|
| << vorige | volgende >> |
Tijdschrift beschrijving
Alle jaargangen van het bijbehorende tijdschrift
Alle afleveringen van het bijbehorende jaargang
Alle artikelen van de bijbehorende aflevering
Details van artikel 47 van 268 gevonden artikelen
| Titel: |
Automated delineation of coastline from polarimetric SAR imagery |
| Auteur: |
Yu, Y. Acton, S. T. |
| Verschenen in: |
International journal of remote sensing |
| Paginering: | Jaargang 25 (2004) nr. 17 pagina's 3423-3438 |
| Jaar: |
2004-09 |
| Inhoud: |
In this paper we present a new diffusion-based method for the delineation of coastlines from space-borne polarimetric SAR imagery of coastal urban areas. Both polarimetric filtering and speckle reducing anisotropic diffusion (SRAD) are exploited to generate a base image where speckle is reduced and edges are enhanced. The primary edge information is then derived from the base image using the instantaneous coefficient of variation edge detector. Next, the resulting edge image is parsed by a watershed transform, which partitions the image into disjoint segments where the division lines between segments are collocated with detected edges. The over-segmentation problem associated with the watershed transform is solved by a region merging technique that combines neighbouring segments with similar radar brightness. As a result, undesired boundary segments are eliminated and true coastlines are correctly delineated. The proposed algorithm has been applied to a space-borne polarimetric SAR dataset, demonstrating a good visual match between the detected coastline and the manually contoured coastline. The performance of the proposed algorithm is compared with those of two polarimetric SAR classification algorithms and two edge-based shoreline detection methods that are tailored to single polarization SAR images. Experimental results are shown using polarimetric SAR data from Hong Kong. |
| Uitgever: |
Taylor & Francis |
| Bronbestand: |
Elektronische Wetenschappelijke Tijdschriften |
Details van artikel 47 van 268 gevonden artikelen
| << vorige | volgende >> |