Interactive Lighting Design with Hierarchical Light Representation


LD-2 LD-1

                          Final result with indirect lighting                                          Assigning shadow                      Drawing highlighted regions

                       (click to see the image in full size)                    (click to see the image in full size)         (click to see the image in full size)




Lighting design plays a crucial role in indoor lighting design, computer cinematograph and many other applications .

Computer-assisted lighting design aims to find a lighting configuration that best approximates the illumination effect
specified  by  designers.  In this paper,  we present  an automatic  approach  for  lighting design, in which discrete and 
continuous  optimization  of  the  lighting  configuration,  including  the number, intensity , and position of lights, are
achieved.  Our  lighting  design  algorithm  consists  of  two  major  steps.  The  first  step  estimates  an  initial lighting
configuration  by  light  sampling  and  clustering. The initial light clusters are then recursively merged to form a light
hierarchy.  The  second  step  optimizes  the  lighting  configuration  by  alternatively  selecting  a  light cut on the light
hierarchy  to  determine the number of representative lights and optimizing the lighting parameters using the simplex
method.   To  speed  up  the  optimization  computation,   only  illumination  at  scene  vertices  that  are  important  to 
rendering  are  sampled  and  taken  into  account  in  the  optimization.  Using  the  proposed  approach, we develop a
lighting  design  system  that  can  compute  appropriate  lighting  configurations  to  generate  the illumination effects
iteratively painted and modified by a designer interactively.




Ÿ       Wen-Chieh Lin, Tsung-Shian Huang, Tan-Chi Ho, Yueh-Tse Chen, and Jung-Hong Chuang, “Interactive Lighting
Design with Hierarchical Light Representation
”, Eurographics Symposium on Rendering 2013. (Computer Graph-
ics Forum, Vol. 32, No. 4, 2013, pp. 133-142).


Additional results