A raster image models geographic features by a
collection of grid cells rather like a scanned map or picture. Such
data is typically created from scanning paper sources, e.g.: ordnance
survey maps, aerial photographs or satellite images. Raster data tends
to be relatively high volume and does not possess intelligence, i.e.:
the attributes, including geo-referenced location, necessary to be
incorporated into spatial analysis. Raster maps are commonly used as a
backdrop to the vector layer.
Vector data are typically described as point,
line or area and are particularly suited to sophisticated geographical
or spatial analysis. Vector data is intelligent in that the map
features are directly linked to the GIS database, so that a selected
set of map features (e.g.: brownfield sites, road junctions, location
of traffic accidents etc) result from a particular database query.