org.openimaj.image.objectdetection.haar

Class Stage

java.lang.Object

org.openimaj.image.objectdetection.haar.Stage

@Reference(type=Inproceedings,
           author={"Viola, P.","Jones, M."},
           title="Rapid object detection using a boosted cascade of simple features",
           year="2001",
           booktitle="Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on",
           pages={" I","511 "," I","518 vol.1"},
           number="",
           volume="1",
           customData={"keywords"," AdaBoost; background regions; boosted simple feature cascade; classifiers; face detection; image processing; image representation; integral image; machine learning; object specific focus-of-attention mechanism; rapid object detection; real-time applications; statistical guarantees; visual object detection; feature extraction; image classification; image representation; learning (artificial intelligence); object detection;","doi","10.1109/CVPR.2001.990517","ISSN","1063-6919 "})
public class Stage
extends Object

A classification stage. The stage is made up of an ensemble of classifiers (which may themselves be trees or stumps). In order to evaluate the stage, the ensemble of classifiers is tested in order and the responses are summed. If the summed response exceeds a threshold (corresponding to a certain error-rate set during training) then the stage will pass and the successStage() should be evaluated. If the stage fails, then the failureStage() will be evaluated. The actual coordination of calling the successStage() or failureStage() is performed by the StageTreeClassifier.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

Constructor Summary

Constructors

Constructor and Description
Stage(float threshold, Classifier[] trees, Stage successStage, Stage failureStage) Construct a new stage.

Method Summary

Modifier and Type	Method and Description
Stage	failureStage() Get the next stage to evaluate should this one fail; if this stage fails and this is null then the whole StageTreeClassifier fails.
boolean	pass(SummedSqTiltAreaTable sat, float wvNorm, int x, int y) Test whether a stage passes.
Stage	successStage() Get the next stage to evaluate should this one pass; if this stage passes and this is null then the whole StageTreeClassifier passes.

Methods inherited from class java.lang.Object

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

Constructor Detail

Stage

public Stage(float threshold,
             Classifier[] trees,
             Stage successStage,
             Stage failureStage)

Construct a new stage.

Parameters:

threshold - the threshold for the stage to pass

trees - the classifier trees in the stage

successStage - the next stage after this one, or null if this is the last

failureStage - the failure stage (for trees rather than cascades)

Method Detail

pass

public boolean pass(SummedSqTiltAreaTable sat,
                    float wvNorm,
                    int x,
                    int y)

Test whether a stage passes. For the stage to pass, the sum of the responses from the internal classification trees or stumps must exceed the stage's threshold.

Parameters:

sat - the summed area tables (integral images)

wvNorm - the normalisation based on the current window variance

x - the x-ordinate of the top-left of the window being tested

y - the y-ordinate of the top-left of the window being tested

Returns:

true if the stage passes; false otherwise

successStage

public Stage successStage()

Get the next stage to evaluate should this one pass; if this stage passes and this is null then the whole StageTreeClassifier passes.

Returns:

the next stage should pass(SummedSqTiltAreaTable, float, int, int) return true.

failureStage

public Stage failureStage()

Get the next stage to evaluate should this one fail; if this stage fails and this is null then the whole StageTreeClassifier fails.

Returns:

the next stage should pass(SummedSqTiltAreaTable, float, int, int) return false.