ABSTRACT
The 3D photo-realistic building model does not only present the building in the three-dimensional perspective but also reveals its real appearance. There have been a number of approaches for reconstructing the geometric model from photogrammetric images, from LiDAR point cloud, or from both of them. However, the façade texture generation and mapping require massive labor works, therefore, are still the bottle neck in the photo-realistic building modeling process. A semi-automated approach to modeling 3D photo-realistic buildings using smartphones or tablet PCs is proposed in this paper. The built-in GPS
and G-sensor are capable of delivering the approximate 3D coordinates and 3 rotation angles around three axes. However, they are not accurate enough for the further photogrammetric operations. The reconstruction of 3D realistic building models consists of three major issues: (1) geometrically modelling the object; (2) determining the image orientation; (3) generating the realistic façade texture from photographs. By introducing the “Floating Model” concept, the object modelling and image orientation problem can be solved efficiently through the semi-automated procedures based on the Least-squares Model-image fitting (LSMIF). A friendly human-machine inter-acting interface is designed for an operator to choose suitable model, and to move, to rotate, or to resize the model so it can approximately fit to all of the images. An ad-hoc LSMIF algorithm is developed to solve the optimal fitting between projected model line segments and extracted edge pixels. Since the object model can be extracted and the photo orientation can be determined, the creation of realistic texture image, which is also called inverse mapping, can be automated by coordinate transforming and image resampling. For better understanding the camera characteristics on the smartphone or tablet PC, a series of camera calibration process has been executed to derive interior orientation parameters before taking façade pictures. Three representative buildings in NTNU campus are selected for the experimental tests. The geometric models are reconstructed by fitting floating models to aerial photogrammetric images. The façade photos are taken by smart phones on the ground. The result shows that the proposed approach is practicable, but the lens distortion must be corrected before creating texture image. Since the iterative LSMIF algorithm requires initial parameters, the position and pose derived from built-in sensors must fall in the pull-in range.
KEY WORDS
Three-dimensional Building Modelling, Photo-realism City Model, Texture Rendering, Direct Georeferencing, Automation, Digital Close-range Photogrammetry, Virtual Reality
FLOWCHART
The proposed photo-realistic 3D building modelling procedure
Attached File
isprsarchives-xxxix-b3-351-2012.pdf |