public class LocalSauvolaThreshold<T extends RealType<T>> extends LocalThreshold<T> implements Ops.Threshold.LocalSauvolaThreshold
This is a modification of Niblack's thresholding method. In contrast to the recommendation on parameters in the publication, this implementation operates on normalized images (to the [0, 1] range). Hence, the r parameter defaults to half the possible standard deviation in a normalized image, namely 0.5.
Sauvola J. and Pietaksinen M. (2000) "Adaptive Document Image Binarization" Pattern Recognition, 33(2): 225-236. PDF
Original ImageJ1 implementation by Gabriel Landini.
SpecialOp.FlavorNAME| Constructor and Description |
|---|
LocalSauvolaThreshold() |
| Modifier and Type | Method and Description |
|---|---|
boolean |
conforms()
True if the contingency holds; i.e., the current circumstances conform to
the contingency's requirements.
|
protected CenterAwareComputerOp<T,BitType> |
unaryComputer(T inClass,
BitType outClass) |
compute, getShape, initializein, out, setInput, setOutputops, setEnvironmentclone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, waitops, setEnvironmentcompute, getIndependentInstance, run, runruncandidates, filterArity, opprotected CenterAwareComputerOp<T,BitType> unaryComputer(T inClass, BitType outClass)
unaryComputer in class AbstractCenterAwareNeighborhoodBasedFilter<T extends RealType<T>,BitType>inClass - First element from the input RandomAccessibleInterval
that may be used for determining the class.outClass - First element from the output IterableInterval that
may be used for determining the class.public boolean conforms()
Contingentconforms in interface Contingentconforms in class LocalThreshold<T extends RealType<T>>Copyright © 2014–2022 ImageJ. All rights reserved.