Generation of 3d-city Models and Their Utilisation in Image Sequences

In this paper we describe the construction of a city model and its support for the analysis of image sequences. Our city model consists of building models which were generated from large scale vector maps and laser altimeter data. First the vector map is analysed to group the outlines of buildings and to obtain a hierarchical description of buildings or building complexes. The base area of single buildings are used to mask the corresponding elevation data. Depending on the task prismatic or polyhedral object models are reconstructed from the masked elevation data. The interpretation of image sequences taken by an airborne sensor in oblique view can be supported by a 3D city model. Possible GIS applications could be automated overlaying of selected buildings or query detailed building information from a database by interactively pointing on a frame in a sequence. The projection parameters of the model data are derived from GPS and INS. In practice the projected building contours do not exactly coincide with their image location. To overcome this problem an automated matching approach of image and model description is required. After image and vector map analysis correspondences between both scene descriptions can be found. These correspondences are used to correct the navigation data. This approach can easily be extended to image sequences. The corrected navigation data of a frame can be used as prediction for subsequent frames.