BUILDING FOOTPRINT EXTRACTION FROM HRSI DERIVED DSM AND ORTHOIMAGE

摘要

Three-dimensional building model is essential for city environment studies, such as urban planning, disaster management, loss estimation, risk modelling and assessment, disaster simulation, etc. The goal of this study is to develop a low cost workflow for the extraction of building footprint from Digital Surface Model (DSM) and orthoimage derived from High-Resolution Satellite Imagery (HRSI). The difficulties for this task is majorly due to buildings possessing various colors, textures and shapes of boundary. Several factors still pose challenges that interfere with building footprint extraction. Therefore, the objective of this research is to develop an algorithm that integrates road vector from Open Street Map (OSM), orthoimage and DSM to extract building footprints by Object Based Image Analysis (OBIA). In which, the orthoimage and DSM are derived from Pleiades satellite image stereo-pair. The major processing steps include the generation of image objects by multi-resolution image segmentation using orthoimage and DSM. After segmentation, several object features like spectral value, texture, and geometry can be derived. These object features are used to develop a rule set for classification. In order to explore the potential of the proposed approach, the selected study area contains a wide variety of buildings varying in color, shape, texture, and orientation.