ITK
4.4.0
Insight Segmentation and Registration Toolkit
|
#include <itkTemporalProcessObject.h>
TemporalProcessObject implements a ProcessObject for the itk pipeline with the notion of a temporal region.
TemporalProcessObject acts as a pass-through in the inheritance tree in order to require that subclasses properly implement handeling of temporal regions. The key parameters of a temporal process object are:
m_UnitInputNumberOfFrames m_UnitOutputNumberOfFrames m_FrameSkipPerOutput m_InputStencilCurrentFrameIndex
These parameters are protected by default so that the filter writer can choose which to expose and which to keep constant. For a full explanation of each parameter, see the individual comments with their declarations. Subclasses of TemporalProcessObject must implement one of the following:
GenerateData() TemporalStreamingGenerateData()
By default, GenerateData will split the requested temporal region of the input and output temporal data objects into sub-requests of unit input and unit output size. For example, if the output requests 3 frames and the filter has unit input size of 2 and unit output size of 1, GenerateData will produce 3 sub-requests that each require the input to provide 2 frames. Using this default behavior, TemporalStreamingGenerateData must be implemented and can assume that the RequestedTemporalRegion on the input and output are unit size. If the temporal streaming behavior is not desired, GenerateData can be implemented to directly handle the full output requested temporal region in a single pass.
Definition at line 65 of file itkTemporalProcessObject.h.
Public Types | |
typedef SmartPointer< const Self > | ConstPointer |
typedef SmartPointer< Self > | Pointer |
typedef TemporalProcessObject | Self |
typedef ProcessObject | Superclass |
Public Types inherited from itk::ProcessObject | |
typedef SmartPointer< const Self > | ConstPointer |
typedef DataObject::DataObjectIdentifierType | DataObjectIdentifierType |
typedef DataObject::Pointer | DataObjectPointer |
typedef std::vector < DataObjectPointer > | DataObjectPointerArray |
typedef DataObjectPointerArray::size_type | DataObjectPointerArraySizeType |
typedef std::vector < DataObjectIdentifierType > | NameArray |
typedef SmartPointer< Self > | Pointer |
typedef ProcessObject | Self |
typedef Object | Superclass |
Public Types inherited from itk::Object | |
typedef SmartPointer< const Self > | ConstPointer |
typedef SmartPointer< Self > | Pointer |
typedef Object | Self |
typedef LightObject | Superclass |
Public Types inherited from itk::LightObject | |
typedef SmartPointer< const Self > | ConstPointer |
typedef SmartPointer< Self > | Pointer |
typedef LightObject | Self |
Protected Attributes | |
OffsetValueType | m_FrameSkipPerOutput |
SizeValueType | m_InputStencilCurrentFrameIndex |
SizeValueType | m_UnitInputNumberOfFrames |
SizeValueType | m_UnitOutputNumberOfFrames |
Protected Attributes inherited from itk::ProcessObject | |
TimeStamp | m_OutputInformationMTime |
bool | m_Updating |
Private Member Functions | |
void | operator= (const Self &) |
TemporalProcessObject (const Self &) | |
Additional Inherited Members | |
Static Public Member Functions inherited from itk::Object | |
static bool | GetGlobalWarningDisplay () |
static void | GlobalWarningDisplayOff () |
static void | GlobalWarningDisplayOn () |
static Pointer | New () |
static void | SetGlobalWarningDisplay (bool flag) |
Protected Types inherited from itk::LightObject | |
typedef int | InternalReferenceCountType |
typedef SmartPointer< const Self > itk::TemporalProcessObject::ConstPointer |
Definition at line 75 of file itkTemporalProcessObject.h.
typedef SmartPointer< Self > itk::TemporalProcessObject::Pointer |
Definition at line 74 of file itkTemporalProcessObject.h.
Standard class typedefs
Definition at line 72 of file itkTemporalProcessObject.h.
Definition at line 73 of file itkTemporalProcessObject.h.
|
protected |
Constructor that initializes members
|
inlineprotectedvirtual |
Empty Destructor
Definition at line 148 of file itkTemporalProcessObject.h.
|
private |
|
inlineprotectedvirtual |
Method that gets called by GenerateData after TemporalStreamingGenerateData. Subclasses can override this method to provide post-processing for the data after producing the requested output temporal region.
Definition at line 203 of file itkTemporalProcessObject.h.
|
inlineprotectedvirtual |
Method that gets called by GenerateData before TemporalStreamingGenerateData. Subclasses can override this method to provide pre-processing for the data before splitting up the requested output temporal region.
Reimplemented in itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, and itk::VideoToVideoFilter< TVideoStream, TVideoStream >.
Definition at line 196 of file itkTemporalProcessObject.h.
|
virtual |
Override EnlargeOutputRequestedRegion, GenerateOutputRequestedRegion, and GenerateInputRequestedRegion to handle temporal regions
Reimplemented from itk::ProcessObject.
|
protectedvirtual |
Explicitly handle temporal regions in EnlargeRequestedRegion. The default implementation makes sure that the output requested temporal region is contained in the buffered temporal region or has a duration that is a multiple of the unit output number of frames
|
virtual |
Override GenerateData to do temporal region streaming. This is analogous to the ThreadedGenerateData system implemented in ImageSource, but it functions slightly differently. Since most temporal processes are going to to need more input frames than they produce output frames for a single operation, we cannot use the same model as spatial streaming which assumes that the input requested region is fully populated before producing any output. Instead, once we have split the requested output region, we reset the requested temporal region of the input to each input requested temporal sub-region (in sequence) and re-propagate the temporal region request up the pipeline.
Reimplemented from itk::ProcessObject.
|
protectedvirtual |
Generate a default LargestPossibleRegion. This is used by temporal process objects that have no input. The default implementation starts at frame 0 and has infinite duration.
|
virtual |
What is the input requested region that is required to produce the output requested region? By default, the largest possible region is always required but this is overridden in many subclasses. For instance, for an image processing filter where an output pixel is a simple function of an input pixel, the input requested region will be set to the output requested region. For an image processing filter where an output pixel is a function of the pixels in a neighborhood of an input pixel, then the input requested region will need to be larger than the output requested region (to avoid introducing artificial boundary conditions). This function should never request an input region that is outside the the input largest possible region (i.e. implementations of this method should crop the input requested region at the boundaries of the input largest possible region).
Reimplemented from itk::ProcessObject.
Reimplemented in itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, and itk::VideoToVideoFilter< TVideoStream, TVideoStream >.
|
protectedvirtual |
Explicitly handle temporal regions in GenerateInputRequestedRegion. The default implementation sets the requested temporal region on the input to start at m_FrameSkipPerOutput times the start frame of the requested output temporal region with the duration needed to produce the entire requested output.
NOTE: This default propagation will be overwritten during the UpdateOutputData phase by the temporal streaming mechanism if a subclass implements TemporalStreamingGenerateData, but this propagation is provided so that subclasses which directly implement GenerateData will work correctly.
|
virtual |
Given one output whose requested region has been set, how should the requested regions for the remaining outputs of the process object be set? By default, all the outputs are set to the same requested region. If a filter needs to produce different requested regions for each output, for instance an image processing filter producing several outputs at different resolutions, then that filter may override this method and set the requested regions appropriatedly.
Note that a filter producing multiple outputs of different types is required to override this method. The default implementation can only correctly handle multiple outputs of the same type.
Reimplemented from itk::ProcessObject.
Reimplemented in itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, and itk::VideoToVideoFilter< TVideoStream, TVideoStream >.
|
protectedvirtual |
Explicitly handle temporal regions in GeneratOutputRegion. The default implementation checks to see if a request is currently set and if not, it sets the requested temporal region to match the largest possible temporal region
Note: If the largest possible temporal duration has infinite duration, the requested temporal region will be set to have duration 1.
Reimplemented in itk::VideoSource< TOutputVideoStream >, itk::VideoSource< TVideoStream >, and itk::VideoSource< itk::VideoStream< TImageToImageFilter::OutputImageType > >.
|
protectedvirtual |
|
virtual |
Run-time type information (and related methods).
Reimplemented from itk::ProcessObject.
Reimplemented in itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoToVideoFilter< TVideoStream, TVideoStream >, itk::VideoFileReader< TOutputVideoStream >, itk::VideoSource< TOutputVideoStream >, itk::VideoSource< TVideoStream >, itk::VideoSource< itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoFileWriter< TInputVideoStream >, itk::ImageFilterToVideoFilterWrapper< TImageToImageFilter >, itk::FrameDifferenceVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::FrameAverageVideoFilter< TInputVideoStream, TOutputVideoStream >, and itk::DecimateFramesVideoFilter< TVideoStream >.
|
virtual |
Get the number of frames of input required to produce output. We don't provide a Set method because we want some filters to be able to hold this as a constant.
|
virtual |
Get the number of frames of output produced for a single set of input frames. We don't provide a Set method because we want some filters to be able to hold this as a constant.
|
private |
|
protectedvirtual |
ITK print mechanism
Reimplemented from itk::ProcessObject.
Reimplemented in itk::VideoSource< TOutputVideoStream >, itk::VideoSource< TVideoStream >, itk::VideoSource< itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoToVideoFilter< TVideoStream, TVideoStream >, itk::VideoFileWriter< TInputVideoStream >, itk::VideoFileReader< TOutputVideoStream >, itk::ImageFilterToVideoFilterWrapper< TImageToImageFilter >, itk::FrameDifferenceVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::FrameAverageVideoFilter< TInputVideoStream, TOutputVideoStream >, and itk::DecimateFramesVideoFilter< TVideoStream >.
|
protectedvirtual |
|
protectedvirtual |
|
protectedvirtual |
|
protectedvirtual |
|
protectedvirtual |
Split the output's RequestedTemporalRegion into the proper number of sub-regions. By default it is assumed that each sub-region processed should be generated using the set of input frames starting one frame forward in time from the previous sub-region. To change this, set FrameSkipPerOutput to something other than 1. Positive indicates forward in time while negative indicates backward in time. The set of returned TemporalRegion objects is the set of temporal regions that are needed as input to generate the entire output requested region
|
virtual |
TemporalStreamingGenerateData is in charge of producing output for a single portion of the output requested temporal region. This is where the body of the process will take place. Subclasses that handle spatial data (such as video frames) may instead use this function to split the requested spatial region and process the spatial sub-regions using the mechanism implemented in ImageBase for multithreading.
Reimplemented in itk::VideoFileWriter< TInputVideoStream >, itk::VideoSource< TOutputVideoStream >, itk::VideoSource< TVideoStream >, itk::VideoSource< itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoFileReader< TOutputVideoStream >, and itk::ImageFilterToVideoFilterWrapper< TImageToImageFilter >.
|
virtual |
Override ProcessObject's implementation of UpdateOutputData. This is necessary because by default ProcessObject propagates the call to its inputs before calling GenerateData. This doesn't work here because we need to adjust the requested temporal region of our inputs before calling their UpdateOutputData methods
NOTE: Subclasses that don't use the temporal streaming process must override this function to call the version implemented in ProcessObject
Code: this->ProcessObject::UpdateOutputData( output )
Reimplemented from itk::ProcessObject.
|
virtual |
The default implementation of UpdateOutputInformation to handle temporal regions will compute the proper size of the output largest possible temporal region based on the largest possible temporal region of the input, the unit input/output sizes for the process, and the number of frames skipped per output
Reimplemented from itk::ProcessObject.
Reimplemented in itk::VideoToVideoFilter< TInputVideoStream, TOutputVideoStream >, itk::VideoToVideoFilter< itk::VideoStream< TImageToImageFilter::InputImageType >, itk::VideoStream< TImageToImageFilter::OutputImageType > >, itk::VideoToVideoFilter< TVideoStream, TVideoStream >, and itk::VideoFileReader< TOutputVideoStream >.
|
protected |
Number of frames to move in order to produce each set of output frames. There is no public API for this member, but subclasses can set it internally (or provide access to it) if they wish.
Definition at line 227 of file itkTemporalProcessObject.h.
|
protected |
Member to indicate the location of the "current frame" in the minimum set of input frames. For example, if the unit number of input frames is 6 and m_InputStencilCurrentFrameIndex = 4, this indicates that for output frame n, frames n-4 through n+1 are required, whereas if m_InputStencilCurrentFrameIndex = 0, frames n through n+5 are required.
Definition at line 234 of file itkTemporalProcessObject.h.
|
protected |
Members to indicate the number of input frames and output frames required to perform a single pass through the processing.
Definition at line 215 of file itkTemporalProcessObject.h.
|
protected |
Definition at line 222 of file itkTemporalProcessObject.h.