Text and numerical data that are associated with the classes of a thematic layer or the features in a vector layer are called attributes. This information can take the form of character strings, integer numbers, or floating point numbers. Attributes work much like the data that are handled by database management software. You may define fields, which are categories of information about each class. A record is the set of all attribute data for one class. Each record is like an index card, containing information about one class or feature in a file of many index cards, which contain similar information for the other classes or features.
Attribute information for raster layers is stored in the image file. Vector attribute information is stored in either an INFO file, dbf file, or SDE database. In both cases, there are fields that are automatically generated by the software, but more fields can be added as needed to fully describe the data. Both are viewed in CellArrays to display and manipulate the information. However, raster and vector attributes are handled slightly differently, so a separate section on each follows.
In ERDAS IMAGINE, raster attributes for image files are accessible from the Table tab > Show Attributes option, or from the Raster Attribute Editor. Both consist of a CellArray, which is similar to a table or spreadsheet that not only displays the information, but also includes options for importing, exporting, copying, editing, and other operations.
The figure below shows the attributes for a land cover classification layer.
Raster Attributes for lnlandc.img
Most thematic layers contain the following attribute fields:
- Class Name
- Class Value
- Color table (red, green, and blue values)
- Opacity percentage
- Histogram (number of pixels in the file that belong to the class)
As many additional attribute fields as needed can be defined for each class.
See Classification for more information about the attribute information that is automatically generated when new thematic layers are created in the classification process.
Viewing Raster Attributes
Simply viewing attribute information can be a valuable analysis tool. Depending on the type of information associated with the layers of a database, processing may be further refined by comparing the attributes of several files. When both the raster layer and its associated attribute information are displayed, you can select features in one using the other. For example, to locate the class name associated with a particular area in a displayed image, simply click in that area with the mouse and the associated row is highlighted in the Raster Attribute CellArray.
Attribute information is accessible in several places throughout ERDAS IMAGINE. In some cases it is read-only and in other cases it is a fully functioning editor, allowing the information to be modified.
Manipulating Raster Attributes
The applications for manipulating attributes are as varied as the applications for GIS. The attribute information in a database depends on the goals of the project. Some of the attribute editing capabilities in ERDAS IMAGINE include:
- import/export ASCII information to and from other software packages, such as spreadsheets and word processors
- cut, copy, and paste individual cells, rows, or columns to and from the same Raster Attribute CellArray or among several Raster Attribute CellArrays
- generate reports that include all or a subset of the information in the Raster Attribute CellArray
- use formulas to populate cells
- directly edit cells by entering in new information
The Raster Attribute CellArray in ERDAS IMAGINE also includes a color cell column, so that class (object) colors can be viewed or changed. In addition to direct manipulation, attributes can be changed by other programs. For example, some of the functions calculate statistics that are automatically added to the Raster Attribute CellArray. Models that read and/or modify attribute information can also be written.
Vector attributes are stored in the Vector Attributes CellArrays. You can simply view attributes or use them to:
- select features in a vector layer for further processing
- determine how vectors are symbolized
- label features
The figure below shows the attributes for a vector layer with line features.
Vector Attributes CellArray
See Attribute Information for more information about vector attributes.