Digital change detection technique by using multi temporal satellite imagery is important to understanding landscape dynamic. This research objective is to map forest changes in south east sulawesi using OBIA approach by remote sensing technique. The analysis was carried out by landsat 5 TM and landsat 8 OLI from 1996 to 2015. Firstly, image was reprojected using UTM south coordinate, secondly it was than choose the (ROI) regionof interest and composite the band, thirdly, OBIA segmentation was doing for each land use with SAGA GIS open source software. Then, segmentation result was classified using maximum likelihood classifier (MLC) TO PRODUCE lc THEMATIC MAP. Finally, change detection of MLC are performed by post classification algorithm. The accuary assessment of two classificationt map show the satisfactory agreement with the overall accuary higher than 89% and the kappa coeficient is 0.73. From 1996 to 2015, the area of closed forest and grass in the study has decreased by 30.70% and 35.8%, respectively. The open forest and other categories increased about 188.9% and 5% respectively.