Automated Feature and Name Placement on Parallel Computers
James E. Mower · Cartography and Geographic Information Systems · 1993
Implementations of general-purpose automated name-placement algorithms characteristically require extensive amounts of serial computing time to select names from large databases and place them onto small-scale maps. This paper presents a parallel algorithm for the automated selection of point features from a scale-independent database, and their placement on maps at a continuous range of presentation scales. The algorithm has been implemented and evaluated on a Connection Machine 2, a single-instruction-stream, multiple-data-stream computer. The execution performance evaluations presented here suggest that parallel computing environments offer cartographers and geographic information systems specialists fast and flexible alternatives to serial models of computation.