Package sc.fiji.snt.analysis.curation

Class PlausibilityCheck.IntensityValley

java.lang.Object

sc.fiji.snt.analysis.curation.PlausibilityCheck.DeepCheck

sc.fiji.snt.analysis.curation.PlausibilityCheck.IntensityValley

Enclosing class:

PlausibilityCheck

public static class PlausibilityCheck.IntensityValley
extends PlausibilityCheck.DeepCheck

Flags localized intensity valleys along a path's intensity profile: discrete dips of one or a few nodes whose intensity is markedly below the surrounding peaks. Useful for catching nodes that pass through ambiguous regions (low local signal flanked by bright signal on both sides) where the trace's exact position is suspect.

Detection mirrors the inverted form of scipy.signal.find_peaks(prominence=...): for each candidate node, the prominence is computed as (min(leftMax, rightMax) - intensity[i]) / min(leftMax, rightMax) with the left/right maxima taken over a half-window around the candidate. Two-sided constraint (must be strictly below both window maxima) prevents flagging the start or end of a path's intensity profile. Adjacent candidates are merged into a single valley centered on the deepest node.

Off by default; intended to be enabled only on images where the pattern is meaningful (e.g., en-passant axons crossing dim regions).

Constructor Summary

Constructors

Constructor	Description
IntensityValley()

Method Summary

Modifier and Type	Method	Description
double	getMinProminence()
String	getName()
boolean	hasImage()
PlausibilityCheck.ImpactKind	impactKind()	Declares how this check's errors propagate.
PlausibilityCheck.Measurements	measure(Collection<Path> paths)	Computes the full distribution of this check's metric across paths, without applying the warning threshold.
List<PlausibilityCheck.Warning>	scan(Collection<Path> paths)	Scans a collection of paths for plausibility issues.
void	setImage(net.imglib2.RandomAccessibleInterval<? extends net.imglib2.type.numeric.RealType<?>> image)
void	setMinProminence(double v)

Methods inherited from class sc.fiji.snt.analysis.curation.PlausibilityCheck.DeepCheck

isEnabled, setEnabled

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

IntensityValley

public IntensityValley()

Method Details

getName

public String getName()

Specified by:

getName in class PlausibilityCheck.DeepCheck

Returns:

a short human-readable name for this check

impactKind

public PlausibilityCheck.ImpactKind impactKind()

Description copied from class: PlausibilityCheck.DeepCheck

Declares how this check's errors propagate. Default LOCAL for deep checks because most deep checks (radius jumps, monotonicity, signal quality) measure intrinsic per-path properties that don't propagate to descendants. Topology-class checks (e.g., Crossovers) override to PlausibilityCheck.ImpactKind.SUBTREE.

Overrides:

impactKind in class PlausibilityCheck.DeepCheck

setMinProminence

public void setMinProminence(double v)

getMinProminence

public double getMinProminence()

setImage

public void setImage(net.imglib2.RandomAccessibleInterval<? extends net.imglib2.type.numeric.RealType<?>> image)

hasImage

public boolean hasImage()

scan

public List<PlausibilityCheck.Warning> scan(Collection<Path> paths)

Description copied from class: PlausibilityCheck.DeepCheck

Scans a collection of paths for plausibility issues.

Specified by:

scan in class PlausibilityCheck.DeepCheck

Parameters:

paths - the paths to scan

Returns:

list of warnings (empty if no issues found)

measure

public PlausibilityCheck.Measurements measure(Collection<Path> paths)

Description copied from class: PlausibilityCheck.DeepCheck

Computes the full distribution of this check's metric across paths, without applying the warning threshold. Default implementation returns PlausibilityCheck.Measurements.EMPTY; subclasses override to enable per-check histogram popups in the Curation Assistant.

Overrides:

measure in class PlausibilityCheck.DeepCheck

Parameters:

paths - the paths to evaluate; null or empty yields PlausibilityCheck.Measurements.EMPTY.

Returns:

a PlausibilityCheck.Measurements record; never null.