ERDAS IMAGINE Analysis Tools
In ERDAS IMAGINE, GIS analysis functions and algorithms are accessible through these main tools:
- script models created with SML
- graphical models created with Spatial Modeler or Model Maker
- prepackaged functions in the IMAGINE Workspace
Spatial Modeler Language
SML is the basis for all ERDAS IMAGINE GIS functions. SML is a modeling language for creating script (text) models for a variety of applications. Models may be used to create custom algorithms that best suit your data and objectives.
Model Maker is the legacy module that is essentially SML linked to a graphical interface where you can create graphical models using a palette of easy-to-use tools. Graphical models can be run, edited, saved in libraries, or converted to script form and edited further, using SML.
References to Spatial Modeler in this chapter mean that the named procedure can be accomplished using both Model Maker and SML.
Spatial Modeler is the next generation spatial modeling suite of ERDAS IMAGINE. Spatial Modeler has been redesigned from the ground up, to provide a robust extensible processing architecture, coupled with a more responsive and intuitive user experience. The core engine is based on a Pull architecture, meaning that no computation is done until the data is needed, providing the ability to have real time feedback of the results. The system is extensible in both data types and operations on data types, providing support to a rich set of vector operators, in addition to the traditional raster functions.
The user interface is completely integrated into the IMAGINE Workspace as a new type of Application, which presents you with a blank sheet of paper, and set of tools for building a model. The model is built by dragging and dropping operators in the page, and then connecting their inputs and outputs to build a running model. The results of the model can be directly viewed in real time in the Preview window, or it can be run in a file to file mode.
The Image Interpretation tools house a set of common functions that were all created using either Model Maker or SML. They have been given a dialog interface to match the other processes in ERDAS IMAGINE. In most cases, these processes can be run from a single dialog. However, the actual models are also provided with the software to enable customized processing.
Many of the functions described in the following sections can be accomplished using any of these tools. SML is intended for more advanced analyses, and has been designed using natural language commands and simple syntax rules. Some applications may require a combination of these tools.
Customizing ERDAS IMAGINE Tools
Use ERDAS Macro Language (EML) to create and add new or customized dialogs. If new capabilities are needed, they can be created with the IMAGINE Developers Toolkit™. Using these tools, a GIS that is completely customized to a specific application and its preferences can be created.
See ERDAS IMAGINE Help for more information about EML and the IMAGINE Developers Toolkit.
Once the database (layers and attribute data) is assembled, the layers can be analyzed and new information extracted. Some information can be extracted simply by looking at the layers and visually comparing them to other layers. However, new information can be retrieved by combining and comparing layers using the following procedures:
- Proximity analysis—the process of categorizing and evaluating pixels based on their distances from other pixels in a specified class or classes.
- Contiguity analysis—identify regions of pixels in the same class and to filter out small regions.
- Neighborhood analysis —any image processing technique that takes surrounding pixels into consideration, such as convolution filtering and scanning. This is similar to the convolution filtering performed on continuous data. Several types of analyses can be performed, such as boundary, density, mean, sum, and so forth.
- Recoding—assign new class values to all or a subset of the classes in a layer.
- Overlaying—creates a new file with either the maximum or minimum value of the input layers.
- Indexing—adds the values of the input layers.
- Matrix analysis—outputs the coincidence values of the input layers.
- Graphical modeling—combine data layers in an unlimited number of ways. For example, an output layer created from modeling can represent the desired combination of themes from many input layers.
- Script modeling—offers all of the capabilities of graphical modeling with the ability to perform more complex functions, such as conditional looping.
Using an Area of Interest
Any of these functions can be performed on a single layer or multiple layers. You can also select a particular AOI that is defined in a separate file (AOI layer, thematic raster layer, or vector layer) or an AOI that is selected immediately preceding the operation by entering specific coordinates or by selecting the area in a Viewer.