Abstract:
Coral reefs are the most important benthic habitat in nearshore areas as they provide important ecosystem services and economic benefits. Most coral reef are threatened, degradation and loss, and hence, monitoring and management is very important. A study
was done in Lang Tengah Island to monitor the impact of development. This study developing a habitat map for documenting the changing conditions associated with Lang Tengah Island coastal ecosystem with QuickBird image. QuickBird image was used
because it is fine spatial resolution image which suitable for evaluating benthic habitats. Mapping of the benthic habitat was done by using unsupervised and supervised classification. Different classification techniques were evaluated to determine the best
classification technique for tropical benthic habitat mapping. This study is more focus on evaluates the common used supervised classification techniques, which are maximum likelihood classification algorithm, minimum distance to mean classification algorithm and parallelepiped classification algorithm. QuickBird image had successfully identified the benthic habitat but not until species. Band 1 (blue) and band 2 (green) image with accuracy of minimum 70.00 % for Pasir Air and Pasir Pandak was suitable for benthic mapping. Band 1 image more suitable for deeper water mapping, while band 2 image more suitable for shallower water mapping. Maximum likelihood classification algorithm
was the best classification techniques with highest accuracy compared to other supervised classification techniques. Further mapping project should carried out from time to time to
assess the changes of benthic habitat in Lang Tengah Island.
