Mapping mangrove forest with multispectral images: contribution of spectral unmixing technique and object based image analysis - Archive ouverte HAL
Chapitre D'ouvrage Année : 2010

Mapping mangrove forest with multispectral images: contribution of spectral unmixing technique and object based image analysis

Résumé

This chapter is a contribution to ongoing efforts to map and monitor mangrove forest. It attempts to explain that statistical results are very dependant of methodology, given many mistakes and discrepancies when date to date comparison is made without the same methodological protocol at each date. The use of spectral unmixing technique and Object Based Image Analysis (OBIA) can widely improve accuracy of maps and spatial statistics. It gives opportunity to lay out a standard reproducible protocol which can avoid main spatial errors. We focus on multispectral images and don’t take account hyperspectral images.
Fichier non déposé

Dates et versions

hal-01109188 , version 1 (25-01-2015)

Identifiants

  • HAL Id : hal-01109188 , version 1

Citer

Françoise Debaine, Marc Robin, Eddy Renoux, Dabe Rakotonavalona Hobialisoa. Mapping mangrove forest with multispectral images: contribution of spectral unmixing technique and object based image analysis. Geomatic Solutions For Coastal Environments, Nova Science Publishers, Chapter 3, 2010, 978-1-61668-140-1. ⟨hal-01109188⟩
255 Consultations
0 Téléchargements

Partager

More