Living Textbook
Dashboard
Show
Concepts
Instances
Contributors
External resources
Learning outcomes
List
1604 - Apply k-Means clustering to an image to extract spectrally homogeneous clusters
×
You've loaded this page without map support, so map tools will not work. Open this page
with map
instead.
Apply k-Means clustering to an image to extract spectrally homogeneous clusters
Concepts
[IP3-4-4] Clustering (unsupervised)
Unsupervised methods are defined as the identification of natural groups, or structures, within existing data. Clustering requires only the number of to-be generated classes as an input parameter and assigns spectrally defined classes to an image.