stream.news.learner.anchorshotdetection

Class ApplyImageModel

java.lang.Object

stream.AbstractProcessor

stream.image.AbstractImageProcessor

stream.news.learner.anchorshotdetection.ApplyImageModel

All Implemented Interfaces:

stream.Processor, stream.StatefulProcessor

public class ApplyImageModel
extends AbstractImageProcessor

This processor classifies a news video into anchorshots and news report shots, by matching each shot against an image, representing an anchorshot model. Pixels, that are black in the model, get ignored, all other pixels get compared pixelwise. If the difference between the current image and the model does not exceed a given threshold t, the shot is predicted to be an anchorshot.

Author:

Matthias

Field Summary

Fields inherited from class stream.image.AbstractImageProcessor

imageKey

Fields inherited from class stream.AbstractProcessor

context

Constructor Summary

Constructors

Constructor and Description
ApplyImageModel()

Method Summary

Methods

Modifier and Type	Method and Description
String	getModel() Delivers the location, where the processor currents looks for the model image.
String	getPredictionkey() Delivers the key under which the classifier currently stores the predicted label.
Integer	getT() Returns the current threshold the processor is working with.
void	init(stream.ProcessContext ctx)
stream.Data	process(stream.Data item, ImageRGB img)
void	setModel(String model) Tells the processor, where to find the model to be matched.
void	setPredictionkey(String predictionkey) Sets the key under which the classifier shall store the predicted label.
void	setT(Integer t) Sets the threshold t to a new value.

Methods inherited from class stream.image.AbstractImageProcessor

getImage, process, setImage

Methods inherited from class stream.AbstractProcessor

finish, resetState

Methods inherited from class java.lang.Object

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

Constructor Detail

ApplyImageModel

public ApplyImageModel()

Method Detail

setModel

public void setModel(String model)

Tells the processor, where to find the model to be matched.

Parameters:

model - String holding the location of the model image

getModel

public String getModel()

Delivers the location, where the processor currents looks for the model image.

Returns:

String holding the location of the model image

setT

public void setT(Integer t)

Sets the threshold t to a new value.

Parameters:

t - Threshold

getT

public Integer getT()

Returns the current threshold the processor is working with.

Returns:

Threshold

setPredictionkey

public void setPredictionkey(String predictionkey)

Sets the key under which the classifier shall store the predicted label.

Parameters:

predictionkey -

getPredictionkey

public String getPredictionkey()

Delivers the key under which the classifier currently stores the predicted label.

Returns:

init

public void init(stream.ProcessContext ctx)
          throws Exception

Specified by:

init in interface stream.StatefulProcessor

Overrides:

init in class stream.AbstractProcessor

Throws:

Exception

process

public stream.Data process(stream.Data item,
                  ImageRGB img)

Specified by:

process in class AbstractImageProcessor