| Titre | Wetland monitoring using the curvelet-based change detection method on polarimetric SAR imagery |
| Auteur | Schmitt, A; Brisco, B |
| Source | Water vol. 5, no. 3, 2013 p. 1036-1051, https://doi.org/10.3390/w5031036 (Accès ouvert) |
| Année | 2013 |
| Séries alt. | Ressources naturelles Canada, Contribution externe 20181098 |
| Éditeur | MDPI AG |
| Document | publication en série |
| Lang. | anglais |
| DOI | https://doi.org/10.3390/w5031036 |
| Media | papier; en ligne; numérique |
| Formats | pdf |
| Sujets | télédétection; géophysique |
| Programme | Géosciences de changements climatiques |
| Diffusé | 2013 07 11 |
| Résumé | (disponible en anglais seulement)
One fundamental task in wetland monitoring is the regular mapping of (temporarily) flooded areas especially beneath vegetation. Due to the independence of
weather and illumination conditions, Synthetic Aperture Radar (SAR) sensors could provide a suitable data base. Using polarimetric modes enables the identification of flooded vegetation by means of the typical double-bounce scattering. In this paper
three decomposition techniques-Cloude-Pottier, Freeman-Durden, and Normalized Kennaugh elements-are compared to each other in terms of identifying the flooding extent as well as its temporal change. The image comparison along the time series is
performed with the help of the Curvelet-based Change Detection Method. The results indicate that the decomposition algorithm has a strong impact on the robustness and reliability of the change detection. The Normalized Kennaugh elements turn out to
be the optimal representation for Curvelet-based change detection processing. Furthermore, the co-polarized channels (same transmit and receive polarization in horizontal (HH) and vertical (VV) direction respectively) appear to be sufficient for
wetland monitoring so that dual-co-polarized imaging modes could be an alternative to conventional quad-polarized acquisitions. |
| GEOSCAN ID | 311452 |
|
|