A comparison of Sub-Pixel and maximum likelihood classification of Landsat ETM+ images to detect illegal logging in the Tropical Rain Forest of Berau, East Kalimantan, Indonesia

Selective logging is currently a widely adopted management practice throughout the tropics. Monitoring of spatial extent and intensity of such logging is, therefore, becoming an important issue for sustainable management of forest. In spite of successful use of RS in various field of forestry, detection and quantification of selective logging is still a problem. This study explores the possibility of using various approaches and Landsat-7 ETM+ image for the purpose. Two dataset of Landsat-7 ETM+ acquired on 16 August 2002 and 26 August 2000 of Labanan concession area East Kalimantan, were used. Field data of newly logged points, unlogged forest etc were collected during a fieldwork in September 2002. Maximum likelihood (ML) classification of original dataset of ETM+ 2002 and fused dataset with panchromatic were carried out. Sub-pixel classification approach was also tested. The results showed that the ML classification of fused image and sub-pixel classification approaches were found reasonable with overall accuracy and kappa 84%, 0.75; 86%, 0.71.