RANSAC: Random sample consensus (RANSAC)

In stnava/patchMatchR: Local patch matching with flexible transformation models in n-dimensional images

Random sample consensus (RANSAC)

Description

Random sample consensus (RANSAC)

Usage

RANSAC(
  fixedPoints,
  movingPoints,
  transformType = "Affine",
  minNtoFit = 16,
  maxIterations = 20,
  errorThreshold = 1,
  goodProportion = 0.5,
  lambda = 1e-06,
  verbose = FALSE
)

Arguments

fixedPoints	fixed points matrix
movingPoints	moving points matrix
transformType	Affine, Rigid and Similarity currently supported
minNtoFit	the minimum number of data values required to fit the model.
maxIterations	the maximum number of iterations allowed in the algorithm
errorThreshold	a threshold value for determining when a test data point fits a model. this parameter is set based on the standard deviation in the random subset model. that is, a point fits the model error distribution if it is within the bracket of values between mean error plus or minus sd error times errorThreshold.
goodProportion	the fraction of close data values required to assert that a model fits well to data. that is, if equal to 0.5, then one would need 50 points to assert that a model fit is good if the whole dataset contains 100 points.
lambda	ridge penalty in zero to one
verbose	boolean

Value

output list contains best fitted model, inliers, outliers

stnava/patchMatchR documentation built on March 23, 2022, 6:47 a.m.