KMeans Thresholding with Gaussian Filter

Short Description

Versions

• Version 2 • Released on: 2015-07-21 06:07:20
  Download
  Description:

  Added documentation

  /*
  import the different Classes needed
  	- FileDialog is used to select file and Loader to load the image
  	- Package filtering is used to create the kernel and perform the convolution
  	- Package thresholder is needed to do the threshold with 
  	  Thresholder and KMeans classes.
  */
  importClass(Packages.plugins.tprovoost.scripteditor.uitools.filedialogs.FileDialog)
  importClass(Packages.icy.file.Loader)
  importClass(Packages.plugins.adufour.filtering.Convolution1D)
  importClass(Packages.plugins.adufour.filtering.Kernels1D)
  importClass(Packages.plugins.adufour.thresholder.Thresholder)
  importClass(Packages.plugins.adufour.thresholder.KMeans)
  importClass(Packages.icy.file.Saver)
  
  /*
   * Request image file and verify if path is not empty
   */
  f = FileDialog.open()
  if (!f.toString().equals("")) {
  	/*
  	 * Load sequence from f and check if seq is valid
  	 */
  	seq = Loader.loadSequence(f, 0, true)
  	if (seq != null) {
  		/*
  		 * Set the kernel of 3x3
  		 */
  		sigma = 3
  		kernelX = Kernels1D.CUSTOM_GAUSSIAN.createGaussianKernel1D(sigma).getData()
  		kernelY = Kernels1D.CUSTOM_GAUSSIAN.createGaussianKernel1D(sigma).getData()
  		/*
  		 * Create a copy of the original sequence and perform convolution
  		 */
  		seqf = seq.getCopy()
  		Convolution1D.convolve(seqf, kernelX, kernelY, null)
  	
  		/*
  		 * Show gaussian result
  		 */
  		gui.addSequence(seqf)
  	
  		/*
  		 * Find threshold using KMeans method and perform binary threshold
  		 */
  		kmeans = KMeans.computeKMeansThresholds(seq, 2, 2, 255)
  		result = Thresholder.threshold(seqf, 2, kmeans, false)
  		result.setName(seq.getName() + " - thresholded")
  	
  		/*
  		 * Show result of binary Threshold
  		 */
  		gui.addSequence(result)
  
  		/*
  		 * Save result to file
  		 */
  		//f = FileDialog.save()
  		//Saver.save(result, f, true, true)
  	}
  }
  

• Version 1 • Released on: 2015-07-21 05:48:57
  Download
  Description:

  initial version

  importClass(Packages.plugins.adufour.filtering.Convolution1D)
  importClass(Packages.plugins.adufour.filtering.Kernels1D)
  importClass(Packages.icy.file.Saver)
  importClass(Packages.icy.file.Loader)
  importClass(Packages.plugins.tprovoost.scripteditor.uitools.filedialogs.FileDialog)
  importClass(Packages.plugins.adufour.thresholder.Thresholder)
  importClass(Packages.plugins.adufour.thresholder.KMeans)
  importClass(Packages.icy.sequence.SequenceUtil)
  f = FileDialog.open()
  seq = Loader.loadSequence(f, 0, true)
  
  sigma = 3
  kernelX = Kernels1D.CUSTOM_GAUSSIAN.createGaussianKernel1D(sigma).getData()
  kernelY = Kernels1D.CUSTOM_GAUSSIAN.createGaussianKernel1D(sigma).getData()
  seqf = seq.getCopy()
  Convolution1D.convolve(seqf, kernelX, kernelY, null)
  
  gui.addSequence(seqf)
  
  kmeans = KMeans.computeKMeansThresholds(seq, 2, 2, 255)
  result = Thresholder.threshold(seqf, 2, kmeans, false)
  result.setName(seq.getName() + " - thresholded")
  
  gui.addSequence(result)
  
  //f = FileDialog.save()
  //Saver.save(result, f, true, true)