Dynamic daylighting simulations from static high dynamic range imagery using extrapolation and daylight coefficient methodologies

This paper describes the development of a technique for extrapolation of dynamic daylighting simulations from a limited number of high dynamic range photographs.  This technique allows us to photographically capture and measure per-pixel lighting quantities from existing spaces in a limited time frame; and the measured information is used to establish a statistics based daylight coefficient model for the studied scene. It negates the need to explicitly model the geometry, material and lighting properties in existing environments, as they would be required in a typical simulation and daylight coefficient computation. Statistics based daylight coefficients can be used to perform daylighting simulations under any generic, arbitrary or physically occurring sky conditions.