Link to the USGS Home Page
image
image
image

Add Study Area dialog window:
image

To begin a land cover classification using RLCM, the first steps are to build a DG feature class and a land cover classification table. The DG feature class is the matrix or grid of points that cover the extent of a study area. The land cover classification table or attribute table is where the land cover classifications are stored.

The first time the RLCM is used and the Add Study Area dialog window is opened, there are four steps that the wizard walks through:

1. Enter the name of the study area and a description (Figure 4). Entering the name of the study area is required and, once entered, the study area name cannot be changed. The description is an optional item and can be modified at any time by opening the Add Study Area dialog. image
Figure 4: Step 1
2. At this step, the study area's spatial extent must be entered (Figure 5). There are two options to enter the extent: either upload an ESRI shapefile of the study area in geoimages coordinates, or manually enter the extent in geoimages coordinates (longitude/latitude). By default, the shapefile upload option is presented. To enter the extent manually, select the Enter Coordinates tab. Enter Coordinates tab. image
figure 5: Step 2
3. Provide the spatial resolution of the DG matrix (Figure 6). This value represents the separation in meters between each point or DG. It also will be the pixel or cell size of the DG feature class when converted to a raster format. image
Figure 6: Step 3
4. The final step (Figure 7) is to select a land cover classification system from the drop-down menu. Currently RLCM is delivered with four classification systems: 1. USGS Anderson Level II Land Use and Land Cover; 2. Global Land Cover 2000 (GLC2000); 3. U.S. National Land Cover Data 2001(NLCD); and 4. Sahel-West Africa Land Use and Land Cover. Once a classification system is selected, the bottom-left list box is populated with the complete list of classes. The next requirement is to select which classes will be displayed in the RLCM Attributer window. This allows you to minimize the number of classes to choose from while classifying. If some classes are not present in the study area, there is no reason to display them in the Attributer; removing them reduces the potential for misclassification. The Attributer class list can be appended to at any time during the classification processes, but no classes can be removed. If during the process of interpreting an image a class is not present in the Attributer list, it can be added by opening the Add Study Area dialog and navigating to the Land Cover Classification System step and moving the class from the left list box to the right list box. image
Figure 7: Step 4

Once all the steps are completed, the option is given to cycle back through each step to verify the inputs and change any item. When the verification is complete, move to the final step in the dialog and press the Finish button to build the DG feature class. This process may take several minutes (30+) to complete, depending on the density of the DG, the spatial extent of the study area, and the processing power of the computer building the data set.