Recovering High Dynamic Range Radiance Maps from Photographs.pdf

From the SIGGRAPH’97 Conference Proceedings, August 1997 Recovering High Dynamic Range Radiance Maps from Photographs Paul E. Debevec Jitendra Malik University of California at Berkeley 1 ABSTRACT true measurements of relative radiance in the scene. For example, if one pixel has twice the value of another, it is unlikely that it observed We present a method of recovering high dynamic range radiance twice the radiance. Instead, there is usually an unknown, nonlinear maps from photographs taken with conventional imaging equip- mapping that determines how radiance in the scene becomes pixel ment. In our method, multiple photographs of the scene are taken values in the image. with different amounts of exposure. Our algorithm uses these dif- This nonlinear mapping is hard to know beforehand because it is ferently exposed photographs to recover the response function of the actually the composition of several nonlinear mappings that occur imaging process, up to factor of scale, using the assumption of reci- in the photographic process. In a conventional camera (see Fig. 1), procity. With the known response function, the algorithm can fuse the film is first exposed to light to form a latent image. The film is the multiple photographs into a single, high dynamic range radianc