Share Email Print
cover

Proceedings Paper

Pansharpening in coastal ecosystems using Worldview-2 imagery
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Both climate change and anthropogenic pressure impacts are producing a declining in ecosystem natural resources. In this work, a vulnerable coastal ecosystem, Maspalomas Natural Reserve (Canary Islands, Spain), is analyzed. The development of advanced image processing techniques, applied to new satellites with very high resolution sensors (VHR), are essential to obtain accurate and systematic information about such natural areas. Thus, remote sensing offers a practical and cost-effective means for a good environmental management although some improvements are needed by the application of pansharpening techniques. A preliminary assessment was performed selecting classical and new algorithms that could achieve good performance with WorldView-2 imagery. Moreover, different quality indices were used in order to asses which pansharpening technique gives a better fused image. A total of 7 pansharpening algorithms were analyzed using 6 spectral and spatial quality indices. The quality assessment was implemented for the whole set of multispectral bands and for those bands covered by the wavelength range of the panchromatic image and outside of it. After an extensive evaluation, the most suitable algorithm was the Weighted Wavelet ‘à trous’ through Fractal Dimension Maps technique which provided the best compromise between the spectral and spatial quality for the image. Finally, Quality Map Analysis was performed in order to study the fusion in each band at local level. As conclusion, novel analysis has been conducted covering the evaluation of fusion methods in shallow water areas. Hence, the excellent results provided by this study have been applied to the generation of challenging thematic maps of coastal and dunes protected areas.

Paper Details

Date Published: 18 October 2016
PDF: 14 pages
Proc. SPIE 10004, Image and Signal Processing for Remote Sensing XXII, 100041U (18 October 2016); doi: 10.1117/12.2241506
Show Author Affiliations
Edurne Ibarrola-Ulzurrun, Univ. de Las Palmas de Gran Canaria (Spain)
Univ. Politécnica de Madrid (Spain)
Javier Marcello-Ruiz, Univ. de Las Palmas de Gran Canaria (Spain)
Consuelo Gonzalo-Martin, Univ. Politécnica de Madrid (Spain)


Published in SPIE Proceedings Vol. 10004:
Image and Signal Processing for Remote Sensing XXII
Lorenzo Bruzzone; Francesca Bovolo, Editor(s)

© SPIE. Terms of Use
Back to Top