Commit 0af36559 authored by Sebastien Harasse's avatar Sebastien Harasse

DOC: Mean shift. Improved comments.

parent 2febfd3d
......@@ -27,6 +27,14 @@
namespace otb
{
/** \class SpatialRangeJointDomainTransform
*
*
* Functor returning the joint spatial-range representation of a pixel, i.e. the
* concatenation of range components and coordinates as a vector. Components are
* scaled by their respective spatial and range bandwidth.
*/
template<class TInputImage, class TOutputJointImage>
class SpatialRangeJointDomainTransform
{
......@@ -74,36 +82,6 @@ private:
RealType m_RangeBandwidth;
};
/** \class MeanShiftImageFilter2
*
*
* Mean shift is a data clustering algorithm often used in image processing and segmentation.
* For a given pixel, the mean shift will build a set of neighboring pixels within a given spatial
* bandwidth (can be set using SetSpatialBandwidth()) and a spectral range (can be set using SetRangeBandwidth()).
* The spatial and spectral center of this set is then computed and the algorithm iterates with this new spatial
* and spectral center.
*
* Mean shift can be used for edge-preserving smoothing, or for clustering. The GetOutput() method
* return concatenation of spatial and spectral meanshift filtered data GetSpatialOutput() and GetSpectralOutput() gives
* resp. spatial and Spectral filtering parts
*
*
* GetMetricOutput() method gives mean shift vector
* GetIterationOutput() returns the number of iterations performed for each pixel.
*
* For more information on mean shift techniques, one might consider reading the following article:
*
* D. Comaniciu, P. Meer, "Mean Shift: A Robust Approach Toward Feature Space Analysis," IEEE Transactions on
* Pattern Analysis and Machine Intelligence, vol. 24, no. 5, pp. 603-619, May, 2002
* D. Comaniciu, P. Meer, "Robust analysis of feature spaces: color image segmentation," cvpr, p. 750, 1997
* IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'97), 1997
* D. Comaniciu, P. Meer, "Mean Shift Analysis and Applications," iccv, p. 1197, Seventh International Conference
* on Computer Vision (ICCV'99) - Volume 2, 1999
*
* \sa MeanShiftImageFilter
*
* \ingroup ImageEnhancement
*/
class KernelUniform
{
......@@ -143,7 +121,36 @@ class NormL2
{
};
/** \class MeanShiftImageFilter2
*
*
* Mean shift is a data clustering algorithm often used in image processing and segmentation.
* For a given pixel, the mean shift will build a set of neighboring pixels within a given spatial
* bandwidth (can be set using SetSpatialBandwidth()) and a spectral range (can be set using SetRangeBandwidth()).
* The spatial and spectral center of this set is then computed and the algorithm iterates with this new spatial
* and spectral center.
*
* Mean shift can be used for edge-preserving smoothing, or for clustering. The GetOutput() method
* return concatenation of spatial and spectral meanshift filtered data GetSpatialOutput() and GetSpectralOutput() gives
* resp. spatial and Spectral filtering parts
*
*
* GetMetricOutput() method gives mean shift vector
* GetIterationOutput() returns the number of iterations performed for each pixel.
*
* For more information on mean shift techniques, one might consider reading the following article:
*
* D. Comaniciu, P. Meer, "Mean Shift: A Robust Approach Toward Feature Space Analysis," IEEE Transactions on
* Pattern Analysis and Machine Intelligence, vol. 24, no. 5, pp. 603-619, May, 2002
* D. Comaniciu, P. Meer, "Robust analysis of feature spaces: color image segmentation," cvpr, p. 750, 1997
* IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'97), 1997
* D. Comaniciu, P. Meer, "Mean Shift Analysis and Applications," iccv, p. 1197, Seventh International Conference
* on Computer Vision (ICCV'99) - Volume 2, 1999
*
* \sa MeanShiftImageFilter
*
* \ingroup ImageEnhancement
*/
template <class TInputImage, class TOutputImage, class TKernel = KernelUniform, class TNorm = NormL2, class TOutputMetricImage = TOutputImage, class TOutputIterationImage = otb::Image<unsigned int, TInputImage::ImageDimension> >
class ITK_EXPORT MeanShiftImageFilter2
: public itk::ImageToImageFilter<TInputImage, TOutputImage>
......@@ -233,8 +240,6 @@ protected:
virtual void GenerateInputRequestedRegion();
//virtual void EnlargeOutputRequestedRegion( itk::DataObject *output );
virtual void BeforeThreadedGenerateData();
/** MeanShiftFilter can be implemented as a multithreaded filter.
......@@ -289,6 +294,7 @@ private:
KernelType m_SpatialKernel;
KernelType m_RangeKernel;
/** Number of components per pixel in the input image */
unsigned int m_NumberOfComponentsPerPixel;
/** Input data in the joint spatial-range domain, scaled by the bandwidths */
......
......@@ -296,6 +296,10 @@ MeanShiftImageFilter2<TInputImage, TOutputImage, TKernel, TNorm, TOutputMetricIm
m_NumberOfComponentsPerPixel = this->GetInput()->GetNumberOfComponentsPerPixel();
// m_JointImage is the input data expressed in the joint spatial-range
// domain, i.e. spatial coordinates are concatenated to the range values.
// Moreover, pixel components in this image are normalized by their respective
// (spatial or range) bandwith.
typedef SpatialRangeJointDomainTransform<InputImageType, RealVectorImageType> FunctionType;
typedef otb::UnaryFunctorWithIndexWithOutputSizeImageFilter<InputImageType, RealVectorImageType, FunctionType> JointImageFunctorType;
......@@ -577,6 +581,11 @@ MeanShiftImageFilter2<TInputImage, TOutputImage, TKernel, TNorm, TOutputMetricIm
modeCandidate[comp] = jointPixel[comp] * m_SpatialBandwidth + 0.5;
}
// Check status of candidate mode
// If pixel candidate has status 0 (no mode assigned) or 1 (mode assigned)
// but not 2 (pixel in current search path), and pixel has actually moved
// from its initial position, and pixel candidate is inside the output
// region, then perform optimization tasks
if (m_modeTable->GetPixel(modeCandidate) != 2 && modeCandidate != currentIndex && outputRegionForThread.IsInside(modeCandidate))
{
// Obtain the data point to see if it close to jointPixel
......@@ -592,15 +601,17 @@ MeanShiftImageFilter2<TInputImage, TOutputImage, TKernel, TNorm, TOutputMetricIm
if (diff < 0.5) // Spectral value is close enough
{
// if no mode has been associated to the candidate pixel then
// If no mode has been associated to the candidate pixel then
// associate it to the upcoming mode
if( m_modeTable->GetPixel(modeCandidate) == 0)
{
// Add the candidate to the list of pixels that will be assigned the
// finally calculated mode value
pointList[pointCount++] = modeCandidate;
m_modeTable->SetPixel(modeCandidate, 2);
} else // == 1
{
// the candidate pixel has already been assigned to a mode
// The candidate pixel has already been assigned to a mode
// Assign the same value
rangePixel = rangeOutput->GetPixel(modeCandidate);
for (unsigned int comp = 0; comp < m_NumberOfComponentsPerPixel; comp++)
......@@ -608,8 +619,7 @@ MeanShiftImageFilter2<TInputImage, TOutputImage, TKernel, TNorm, TOutputMetricIm
jointPixel[ImageDimension + comp] = rangePixel[comp] / m_RangeBandwidth;
}
// Update the mode table because pixel will be assigned just now
m_modeTable->SetPixel(currentIndex, 2); // Note: in multithreading, = 1 would
// not be safe
m_modeTable->SetPixel(currentIndex, 2);
// bypass further calculation
numBreaks++;
break;
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment