TIPS

Masking by Layer

Feature Analyst® allows you to include or exclude certain areas of your image in order to either focus on a specific area in the image or ignore a selected area. The first type of masking, Masking by Pixel Value, allows you to eliminate either black or white pixels from consideration during an extraction. This can speed up an extraction if you have white or black borders around your image or if you have “blacked out” an area in an image. The other type of masking is called Masking by Layer, which allows you to isolate a Region of Exclusion or a Region of Interest.

Region of Exclusion

Masking by Region of Exclusion, is helpful when there are multiple features that closely resemble the target feature class. In the following example, some of the buildings are nearly the same color as the street. These areas can be masked out thus avoiding any confusion for the Feature Analyst Learner. The results in the above left example are extremely cluttered, with portions of the street that have been misidentified as building features. In the image at right, the extracted streets were used as a mask, yielding much better results. Note: You can select more than one layers to use for masking.

Region of Interest

The example below illustrates masking by Region of Interest. Feature Analyst allows users to select regions of interest so that the software classifies this specified area and ignores the rest of the image. For example, after extracting all of the pervious versus impervious surfaces in an image, one could further analyze the pervious surfaces in order to distinguish between multiple land cover types, such as trees and grass. By selecting an area of interest, the extraction process is expedited, and the likelihood that the software will return clutter is drastically reduced.

Click here to view the print version of this Tip of the Week.