Hi Berkels,

I tried to use your code to simply process my experimental data, it seems the code fails.

log:

reading reference image from: D:/sampledata/sample0000.dm4
reading template image from: D:/sampledata/sample0000.dm4
Created directory results/stage1/
Using templates images ---------------------------------------
D:/sampledata/sample0001.dm4
D:/sampledata/sample0002.dm4
D:/sampledata/sample0003.dm4
D:/sampledata/sample0004.dm4
D:/sampledata/sample0005.dm4
D:/sampledata/sample0006.dm4
D:/sampledata/sample0007.dm4
D:/sampledata/sample0008.dm4
D:/sampledata/sample0009.dm4
D:/sampledata/sample0010.dm4
D:/sampledata/sample0011.dm4
D:/sampledata/sample0012.dm4
D:/sampledata/sample0013.dm4
D:/sampledata/sample0014.dm4

Created directory results/stage1/0/

Registration on level 0 started

The par file content:

templateNamePattern D:/sampledata/sample%04d.dm4
templateNumOffset 0
templateNumStep 1
numTemplates 15
deformationModel 0
#templateSkipNums { 1 9 10 173 }

cropInput 0
cropStartX 512
cropStartY 512
dontResizeOrCropReference 0

preSmoothSigma 1

saveRefAndTempl 0

numExtraStages 2

saveDirectory results/

dontNormalizeInputImages 0
enhanceContrastSaturationPercentage 0.15
normalizeMinToZero 1

lambda weights the deformation regularization term

lambda 200

lambdaFactor scales lambda dependening on the current level: On level d, lambda is multiplied by pow ( lambdaFactor, stopLevel - d )

lambdaFactor 1

maxGDIterations 500
stopEpsilon 1e-6

startLevel 6
stopLevel 10
precisionLevel 10
refineStartLevel 10
refineStopLevel 10

checkboxWidth 16

resizeInput 0

dontAccumulateDeformation 0
reuseStage1Results 1
extraStagesLambdaFactor 0.1
useMedianAsNewTarget 1
calcInverseDeformation 0
skipStage1 0

saveNamedDeformedTemplates 1
saveNamedDeformedTemplatesUsingNearestNeighborInterpolation 1
saveNamedDeformedTemplatesExtendedWithMean 1
saveDeformedTemplates 1
saveNamedDeformedDMXTemplatesAsDMX 1

Hi Dr. Berkels, Thanks for sharing your work. I am wondering if there is any plan to support multithreaded CPU implementation or GPU implementation? Thanks.

Thanks for your good job. However, When I run the code, I meet a problem. I do not know how to solve it.

Here is an example of a C++/CUDA code with python bindings:
https://github.com/prism-em/prismatic
Everything is in the same repository.
I have no idea how it is done.

I currently try to create a conda package:
conda-forge/staged-recipes#11956

To be able to install match-series with all the required dependencies as easy as:

conda install -c conda-forge match-series

So to be able to fix one version of match-series for future reference and to increase the reproducibility it would be great if you could create a Github release.

When I performed the "calculation.run()" on my data, it told me that 'utf-8' codec can't decode byte 0xd7 in position 0: invalid continuation byte.

What causes this issue and how to solve it?