digilib@itb.ac.id +62 812 2508 8800

ABSTRAK Achmad Widiyanto
PUBLIC Irwan Sofiyan

Landslides are the movement of masses of rock, soil or debris down a slope. In recent years, the intensity of landslides in Indonesia has increased, with a wider distribution of disaster areas. Losses resulting from landslides can be in the form of casualties, environmental damage, loss of property, and psychological impacts. Lebak Regency has experienced several landslides, such as what happened in Cihuni Hamlet, Curug Panjang Village, Cikulur District in 2022 which resulted in residents' houses being damaged and will be relocated. Ground movement can be an indicator of the potential for landslides to occur, landslide detection is necessary for mitigation efforts. This study aims to analyze indications of ground motion for landslide detection with a case study in Curug Panjang Village, Cikulur District, Lebak Regency. The data used in this study are very high resolution satellite image data with the recording years 2013 and 2020. Ground motions are identified based on the deformation of objects recorded in the two images. The method used is Object Based Image Analysis (OBIA). Calculation of deformation with OBIA for ground motion detection consists of two stages, namely segmentation and classification. Segmentation is the process of building objects from pixels into the same object or segment. Furthermore, classification is carried out using objects or segments that are built in the image segmentation process based on the classification scheme and predetermined input features. Deformation is calculated from the objects formed from the OBIA classification process. From this research it can be concluded: (1) Curug Panjang Village, Cikulur District, Lebak Regency has experienced ground motion with an average speed of 19 cm/year, the speed of ground movement is included in the very slow category; (2) The horizontal vector of ground motion is random or does not form a certain pattern; (3) Residential locations in Curug Panjang Village have a very high risk of landslides.