Rasterizing the CIA World Data Bank II

Abstract

The geographic database is central to a Geographical Information System. The database may commonly be either in vector or raster format. The CIA World Data Bank II is a public domain vector database of the entire world to one second resolution. However, the database, in its standard distribution format, contains numerous unmarked line segment intersections and ‘orphan’ line segments that represent unclosed polygons. This makes it nearly impossible to do many GIS operations such as point-in-polygon location and boundary following. This project makes use of the PMR (Polygonal Map Random) quadtree data structure as a spatial index to the vector database to support editing and rasterizing operations. Using the PMR quadtree, the entire database for internal boundaries and coastlines is searched to remove unmarked intersections and orphans. Then various additional boundaries for major seas, gulfs and bays are added. Finally, a novel approach for converting the database from its current form to a raster is developed. This technique first builds the database into a line quadtree structure and then uses a flood-fill algorithm to color the different polygons (countries, islands and lakes). In the final stage, the line quadtree is converted to a region quadtree. All editing and display programs were developed using the X Window System.