locate.calsphere.MS70.voxel: Locates the voxel containing the calibration sphere of...

In arnejohannesholmin/AJ: Holmins functions, used also for loading all packages by library(AJ)

Description

Locates the voxel containing the calibration sphere of calibration data from the MS70 sonar. The method used is a geometrical clustering of the 'K' voxels of highest intensity, folowed by a ranging of the clusters based on the sum of the intensities of each cluster. If 'oldpos' is missing the calibration sphere is assigned the voxel position of the largest cluster. If 'oldpos' is present, the location needs to be in the proximity of the old position, given by 'step'. Also the size of the chosen cluster needs to be close to the old size. If these conditions are not met, NULL is returned for the 'voxel'.

Usage

locate.calsphere.MS70.voxel(data, ind = list(-(1:30), NULL, NULL), K = 100,
  oldpos = NULL, oldsize = NULL, step = c(2, 2, 2), margin = c(3, 2, 2),
  sizemargin = c(0.5, 1.5))

Arguments

data	is a list of data as returned from read.event(...,var=c("vbsc","voxels","beams")).
ind	is a list of indexes subsetting the data using subset_TSD() (see subset_TSD() for details).
K	is the number of the largest voxels to include in the clustering forming the basis of the funstion (see DESCRIPTION above).
oldpos	is a vector of three elements (radial, horizontal, vertical) representing the old voxel position of the calibration sphere.
oldsize	is a the sum of the voxels in old cluster.
step	is a vector of three elements (radial, horizontal, vertical) representing the proximity around the old position in which the new position is accepted.
margin	is a matrix of three elements (radial, horizontal, vertical) representing the distances separating clusters.
sizemargin	is a vector of two factors representing the interval on each size of 'oldsize' in which the new size must be contained. If sizemargin=[0.5,1.5] and 'size' of the relevant cluster is 10, the interval will be [5,15].

arnejohannesholmin/AJ documentation built on May 27, 2019, 4:06 a.m.