EXPERT CLASSIFICATION FOR EUCALYPTUS PLANTATION USING LANDSAT-5 DATA: A CASE STUDY OF LAM THAMENCHAI DISTRICT, NAKHON RATCHASIMA PROVINCE, THAILAND

摘要

This study aims to develop an expert classification technique to identify eucalyptus plantation areas of Lam Thamenchai District, Nakhon Ratchasima province in Thailand.Multi-date LANDSAT-5 data acquired in January and March 2011, were used in development of the expert classification technique, since Landsat-5 data is found that these can be useful for providing information on Land use and Land cover classification. All imageries are finished on the digital image pre-processing steps. In addition, the important numerical data sets, including NDVI data, NDWI data, band ratio data and result from unsupervised classification data using Isodata method, are created to be the rule set for classification the eucalyptus plantation areas. Moreover, spectral signature of eucalyptus plantation derived from LANDS AT - 5 is also applied to make the new rule set more reliable. The result is compared with the result derived from maximum likelihood classification method, which is found that the result from the new technique is more accurate than the traditional method.In conclusion, the expert classification technique considering both spectral signature and other significant numeric indicators, is a proficient method to efficiently discriminate eucalyptus plantation area of Lam Thamenchai District, Nakhon Ratchasima province in Thailand.