Add Hamamatsu DCIMG Reader

[zacsimile]

I found myself in the fortunate position of regularly needing to investigate bead stacks acquired as .dcimg files. My goal was to load them and reslice in xz. This seemed easiest to do in ImageJ, so I ported a merge of some existing DCIMG readers to Java for use as a local tool. The result is presented here, in case there is broader interest in the reader.

There are a few things to note:

1. On a Mac (works fine on Windows), while the stacks load fine, reslicing does not work on the loaded stack. I receive the error

java.lang.IllegalArgumentException: Invalid bitDepth: 0
  at ij.gui.NewImage.createImage(NewImage.java:397)
  at ij.plugin.Slicer.createOutputStack(Slicer.java:462)
  at ij.plugin.Slicer.resliceRectOrLine(Slicer.java:442)
  at ij.plugin.Slicer.reslice(Slicer.java:115)
  at ij.plugin.Slicer.run(Slicer.java:80)
  at ij.IJ.runPlugIn(IJ.java:216)
  at ij.Executer.runCommand(Executer.java:152)
  at ij.Executer.run(Executer.java:70)
  at java.lang.Thread.run(Thread.java:750)

If I duplicate the stack, reslicing works fine. Does anyone know what is likely to cause such an error just on a Mac?

2. I’ve made a design choice, where multiple frames within a single DCIMG file are loaded in T and multiple DCIMG files are loaded along Z.

3. I’ve done the bare minimum on loading metadata.

4. I have only tested the 4 pixel correction on older DCIMG 0x7. I don’t have any examples of DCIMG 0x1000000 or 0x2000000 that need the 4 pixel correction, though I would be happy to take a look at one.

[melissalinkert]

Thank you, @zacsimile.

Before we proceed with review, there are two things we will need:

• a signed Contributor License Agreement (see https://ome-contributing.readthedocs.io/en/latest/cla.html)
• example datasets we can use to test this reader (uploaded to https://zenodo.org/communities/bio-formats)

Once we have those two things, we'll assign reviewers to this pull request. Please note we'll do our best to review in a timely fashion, but can't guarantee a timeline at this point. Please see https://forum.image.sc/t/bio-formats-update/101734 for more information about the current state of Bio-Formats maintenance.

1. On a Mac (works fine on Windows), while the stacks load fine, reslicing does not work on the loaded stack. I receive the error

java.lang.IllegalArgumentException: Invalid bitDepth: 0
  at ij.gui.NewImage.createImage(NewImage.java:397)
  at ij.plugin.Slicer.createOutputStack(Slicer.java:462)
  at ij.plugin.Slicer.resliceRectOrLine(Slicer.java:442)
  at ij.plugin.Slicer.reslice(Slicer.java:115)
  at ij.plugin.Slicer.run(Slicer.java:80)
  at ij.IJ.runPlugIn(IJ.java:216)
  at ij.Executer.runCommand(Executer.java:152)
  at ij.Executer.run(Executer.java:70)
  at java.lang.Thread.run(Thread.java:750)

If I duplicate the stack, reslicing works fine. Does anyone know what is likely to cause such an error just on a Mac?

I don't have any immediate ideas, but once we have example data I expect we would try to reproduce this error as part of the review process.

[melissalinkert]

Apologies, it looks like we do already have a signed CLA. So we'll just need example data in order to proceed here.

[zacsimile]

No worries! I apologize, after reading that forum post, it seems this is exactly the kind of PR you are not looking for at the moment. Thanks for having a look anyway, and please take your time. I have compiled a local version of bioformats that works for me at the moment.

I will upload some data ASAP. I just have to email a few people because a couple of the files I tested on are not mine and I'm not sure if uploading them is OK. In this case, I will have to record a couple more files in different DCIMG formats.

[zacsimile]

Please find examples for DCIMG 0x7 and 0x1000000 at https://zenodo.org/records/14281237 and https://zenodo.org/records/14268554, respectively. The first one is a stack with multiple frames in a single DCIMG file and the second is a sequence of single frames, both monochromatic. I still do not have a DCIMG version 0x2000000 or a DCIMG with multiple color channels, but I do not claim to be able to load these with the current code. Happy to update it later if someone provides one or both of these examples.

[zacsimile]

Fixed 4-pixel correction for DCIMG 0x1000000 upon receiving example at https://zenodo.org/records/14287640.

[melissalinkert]

Thanks for providing test data, @zacsimile. I've done a first round of code review - let us know when you have had a chance to address all of the comments, and then we can make a plan for further testing.

[zacsimile]

Thanks so much for the feedback! Addressed.

[melissalinkert]

https://merge-ci.openmicroscopy.org/jenkins/job/BIOFORMATS-test-repo/293/ continues to pass with this pull request included, and the last set of changes looks good to me.

Propose to get this in for 8.1.0, but will hold off on merging until the new year in case any final comments.