Trackmate SpotsID initial value

[manerotoni]

I encountered following issues with the numbering of spotsID

• spotsID on the first frame do not start with ID0001. This is true for different detectors.
• The value of the smallest spotsID incrementally increases when trackmate is restarted. The value increases by the number of spots found in the previous round.
• For many spots we easily reach the 16bit limit in case we need to generate a labeled image as output and do not want to restart imageJ every time.
• Export the label image creates labels with an offset of 1. SpotID 2000 has a label 2001. This is a little confusing and not really necessary.

See also entry in forum.sc https://forum.image.sc/t/trackmate-spotsid-initial-value/76753

[tinevez]

Hello @manerotoni
Yes, this is the expected behavior. The ID will increase continuously within a Fiji session.
Also the label ID is also the expected behavior. We want to reserve the label 0 for the background, even if a spot has an ID of 0.
I do not know how to solve the issue with 16-bit limit....

[manerotoni]

Thanks, I guess we need to restart Fiji at every round, but still use this very nice and useful tool.

Also the label ID is also the expected behavior. We want to reserve the label 0 for the background, even if a spot has an ID of 0.

I understand this, but the initial spotsID after a fresh restart in Fiji starts with >=2000. This is what I meant "unnecessary" as we are already far above 0.

[tinevez]

Hello @manerotoni
Is this issue still valid? Can I close it?

[manerotoni]

Hello @tinevez,
thanks for asking. In my opinion the issue is still open (trackmate version 7.10.2). I am not sure if I am the only one that has a problem with it.

Using the example data I get a starting value of a spot ID of 2048 ending at 2207. Starting again trackmate gives an initial spot ID of 2439.

If you run trackmate via a script on several movies or for long movies you quickly reaches a point where the maximal number of spots ID in a 16bit labeled image is reached and you get non-sense results. Also the fact that the counting of the first spot does not starts at 1 but with a rather high and not so reproducible value is an issue when porting the labels to another software.