grow_crown: Grow a Spectrally Homogenous Object

In pieterbeck/CanHeMonR: Tools for canopy health monitoring through image analysis

Description

Grow a spectrally homogenous object from the center of a raster object outwards, based on Mahalonobis distance criteria

Usage

grow_crown(testim, prox_ind, startclus, prob_cut)

Arguments

testim	A single or multi-band raster image to grow a region in
prox_ind	A matrix, as produced by Proximity.indices, that details proximity to the center and neighbours for all the cells in testim
startclus	The number of cells at the center of the testim that should be taken for granted as belonging to the region
prob_cut	The cutoff to apply when considering the probability that a new pixel at the edge of the region, belongs to the region. The value should fall between 0 an 100. Higher values tend to produce larger regionss

Value

A list containing $crown: binary raster of the region grown, and matching the attributes of testim and $startcluster: the startcluster size required to produce a non-singular covariance matrix (required to calculate Mahalanobis distance to the startcluster)

Note

called in region_growing_wrapper.r

References

http://w3facility.org/question/r-how-to-split-spatial-multipolygon-to-spatialpolygons-element/

pieterbeck/CanHeMonR documentation built on May 25, 2019, 7:11 a.m.