Tree species classification using worldview-3 data

Abstract

Tree species identification is a crucial matter in forest managing and supervision. In recent years, tree species identification has been studying a lot using high-resolution satellite imagery data of 0.5m spatial resolution Worldview-3. This has because of some of the tree species have very high market value type of wood. Besides that, Tropical forest stored a large stock of carbon that contributes to the huge amount of above and below ground biomass. Hence, tree species identification is important in order to manage and control such activities like critical timber logged, forest cut for agricultural and development purposes that would affect the carbon cycle and lead to global warming affect. Tree classification using conventional method is required more time and limited accessibility for site survey, so the remote sensing technology was applied to identify tree species. In this study, tree species classification was done to identify dominant tree species had in Ayer Hitam Forest Reserve Puchong to estimate biomass values. Support Vector Machine (SVM) supervised classification method was applied in this process. This study can give some ideas to researchers in the future on what type of improvements can be made for next research and is hoped to assist the forest managers to control and manage forest ecosystems in order to reduce climate change.