Maximum likelihood is a common classifier used for land use classification. It calculates the likelihood of an object to belong to each class based on an expected distribution and a metric of distance.
Maintainer: Caio Hamamura caiohamamura@gmail.com (ORCID)
The most common implementation, like in this package, will assume normal distributed variables within classes, and calculate the distance, based on Mahalanobis distance.
Mather, P. M. (1985). Remote sensing letters: A computationally efficient maximum-likelihood classifier employing prior probabilities for remotely-sensed data. International Journal of Remote Sensing, 6(2), 369–376. tools:::Rd_expr_doi("10.1080/01431168508948456")
Imports
Useful links: