We present the algorithms and implementation of an interactive system to capture and interpret an architectural design from a collection of small-scale physical elements. The user sketches a proposed design by arranging 3D wall modules and simple markers for windows, materials, and other design features on a table. The color-coded elements are captured by a camera mounted above the scene and recognized using computer vision techniques. The architectural design is automatically inferred from this rough physical sketch and a closed, 3D triangle mesh representation is constructed. We apply the system to architectural daylighting analysis using an interactive global illumination simulation that allows designers to explore alternative designs and new technologies for improving the sustainability of their buildings. The participants may interactively redesign the geometry and materials of the space by manipulating the physical design elements and visualize the revised lighting simulation.
[1]
Daniel C Glaser,et al.
USING VIDEO FOR ANALYZING DAYLIGHT SIMULATION TOOLS
,
2006
.
[2]
Barbara Cutler,et al.
Interactive selection of optimal fenestration materials for schematic architectural daylighting design
,
2008
.
[3]
Robert C. Bolles,et al.
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography
,
1981,
CACM.
[4]
Bernhard P. Wrobel,et al.
Multiple View Geometry in Computer Vision
,
2001
.
[5]
Christopher Young,et al.
Virtual Heliodon: Spatially Augmented Reality for Architectural Daylighting Design
,
2009,
2009 IEEE Virtual Reality Conference.
[6]
Greg Welch,et al.
Shader Lamps: Animating Real Objects With Image-Based Illumination
,
2001,
Rendering Techniques.