Extraction of Polygonal Information from Gridded Data

Rectangularly sampled, or gridded, spatial data such as classified LANDSAT imagery and Digital Terrain Models (DTM's), are increasingly ubiquitous sources of cartographic information. For environmental analysis, gridded data must be related to point, line, and polygon descriptions of phenomena, such as highway networks and soil zones. In combining such disparate forms of data, analysts must convert the irregular data structures to regular ones (or vice versa). That is, either points, lines and polygons describing objects may be gridded , or objects may be extracted from gridded representations of phenomena. This paper discusses integrating such data sources. Principally, issues involved in extracting polygonal features from grids will be addressed.