Using linear quadtrees to store vector data
The linear quadtree is adapted to store vector data by defining a new data structure called a segment quadtree. It uses a constant or bounded, amount of storage per node, represents straight lines exactly (i.e., it is not a digitized representation), and enables updates in a consistent manner (i.e., when a vector feature is deleted, the database can be restored to the state it would have been in had the deleted feature never been inserted). The segment quadtree is shown to meet these requirements whereas existing quadtree-like methods (e.g., the edge quadtree, strip tree, etc.) fail to satisfy them. In order to illustrate the usefulness of the segment quadtree, sample algorithms are discussed to insert and delete line segments as well as perform boundary following. The space requirements of segment quadtrees are also investigated.
Unable to display preview. Download preview PDF.
- [Klin7l]A. Klinger, Patterns and search statistics, in Optimizing Methods in Statistics, J.S. Rustagi, ED., Academic Press, New York, 1971, 303–337.Google Scholar
- [Same85a]H. Samet and C.A. Shaffer, A model for the analysis of neighbor finding in pointer-based quadtrees, to appear in IEEE Transactions on Pattern Analysis and Machine Intelligence, 1985 (see also University of Maryland Computer Science TR-1432).Google Scholar
- [Same85b]H. Samet and R. E. Webber, Storing a collection of polygons using quadtrees, to appear in ACM Transactions on Graphics, 1985 (see also Proceedings of Computer Vision arid Pattern Recognition 83, Washington, DC, June 1983, 127–132 and University of Maryland Computer Science TR-1372).Google Scholar