The Feudal Priority Algorithm on Hidden-Surface Removal
Han-Ming Chen, Wen-Teng Wang
Proceedings of SIGGRAPH 96, August 1996, pp. 55--64.
Abstract: Development of a real-time shaded rendering approach for a frequently changing viewpoint or view vector is very important in the simulation of 3-D objects in Computer-Aided Design. A new approach is proposed in this paper to meet this demand in a very efficient manner. A pre-processing phase, in which a feudal priority tree is established for all polygons of an object, and a post-processing phase, in which a rendering priority list is searched for from the feudal priority tree for a new viewpoint or view vector, are included in our approach. The most time-consuming work is finished in the pre-processing phase which only has to be executed once for an object, and the relatively simple task is left to the post-processing phase, which is repeated when the viewpoint or view vector is changed. For the pre-processing phase, a static version and a dynamic version are proposed in this paper. The one-way priority relations of all polygons are computed in the former part of the dynamic pre-processing in a more efficient way than that in the static pre-processing, but the latter part of the dynamic pre-processing is still based on the static pre-processing. A new concept of "absolute priority" is introduced to systematically reduce the polygons in which a separating plane is to be searched for so the probability of finding the separating plane is much increased. This is the basis to implement another important concept of "separating before splitting" by which the polygon splittings are much reduced. Hence the efficiency in the pre-processing and the post-processing phases is highly increased.
@inproceedings{Chen:1996:TFP,
author = {Han-Ming Chen and Wen-Teng Wang},
title = {The Feudal Priority Algorithm on Hidden-Surface Removal},
booktitle = {Proceedings of SIGGRAPH 96},
pages = {55--64},
month = aug,
year = {1996},
}
Return to the search page.
graphbib: Powered by "bibsql" and "SQLite3."