ITK/Examples

These are fully independent, compilable examples. There is significant overlap in the examples, but they are each intended to illustrate a different concept and be fully stand alone compilable. Please add examples in your areas of expertise! You can checkout the entire set of examples from this repository:

http://gitorious.org/itkwikiexamples/itkwikiexamples

About the Examples

• Official announcement

Simple Operations

Example Name	ITK Classes Demonstrated	Description
Translate a vector image	TranslationTransform, VectorResampleImageFilter
Translate an image	TranslationTransform, ResampleImageFilter
Get some basic information about a type	NumericTraits
Apply a color map to an image	ScalarToRGBColormapImageFilter
Catch an ITK exception		Try/Catch blocks
Get the points on a Bresenham line between two points	BresenhamLine
Add an offset to a pixel index	Offset
Distance between two points	Point
Create a vector	Vector
Create a covariant vector	CovariantVector
Get the name/type/class of an object		GetNameOfClass()
An object which holds the index of a pixel	Index
An object which holds the size of an image	Size
An object which holds the index (start) and size of a region of an image	ImageRegion
Make part of an image transparent	RGBAPixel	Transparency, RGBA, alpha
Determine if one region is fully inside another region	ImageRegion	Region intersection
Determine the overlap of two regions	ImageRegion	Region intersection
Duplicate an image	ImageDuplicator
Produce an image of noise	RandomImageSource
Create a vector image	VectorImage	An image with an ND vector at each pixel
Set specified pixels to specified values	Image

Input/Output (IO)

Example Name	ITK Classes Demonstrated	Description
Convert a C-style array to an itkImage	ImportImageFilter
Read an image file without knowing its type before hand	ImageFileReader
Write an image	ImageFileWriter
Read an image	ImageFileReader
Write a TIFF image	TIFFImageIO	This is a general demonstration of how to use a specific writer rather than relying on the ImageFileWriter to choose for you.
Display an ITK image	ImageToVTKImageFilter

Image Processing

Example Name	ITK Classes Demonstrated	Description
Normalized correlation	NormalizedCorrelationImageFilter
Multiply every pixel in an image by a constant	MultiplyByConstantImageFilter
Compute the squared difference of corresponding pixels in two images	SquaredDifferenceImageFilter
Add a constant to every pixel without duplicating the image in memory (an accessor)	AddPixelAccessor
RegionalMaximaImageFilter	RegionalMaximaImageFilter
ValuedRegionalMaximaImageFilter	ValuedRegionalMaximaImageFilter
Pixel wise compare two input images and set the output pixel to their max	MaximumImageFilter
Pixel wise compare two input images and set the output pixel to their min	MinimumImageFilter
Binary AND two images	AndImageFilter
Invert an image using the Binary Not operation	BinaryNotImageFilter
Compose a vector image (with 3 compoennts) from three scalar images	Compose3DCovariantVectorImageFilter	Create an RGB image from scalar images. Combine individual channels into a color image.
Extract a component of a vector image	NthElementImageAdaptor
Process the nth component/element of a vector image	NthElementImageAdaptor
Extract a component of a vector image	ImageAdaptorExtractVectorComponent
Label binary regions in an image and get their properties	BinaryImageToShapeLabelMapFilter	Region bounding box, etc.
Convolve an image with a kernel	ConvolutionImageFilter	Convolution.
Assign contiguous labels to connected regions of an image	RelabelComponentImageFilter
Crop an image	CropImageFilter
Compute the absolute value of an image	AbsImageFilter
Invert an image	InvertIntensityImageFilter
Apply a mask to an image	MaskImageFilter
Pass image pixels through a sigmoid function	SigmoidImageFilter	The qualitative description of how Alpha and Beta affect the function from the ITK Software Guide and the associated images would be nice to add to the doxygen.
Threshold an image	ThresholdImageFilter	The result is the original image but with the values below (or above) the threshold "clamped" to an output value.
Threshold an image	BinaryThresholdImageFilter	The result is a binary image (inside the threshold region or outside the threshold region).
Apply a custom operation to each pixel in an image	UnaryFunctorImageFilter	Perform a custom operation on every pixel in an image. This example rotate the vector-valued pixels by 90 degrees.
Find the minimum and maximum value (and the position of the value) in an image	MinimumMaximumImageCalculator
Add two images together	AddImageFilter
Subtract two images	SubtractImageFilter
Paste a part of one image into another image	PasteImageFilter
Multiply two images together	MultiplyImageFilter
Extract a portion of an image (region of interest)	RegionOfInterestImageFilter
Rescale the intensity values of an image to a specified range	RescaleIntensityImageFilter
Label connected components in a binary image	ConnectedComponentImageFilter
Normalize an image	NormalizeImageFilter
Cast an image from one type to another	CastImageFilter
Switch the axes of an image	PermuteAxesImageFilter
Linearly interpolate a position in an image	LinearInterpolateImageFunction
HammingWindowFunction	HammingWindowFunction

Iterating Over (Traversing) An Image

Example Name	ITK Classes Demonstrated	Description
Iterate over a line through an image	LineConstIterator	Walks a Bresenham line through an image
Iterate over the central region (non-boundary) separately from the face-regions (boundary)	ImageBoundaryFacesCalculator
Iterate over a region of an image with a neighborhood (with write access)	NeighborhoodIterator
Iterate over a region of an image with a neighborhood (without write access)	ConstNeighborhoodIterator
Iterate over a region of an image (with write access)	ImageRegionIterator
Iterate over a region of an image (without write access)	ImageRegionConstIterator
Make out of bounds pixels return a constant value	ConstantBoundaryCondition
Iterate over a region of an image with efficient access to the current index (with write access)	ImageRegionIteratorWithIndex
Iterate over a region of an image with efficient access to the current index (without write access)	ImageRegionConstIteratorWithIndex

Image Edges, Gradients, and Derivatives

Example Name	ITK Classes Demonstrated	Description
Compute the derivative of an image in a particular direction	DerivativeImageFilter
Find the gradient magnitude of the image first smoothed with a Gaussian kernel	GradientMagnitudeRecursiveGaussianImageFilter
Find higher derivatives of an image	RecursiveGaussianImageFilter
Extract the boundaries of connected regions in a binary image	BinaryContourImageFilter
Extract the inner and outer boundaries of blobs in a binary image	BinaryContourImageFilter
Compute the gradient magnitude image	GradientMagnitudeImageFilter
Compute the Laplacian of an image	LaplacianRecursiveGaussianImageFilter

Smoothing

Example Name	ITK Classes Demonstrated	Description
Anti alias a binary image	AntiAliasBinaryImageFilter
Mean filter an image	MeanImageFilter	Replace each pixel by the mean of its neighborhood
Median filter an image	MedianImageFilter	Replace each pixel by the median of its neighborhood
Smooth an image with a discrete Gaussian filter	DiscreteGaussianImageFilter
Blur an image	BinomialBlurImageFilter
Bilateral filter an image	BilateralImageFilter	Edge preserving smoothing.
Smooth an image using curvature flow	CurvatureFlowImageFilterType
Gaussian smoothing that works with image adaptors	SmoothingRecursiveGaussianImageFilter

Morphology

Example Name	ITK Classes Demonstrated	Description
Erode a binary image	BinaryErodeImageFilter
Dilate a binary image	BinaryDilateImageFilter
Dilate a grayscale image	GrayscaleDilateImageFilter
Erode a grayscale image	GrayscaleErodeImageFilter
Erode a binary image using a flat (box) structuring element	FlatStructuringElement

Curves

Example Name	ITK Classes Demonstrated	Description
Compute the mean distance between all points of two curves	ContourMeanDistanceImageFilter

Spectral Analysis

Example Name	ITK Classes Demonstrated	Description
Compute the FFT of an image	VnlFFTRealToComplexConjugateImageFilter
Compute the cross-correlation of two images in the Fourier domain	VnlFFTRealToComplexConjugateImageFilterVnlFFTComplexConjugateToRealImageFilter
Convert a real image and an imaginary image to a complex image	RealAndImaginaryToComplexImageFilter

Utilities

Example Name	ITK Classes Demonstrated	Description
Map scalars into a jet colormap	JetColormapFunctor
Monitor a filter	SimpleFilterWatcher	See debug style information.
Time probe	TimeProbe	Compute the time between points in code.
Observe an event	Command
Vector container	VectorContainer
Bounding box	BoundingBox

Statistics

Example Name	ITK Classes Demonstrated	Description
Create a Gaussian distribution	GaussianDistribution
Create a histogram from a list of sample measurements	SampleToHistogramFilter
Create a list of sample measurements	ListSample
Create a list of samples from an image without duplicating the data	ImageToListSampleAdaptor
Create a list of samples with associated class IDs	MembershipSample
2D Gaussian Mixture Model Expectation Maximization	ExpectationMaximizationMixtureModelEstimator	EM
1D KMeans Clustering	KdTreeBasedKMeansClustering
3D KMeans Clustering	KdTreeBasedKMeansClustering
Cluster the pixels in a greyscale image	ScalarImageKmeansImageFilter

Spatial Objects

Example Name	ITK Classes Demonstrated	Description
Convert a spatial object to an image	SpatialObjectToImageFilter
Ellipse	EllipseSpatialObject
Line spatial object	LineSpatialObject, LineSpatialObjectPoint	Specify a piecewise-linear object by specifying points along the line.
Blob	BlobSpatialObject

Inspection

Example Name	ITK Classes Demonstrated	Description
Combine two images by alternating blocks of a checkerboard pattern	CheckerBoardImageFilter

Image Registration

Example Name	ITK Classes Demonstrated	Description
A basic global registration of two images	ImageRegistrationMethod, TranslationTransform	Translation only transform.
A global registration of two images	ImageRegistrationMethod, AffineTransform	Full affine transform.
A global registration of two images	ImageRegistrationMethod, BSplineDeformableTransform	BSpline transform.
Mutual Information	MutualInformationImageToImageMetric, TranslationTransform	Global registration by maximizing the mutual information and using a translation only transform
Mutual Information Affine	MutualInformationImageToImageMetric, AffineTransform	Global registration by maximizing the mutual information and using an affine transform

Image Segmentation

Example Name	ITK Classes Demonstrated	Description
Mean shift clustering
KMeans Clustering
Multiphase Chan And Vese Sparse Field Level Set Segmentation	itkScalarChanAndVeseSparseLevelSetImageFilter, itkScalarChanAndVeseLevelSetFunction

Meshes

Example Name	ITK Classes Demonstrated	Description
Decimation
Add points and edges
Compute normals of a mesh	QuadEdgeMeshNormalFilter

Wish List

Example Name	ITK Classes Demonstrated	Description
Color Normalized Correlation	NormalizedCorrelationImageFilter
ContourSpatialObject	ContourSpatialObject
IntensityWindowingImageFilter	IntensityWindowingImageFilter
Store non-pixel associated data in an image	MetaDataDictionary
Level Sets
Region Growing
Mesh subdivision
Separate foreground and background using Otsu's method	OtsuThresholdImageFilter
Extract contours	SimpleContourExtractorImageFilter
Present an image of RGBPixel pixels as an image of vectors	RGBToVectorImageAdaptor

Included in the ITK Repository

Example Name	ITK Classes Demonstrated	Description
Image registration

Matlab

Example Name	ITK Classes Demonstrated	Description
Write data from Matlab in a format readable by ITK
Write data from ITK in a format readable by Matlab

Developer Examples

Example Name	ITK Classes Demonstrated	Description
Produce an image programmatically.	ImageSource	Nothing in, image out.
Filter an image	ImageToImageFilter	Image in, same type of image out.
Write a filter with multiple inputs of the same type.	ImageToImageFilter	Two images in (same type), same type of image out.
Write a filter with multiple inputs of different types.	ImageToImageFilter	Two images in (different type), image (same type as first input) out.
Write a filter with multiple outputs of the same type.	ImageToImageFilter	Images in, two images out (same type as first input).

Problems

Small Problems

Example Name	ITK Classes Demonstrated	Description
Compute the gradient of a vector image	GradientImageFilter	How to do this?
Compute the Laplacian of an image	GaussianMixtureModelComponent	Input image type must be double or float (probably should smooth before this anyway, so do this type conversion in the smoothing filter?
Compute distributions of image pixels using GMM EM	GaussianMixtureModelComponent
Compute distributions of image pixels using GMM EM	ExpectationMaximizationMixtureModelEstimator
Compute distributions of samples using GMM EM	ExpectationMaximizationMixtureModelEstimator	Someone please confirm that this outputs the mean and the variance (i.e. I used a standard deviation of 30 to create the samples and the second estimated parameter is near 1000 (~30^2) . Is this correct?)
Find edges in an image	CannyEdgeDetectionImageFilter	How to set a reasonable Threshold for the output edges?
Variable length vector	VariableLengthVector	main() works properly, but neither conversion seems to work.
Compute the histogram of an image	Statistics_1_1ImageToHistogramFilter	The last entry of the red histogram should contain several values, but it is 0?
Classifying pixels using KMeans	KmeansModelEstimator	How to apply the labels of the filter to the input image?
Basic region growing	RegionGrowImageFilter	Just getting started with demo...
Find connected components in an image	ConnectedThresholdImageFilter	Just need to finish it.
Convert an image from one type to another	ConvertPixelBuffer
Apply a function to every pixel with access to its neighborhood	NeighborhoodOperatorImageFilter
Apply a function to every pixel with access to its neighborhood	NeighborhoodOperatorImageFunction
In-place filtering of an image	InPlaceImageFilter	This only works for filters which derive from itkInPlaceImageFilter
Pad an image with a constant value	ConstantPadImageFilter	Not sure how to specify the final output size/extent. The documentation should be improved.
Pad an image	MirrorPadImageFilter	Not sure how to specify the final output size/extent. The documentation should be improved.
Pad an image by wrapping	WrapPadImageFilter	Not sure how to specify the final output size/extent. The documentation should be improved.
Shrink an image	ShrinkImageFilter	Image doesn't seem to be shrunk?
Flip an image over specified axes	FlipImageFilter	The axes specification is not quite right - warning: address of axes always evaluates true
Convert a VTK image to an ITK image	VTKImageToImageFilter	Seems to expect an input image with only 1 component? (i.e. greyscale)
Iterate over a region of an image with a shaped neighborhood	ShapedNeighborhoodIterator	Crashes when initializing the iterator loop?

Big Problems

Example Name	ITK Classes Demonstrated	Description
Join images, stacking their components	JoinImageFilter	How to convert the vector image produced by the joinFilter to a standard image?
Find the best position of the moving image in the fixed image.	MeanSquaresImageToImageMetric	Output (0,0) is incorrect.
Compute and display the gradient of an image	GradientImageFilter	Blank output.