Minimum Mapping Unit (MMU)

The minimum mapping unit (MMU) is the smallest size that determines whether a feature is captured from a remotely sensed image, such as an aerial photograph or a satellite image. Determination of the MMU defines the amount of detail captured in the process of image interpretation.

The concept of an MMU can apply either to visual image interpretation, such as photogrammetric compilation from an aerial photograph, or to digital image processing, such as satellite image classification. For visual image interpretation, MMU refers to the size above which an areal feature is represented as a polygon, the size or dimension below which a long, nar-row polygonal feature is represented as a line, and the size below which a small area is represented as a point. In addition, it defines the size below which features would not be captured at all. For example, the U.S. Geological Survey (USGS) 1:24,000 topographic map specifications for capturing lakes and pond features states, “If a lake/pond is < 0.025 inches along the shortest axis and < 0.0025 square inches (10,000 square feet at 1:24,000 scale), then capture.” For streams and rivers, the USGS capture rules are more complex, but, essentially, they state that “if the shortest axis of a stream/river is < 0.025 inches but > 0.01 inches for a distance < 2.64 inches, then capture it as a 2-dimensional feature, but if it is < 0.01 inches for any distance, then capture it as a 1-dimensional feature.” For raster data, such as satellite images, the MMU can be no smaller than the pixel resolution of the image, although the MMU is often set to something larger than the image resolution in order to account for scale differences between analyses and to reduce reporting errors.

The minimum mapping unit may not be the same for all features in the image. Some features with greater importance may have very small MMUs, so that their presence is captured even when the MMU for other features is larger. For example, water features in an arid region may be captured at smaller sizes than in more humid regions because of their scarcity or importance, while the MMU for vegetation features might remain the same in both regions.

[Page 288]The size of the smallest unit mapped is a compromise between the level of detail captured by the image interpretation, the resolution requirements (for inventory, analysis, or mapping) for the GIS database that is being compiled, the legibility requirements for the printed map products to be created, and a reasonable trade-off between project operation costs and quality of the source data. A possible consequence is that even if the interpreter is able to distinguish single features, these may be grouped and categorized as part of a more generalized heterogeneous class.

It is usually necessary and desirable to limit the minimum size of the features that are delineated. Most often, the requirements that dictate the MMU are the resolution of the GIS data to be compiled and/or the scale of the map to be produced. Setting the MMU allows the resulting data to be a reduction of the visual and spatial complexity of the information contained in the image, especially when the information corresponding to the smallest features is of little or no interest for the purposes for which the map or database is developed. In addition, poor image resolution may prohibit the interpretation of features smaller than some minimum size threshold.

...