Image Registration in a Coarse Three-Dimensional Virtual Environment

University of East Anglia, Norwich, England, United Kingdom
Computer Graphics Forum (Impact Factor: 1.64). 03/2006; 25(1):69-82. DOI: 10.1111/j.1467-8659.2006.00918.x
Source: DBLP


Abstract In recent years, the availability of off-the-shelf geometric data for an urban environment has increased. During rendering, ground level images are mapped onto the façades of the buildings to improve the visual quality of the scene. This paper focuses on a technique that enables ground level images to be automatically integrated into an existing coarse three-dimensional environment. The approach utilises the planar nature of architectural scenes to enable the automatic extraction of a building façade from an image and its registration into the virtual environment.

Download full-text


Available from: A. M. Day, Feb 25, 2014
  • Source
    • "A variety of methods in texture mapping have been proposed. On the data source aspect, it can be classified into aerial images (Weinhaus and Devich, 1999; Wu et al., 2007), ground based images (Förstner and Gülch, 1999; Laycock and Day, 2006; Tsai and Lin, 2007) and mixture (Kersten et al., 2004). The aerial images can obtain large area texture information and roof shapes, but the occlusion between neighboring buildings may be more serious. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Cyber city displays 3D geoinformatics for virtual reality in computer vision. On the geometric aspect, the purpose is to establish 3D models. In the visualization, the objects need more texture details to reach reality. This research registers aerial photos onto building facades. In which, building roofs and walls are to be treated. Generally, the occlusion due to the wall is more complicated than the roof. Therefore, before texture mapping, we need to detect occlusions. This research uses traditional aerial photos and oblique helicopter images. The main task is to detect hidden areas for roofs and walls, then compensate them with multi-view images. The wall images that we choose are based on oblique photography. For a target wall, we select the largest oblique angle for the prior image in order to get more homogeneous image resolution. However, the occlusion problem by neighboring buildings is more serious. It is the major task of this study to select the optimal combination of images. For a roof, we use the vertical photos to obtain better resolution. The essence of the step is to register the roof surface with images. Experimental results indicate that high fidelity may be reached.
    Preview · Article · Jan 2008
  • Source

    Preview · Article · Jan 2006
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In the first part of this article, we analyze the relation between local image structures (i.e., homogeneous, edge-like, corner-like or texture-like structures) and the underlying local 3D structure (represented in terms of continuous surfaces and different kinds of 3D discontinuities) using range data with real-world color images. We find that homogeneous image structures correspond to continuous surfaces, and discontinuities are mainly formed by edge-like or corner-like structures, which we discuss regarding potential computer vision applications and existing assumptions about the 3D world. In the second part, we utilize the measurements developed in the first part to investigate how the depth at homogeneous image structures is related to the depth of neighbor edges. For this, we first extract the local 3D structure of regularly sampled points, and then, analyze the coplanarity relation between these local 3D structures. We show that the likelihood to find a certain depth at a homogeneous image patch depends on the distance between the image patch and a neighbor edge. We find that this dependence is higher when there is a second neighbor edge which is coplanar with the first neighbor edge. These results allow deriving statistically based prediction models for depth interpolation on homogeneous image structures.
    Full-text · Article · Jul 2007 · Network Computation in Neural Systems