Error in Labkit Command Line: Datatype: Can't convert datatypes Exception

[staghizadeh]

Hello Labkit team,

I’m encountering an error while using Labkit Command Line with custom HDF5 and XML files. The tool works perfectly with the example files (image.xml and image.h5) provided by Labkit, and I can create and use my own classifier on that dataset without issues. However, whenever I save a new HDF5/XML dataset—even if it’s the example image—using ImageJ, I encounter the following error during segmentation:

picocli.CommandLine$ExecutionException: Error while calling command (sc.fiji.labkit.cli.SegmentChunkCommand@b25b095): java.lang.RuntimeException: java.util.concurrent.ExecutionException: ncsa.hdf.hdf5lib.exceptions.HDF5DatatypeInterfaceException: Datatype:Can't convert datatypes ["....\src\H5Tconv."]line 5377 in H5T__conv_ushort_short(): can't handle conversion exception
at picocli.CommandLine.execute(CommandLine.java:1180)
at picocli.CommandLine.access$800(CommandLine.java:141)
at picocli.CommandLine$RunLast.handle(CommandLine.java:1367)
at picocli.CommandLine$RunLast.handle(CommandLine.java:1335)
at picocli.CommandLine$AbstractParseResultHandler.handleParseResult(CommandLine.java:1243)
at picocli.CommandLine.parseWithHandlers(CommandLine.java:1526)
at sc.fiji.labkit.cli.LabkitCommand.parseAndExecuteCommandLine(LabkitCommand.java:84)

Steps to Reproduce:

Open image.h5 and image.xml (the working example files) in ImageJ/Fiji.
Save the image in XML/HDF5 format using BigDataViewer or ImageJ's Save As functionality, creating new .h5 and .xml files.
Run Labkit Command Line on the newly created .h5 and .xml files.
The segmentation command fails with the Datatype:Can't convert datatypes error shown above.

Could you provide guidance on how to properly save compatible .h5 and .xml files for use with Labkit CLI, or recommend any specific data type settings that would prevent this conversion error?

Thank you for your help!