Package net.preibisch.mvrecon.process.deconvolution.iteration

Interface ComputeBlockThread

All Known Subinterfaces:
ComputeBlockMulThread, ComputeBlockSeqThread
All Known Implementing Classes:
ComputeBlockMulThreadAbstract, ComputeBlockMulThreadCPU, ComputeBlockSeqThreadAbstract, ComputeBlockSeqThreadCPU, ComputeBlockSeqThreadCUDA, ComputeBlockThreadAbstract
public interface ComputeBlockThread
Executes one Lucy-Richardson iteration on one specifc block. When initialized, it is not relevant which block it will be, since it only depends on the block size
Author:
stephan.preibisch@gmx.de

  • Method Details

    • getBlockSize

      int[] getBlockSize()
      Returns:
      the block size in which we process

    • getMinValue

      float getMinValue()
      Returns:
      the minimum value inside the deconvolved image

    • getId

      int getId()
      Returns:
      the unique id of this thread, greater or equal to 0, starting at 0 and increasing by 1 each thread

    • getPsiBlockTmp

      net.imglib2.img.Img<net.imglib2.type.numeric.real.FloatType> getPsiBlockTmp()
      contains the deconvolved image at the current iteration, copied into a block from outside (outofbounds is mirror)
      Returns:
      the Img to use in order to provide the copied psiBlock