analysis), and spatial domain (e.g., high-pass filter, and subtractive resolution merge). The image fusion methods selected for this study were capable of producing pansharpened VHR images with more than three bands. Comparisons of fusion techniques were applied to images from three satellite sensors: United States commercial satellites IKONOS and QuickBird, and the Korean KOMPSAT II. Pansharpened VHR multispectral images were assessed by spectral and spatial quality measurements. Results satisfying both spectral and spatial quality revealed optimum pansharpened techniques necessary for regular coastal mapping of barrier islands. These techniques may also be used to assess the quality of recently available VHR imagery acquired by numerous international, government, and commercial VHR satellite programs.

Original languageEnglish
Pages (from-to)687-710
Number of pages24
JournalGIScience and Remote Sensing
Volume49
Issue number5
DOIs
StatePublished - 1 Sep 2012

Fingerprint

Dive into the research topics of 'Assessing optimal image fusion methods for very high spatial resolution satellite images to support coastal monitoring'. Together they form a unique fingerprint.

Cite this