Selection of an appropriate classification technique for coastal biomass mapping using high and low resolution dataset

• Main Authors: Saeed, U., Fatima, K., Gilani, H., Daud, A., Ashraf, S.
• Format: Conference Proceeding
• Language: English
• Subjects: Biomass; Ecosystems; Image resolution; Marine animals; Maximum likelihood estimation; Remote sensing; Satellites; Sea measurements; Vegetation mapping; Water pollution
• DOI: 10.1109/ICAST.2008.4747691

Mangrove forest is an integral part of inter-tidal zone of the coastal environment extending throughout the tropics and subtropics of the world. There is a wide spectrum of economic and ecological utility of Indus delta mangrove forests, sixth largest man groves forest of 92 countries. Over the last decades, numerous classification techniques/software have been used to extract deltaic vegetation information through remotely sensed data. In the present research paper, two different techniques (maximum likelihood classification and subpixel classification) were applied at high and low resolution satellite data. The aim was to propose best classification technique for the mapping and management of coastal green gold. Study highlights the draw backs of traditional classification results at low resolution satellite dataset. Subpixel classification results applied on medium resolution image were satisfactory and comparable with maximum likelihood classification results at high resolution satellite data.