AbstractSemiAutoTracker (Fiji Javadocs 0-SNAPSHOT API)

java.lang.Object
- fiji.plugin.trackmate.detection.semiauto.AbstractSemiAutoTracker<T>

Type Parameters:

T - the type of the source. Must extend RealType and NativeType to use with most TrackMate SpotDetector s.

All Implemented Interfaces:

Algorithm, MultiThreaded

Direct Known Subclasses:

SemiAutoTracker
```
public abstract class AbstractSemiAutoTracker<T extends RealType<T> & NativeType<T>>
extends Object
implements Algorithm, MultiThreaded
```
A class made to perform semi-automated tracking of spots in TrackMate & friends
The user has to select one spot, one a meaningful location. The spot location and its radius are then used to extract a small rectangular neighborhood in the next frame around the spot. The neighborhood is then passed to a SpotDetector that returns the spot it found. If a spot of Spot.QUALITY high enough is found near enough to the first spot center, then it is added to the model and linked with the first spot.
The process is then repeated, taking the newly found spot as a source for the next neighborhood. The model is updated live for every spot found.
The process halts when:
- no spots of quality high enough are found;
- spots of high quality are found, but too far from the initial spot;
- the source has no time-point left.
Author:

Jean-Yves Tinevez - 2013

Nested Class Summary

Nested Classes
Modifier and Type	Class and Description
`static class`	`AbstractSemiAutoTracker.SearchRegion<R>` A utility class made to return the information on a neighborhood generated from a source around a `Spot`.

Field Summary

Fields
Modifier and Type	Field and Description
`protected static String`	`BASE_ERROR_MESSAGE`
`protected double`	`distanceTolerance` How close must be the new spot found to be accepted, in radius units.
`protected String`	`errorMessage`
`protected Logger`	`logger`
`protected static double`	`NEIGHBORHOOD_FACTOR` Minimal size of neighborhoods, in spot diameter units.
`protected boolean`	`ok`
`protected double`	`qualityThreshold` The fraction of the initial quality above which we keep new spots.

Constructor Summary

Constructors
Constructor and Description

AbstractSemiAutoTracker(Model model, SelectionModel selectionModel, Logger logger)

Constructors
Constructor and Description
`AbstractSemiAutoTracker(Model model, SelectionModel selectionModel, Logger logger)`

Method Summary

All Methods Instance Methods Abstract Methods Concrete Methods
Modifier and Type	Method and Description
`boolean`	`checkInput()`
`protected SpotDetector<T>`	`createDetector(RandomAccessible<T> img, Interval interval, double[] calibration, double radius, double quality)` Returns a new instance of a `SpotDetector` that will inspect the neighborhood.
`protected abstract void`	`exposeSpot(Spot newSpot, Spot previousSpot)` This method is a hook for subclassers.
`String`	`getErrorMessage()`
`protected abstract AbstractSemiAutoTracker.SearchRegion<T>`	`getNeighborhood(Spot spot, int frame)` Returns a small neighborhood around the specified spot, but taken at the specified frame.
`int`	`getNumThreads()`
`boolean`	`process()`
`void`	`processSpot(Spot initialSpot)` Recursively investigates a neighborhood to find the most likely successor of a spot, starting with the specified spot and operating recursively.
`void`	`setNumThreads()`
`void`	`setNumThreads(int numThreads)`
`void`	`setParameters(double qualityThreshold, double distanceTolerance, int nFrames)` Configures this semi-automatic tracker.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Field Detail
  - NEIGHBORHOOD_FACTOR
```
protected static final double NEIGHBORHOOD_FACTOR
```
    Minimal size of neighborhoods, in spot diameter units.
    
    See Also:
    
    Constant Field Values
  - BASE_ERROR_MESSAGE
```
protected static final String BASE_ERROR_MESSAGE
```
    See Also:
    
    Constant Field Values
  - errorMessage
```
protected String errorMessage
```
  - ok
```
protected boolean ok
```
  - logger
```
protected final Logger logger
```
  - distanceTolerance
```
protected double distanceTolerance
```
    How close must be the new spot found to be accepted, in radius units.
  - qualityThreshold
```
protected double qualityThreshold
```
    The fraction of the initial quality above which we keep new spots. The highest, the more intolerant.
- Constructor Detail
  - AbstractSemiAutoTracker
```
public AbstractSemiAutoTracker(Model model,
                               SelectionModel selectionModel,
                               Logger logger)
```
- Method Detail
  - setParameters
```
public void setParameters(double qualityThreshold,
                          double distanceTolerance,
                          int nFrames)
```
    Configures this semi-automatic tracker.
    
    Parameters:
    
    qualityThreshold - the fraction of the initial quality above which we keep new spots. The highest, the more intolerant.
    
    distanceTolerance - how close must be the new spot found to be accepted, in radius units.
    
    nFrames - how many frames at most we track. Set it to 0 or negative to go as far as possible.
  - process
```
public boolean process()
```
    Specified by:
    
    process in interface Algorithm
  - processSpot
```
public void processSpot(Spot initialSpot)
```
    Recursively investigates a neighborhood to find the most likely successor of a spot, starting with the specified spot and operating recursively. The found spots are added to the Model given at construction and linked to the previous spot, until:
    - a spot of sufficient quality cannot be found in the close vicinity of the previous spot;
    - the raw image has exhausted all its time-points.
    This method can be called concurrently on several spots.
    Parameters:
    
    initialSpot - the spot to start detection with.
  - exposeSpot
```
protected abstract void exposeSpot(Spot newSpot,
                                   Spot previousSpot)
```
    This method is a hook for subclassers. It exposes the newly found spot just before it is added to the Model. This method allows concrete implementation to add some specific post-treatment to detected spots.
    
    Parameters:
    
    newSpot - the spot that just has been found by the detection mechanism.
    
    previousSpot - the spot in the previous frame whose neighborhood has been investigated to find the new spot. Already part of the model.
  - getNeighborhood
```
protected abstract AbstractSemiAutoTracker.SearchRegion<T> getNeighborhood(Spot spot,
                                                                           int frame)
```
    Returns a small neighborhood around the specified spot, but taken at the specified frame. Implementations have to decide what is the right size for the neighborhood, given the specified spot radius and location.
    Implementations can return null if for instance, the number of time frames in the raw source has been exhausted, or if the specified spot misses some information. This will be dealt with gracefully in the process() method.
    
    Parameters:
    
    spot - the spot the desired neighborhood is centered on.
    
    frame - the frame in the source image the desired neighborhood as to be taken.
    
    Returns:
    
    the neighborhood, as a AbstractSemiAutoTracker.SearchRegion. Concrete implementations have to specify the neighborhood location and calibration, so that the found spots can have their coordinates put back in the raw source coordinate system.
  - createDetector
```
protected SpotDetector<T> createDetector(RandomAccessible<T> img,
                                         Interval interval,
                                         double[] calibration,
                                         double radius,
                                         double quality)
```
    Returns a new instance of a SpotDetector that will inspect the neighborhood.
    
    Parameters:
    
    img - the source image.
    
    interval - defines the neighborhood to inspect.
    
    calibration - the pixel sizes to convert pixel coordinates into image coordinates.
    
    radius - the expected spot radius.
    
    quality - the quality threshold below which found spots will be discarded.
    
    Returns:
    
    a new SpotDetector.
  - checkInput
```
public boolean checkInput()
```
    Specified by:
    
    checkInput in interface Algorithm
  - getErrorMessage
```
public String getErrorMessage()
```
    Specified by:
    
    getErrorMessage in interface Algorithm
  - setNumThreads
```
public void setNumThreads()
```
    Specified by:
    
    setNumThreads in interface MultiThreaded
  - setNumThreads
```
public void setNumThreads(int numThreads)
```
    Specified by:
    
    setNumThreads in interface MultiThreaded
  - getNumThreads
```
public int getNumThreads()
```
    Specified by:
    
    getNumThreads in interface MultiThreaded

Class AbstractSemiAutoTracker<T extends RealType<T> & NativeType<T>>

Nested Class Summary

Field Summary

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

NEIGHBORHOOD_FACTOR

BASE_ERROR_MESSAGE

errorMessage

ok

logger

distanceTolerance

qualityThreshold

Constructor Detail

AbstractSemiAutoTracker

Method Detail

setParameters

process

processSpot

exposeSpot

getNeighborhood

createDetector

checkInput

getErrorMessage

setNumThreads

setNumThreads

getNumThreads