DocumentCode
922175
Title
An exact closed-form formula for d-dimensional quadtree decomposition of arbitrary hyperrectangles
Author
Chen, Shyh-Kwei
Author_Institution
IBM Thomas J. Watson Res. Center, Hawthorne, NY
Volume
18
Issue
6
fYear
2006
fDate
6/1/2006 12:00:00 AM
Firstpage
784
Lastpage
798
Abstract
In this paper, we solve the classic problem of computing the average number of decomposed quadtree blocks (quadrants, nodes, or pieces) in quadtree decomposition for an arbitrary hyperrectangle aligned with the axes. We derive a closed-form formula for general cases. The previously known state-of-the-art solution provided a closed-form solution for special cases and utilized these formulas to derive linearly interpolated formulas for general cases individually. However, there is no exact and uniform closed-form formula that fits all cases. Contrary to the top-down counting approach used by most prior solutions, we employ a bottom-up enumeration approach to transform the problem into one that involves the Cartesian product of d multisets of successive 2´s powers. Classic combinatorial enumeration techniques are applied to obtain an exact and uniform closed-form formula. The result is of theoretical interest since it is the first exact closed-form formula for arbitrary cases. Practically, it is nice to have a uniform formula for estimating the average number since a simple program can be conveniently constructed taking side lengths as inputs
Keywords
combinatorial mathematics; computational geometry; interpolation; quadtrees; Cartesian product; arbitrary hyperrectangle; bottom-up enumeration approach; closed-form formula; combinatorial enumeration technique; d-dimensional quadtree decomposition; linearly interpolated formula; top-down counting approach; Application software; Biomedical image processing; Closed-form solution; Costs; Data structures; Image coding; Image databases; Intrusion detection; Spatial databases; Visual databases; Quadtree; binary coding system.; combinatorial enumeration; geometric data; regular decomposition;
fLanguage
English
Journal_Title
Knowledge and Data Engineering, IEEE Transactions on
Publisher
ieee
ISSN
1041-4347
Type
jour
DOI
10.1109/TKDE.2006.86
Filename
1626233
Link To Document