By Topic

A parallel hierarchical radiosity algorithm for complex scenes

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Chen-Chin Feng ; Dept. of Comput. Sci., Nat. Tsing Hua Univ., Hsinchu, Taiwan ; Shi-Nine Yang

The radiosity method is known for generating high quality images of diffuse environments. However it suffers from extensive computation and requires a large amount of memory, especially when the given scene is rather complex. As parallel computing becomes widely available, it is natural to exploit parallelism to speed up radiosity computations. The paper presents a general parallel radiosity algorithm for complex scenes based on a hierarchical method. To achieve the generality, the algorithm is primarily designed for a parallel model machine with distributed memory so that it can be realized easily on any set of homogeneous or heterogeneous machines with message passing capability. To cope with the huge volume of input data and to solve the task allocation problem in parallel computation, the complex scene is first subdivided into cells by axial occluders such as walls, floors and ceilings, and then every cell is augmented to form its cell-visible set according to cell visibility. Then the radiosity of these cell-visible sets are scheduled for parallel processing. In order to facilitate the parallel construction of visible sets, the notion of visible buffer is introduced. Moreover, by exploiting the principle of spatial coherence, a scheduling scheme for visible sets is incorporated so that both the communication and disk access overhead can be reduced. Finally the authors implement their algorithm on the PVM (Parallel Virtual Machine) programming environment over 8 DEC Alpha 3000 machines to investigate its performance. The experiment results show that the parallel algorithm achieves good speedup

Published in:

Parallel Rendering, 1997. PRS 97. Proceedings. IEEE Symposium on

Date of Conference:

20-21 Oct 1997