Research on the Three-Dimensional Information Extraction of Urban Buildings from Single QiuckBird Remote Sensing Image

The three-dimensional information of urban buildings is an important issue in urban monitoring, urban planning, urban management and various economic activities. In this paper we introduce a probabilistic approach of urban buildings’ height and location extraction on remote sensing image. According to the distance between the corner of building roof and its shadow on remote sensing image, buildings’ height can be extracted from single remote sensing image and buildings’ location also can be extracted from the classified image. Compared with the measured values, the absolute error of extracted urban buildings’ height and location is less than 1 m. The method is simple, fast and efficient, and is useful in the three-dimensional information extraction of urban buildings.