ERDAS IMAGINE is designed to integrate two data types, raster and vector, into one system. While the previous chapter explored the characteristics of raster data, this chapter is focused on vector data. The vector data structure in ERDAS IMAGINE is based on the ArcInfo data model (developed by Esri, Inc.) as well as providing interoperability with the formats supported by GeoMedia Warehouses. This chapter describes vector data, attribute information, and symbolization.
You do not need ArcInfo software or an ArcInfo license to use the vector capabilities in ERDAS IMAGINE. Since the ArcInfo data model is used in ERDAS IMAGINE, you can use ArcInfo coverages directly without importing them.
See Geographic Information Systems for information on editing vector layers and using vector data in a GIS.
Vector data consist of:
Each is illustrated in the following figure:
A point is represented by a single x, y coordinate pair. Points can represent the location of a geographic feature or a point that has no area, such as a mountain peak. Label points are also used to identify polygons (see the figure in the "Vertex" section below).
A line (polyline) is a set of line segments and represents a linear geographic feature, such as a river, road, or utility line. Lines can also represent nongeographical boundaries, such as voting districts, school zones, contour lines, and so forth.
A polygon is a closed line or closed set of lines defining a homogeneous area, such as soil type, land use, or water body. Polygons can also be used to represent nongeographical features, such as wildlife habitats, state borders, commercial districts, and so forth. Polygons also contain label points that identify the polygon. The label point links the polygon to its attributes.
The points that define a line are vertices. A vertex is a point that defines an element, such as the endpoint of a line segment or a location in a polygon where the line segment defining the polygon changes direction. The ending points of a line are called nodes. Each line has two nodes: a from-node and a to-node. The from-node is the first vertex in a line. The to-node is the last vertex in a line. Lines join other lines only at nodes. A series of lines in which the from-node of the first line joins the to-node of the last line is a polygon.
In the figure above, the line and the polygon are each defined by three vertices.
Vector data are expressed by the coordinates of vertices. The vertices that define each element are referenced with x, y, or Cartesian (Planar) coordinates. In some instances, those coordinates may be inches [as in some computer-aided design (CAD) applications], but often the coordinates are map coordinates, such as State Plane, UTM (Universal Transverse Mercator), or Lambert Conformal Conic. Vector data digitized from an ungeoreferenced image are expressed in file coordinates.
Vector layers in coverage format are referenced to coordinates or a map projection system using tic files that contain geographic control points for the layer. Every vector layer must have a tic file. Tics are not topologically linked to other features in the layer and do not have descriptive data associated with them.
Although it is possible to have points, lines, and polygons in a single layer, a layer typically consists of one type of feature. It is possible to have one vector layer for streams (lines) and another layer for parcels (polygons). A vector layer is defined as a set of features where each feature has a location (defined by coordinates and topological pointers to other features) and, possibly attributes (defined as a set of named items or variables). Vector layers contain both the vector features (points, lines, polygons) and the attribute information.
Usually, vector layers are also divided by the type of information they represent. This enables the user to isolate data into themes, similar to the themes used in raster layers. Political districts and soil types would probably be in separate layers, even though both are represented with polygons. If the project requires that the coincidence of features in two or more layers be studied, the user can overlay them or create a new layer.
See Geographic Information Systems for more information about analyzing vector layers.
The spatial relationships between features in a vector layer are defined using topology. In topological vector data, a mathematical procedure is used to define connections between features, identify adjacent polygons, and define a feature as a set of other features (for example, a polygon is made of connecting lines) (Environmental Systems Research Institute, 1990).
Topology is not automatically created when a vector layer is created. It must be added later using specific functions. Topology must be updated after a layer is edited also.
Digitizing describes how topology is created for a new or edited vector layer.
As mentioned above, the ERDAS IMAGINE vector structure is based on the ArcInfo data model used for ARC coverages. This georelational data model is actually a set of files using the computer’s operating system for file management, input and output. An ERDAS IMAGINE vector layer is stored in subdirectories on the disk. Vector data are represented by a set of logical tables of information, stored as files within the subdirectory. These files may serve the following purposes:
- define features
- provide feature attributes
- cross-reference feature definition files
- provide descriptive information for the coverage as a whole
A workspace is a location that contains one or more vector layers. Workspaces provide a convenient means for organizing layers into related groups. They also provide a place for the storage of tabular data not directly tied to a particular layer. Each workspace is completely independent. It is possible to have an unlimited number of workspaces and an unlimited number of vector layers in a workspace. The following table summarizes the types of files that are used to make up vector layers.
Line coordinates and topology
Polygon centroid coordinates
Label point coordinates and topology
Feature Attribute Files
Line (arc) attribute table
Polygon or point attribute table
Feature Cross-Reference File
Polygon/line/node cross-reference file
Layer history file
Coordinate definition file
Layer tolerance file
The figure below illustrates how a typical vector workspace is set up (Environmental Systems Research Institute, 1992).
Because some vector layers are stored in directories rather than in simple files, you must use the utilities provided in ERDAS IMAGINE to copy and rename them. A utility is provided to update path names that are no longer correct due to using regular system commands on vector layers.
See Esri documentation for more information about different vector files.