Trackmate only running in Java

[mingxxu]

Hi everyone,
Because when I use Trackmate on Fiji, It’s out of memory if I don’t divide the input data to several parts, and I need to combine all output data files together to one. So recently, I am working on Java files that @tinevez provided on Github. The problem is when I run main.java, It’s still running through Fiji. How should I do to make the trackmate running on Java rather than Fiji?

I think Fiji only supports loading all input first and do analyze together, and what I want to achieved on Java is running input a small section by a small section to get rid of memory outage.

Sorry if the question is unclear!

[stelfrich]

Dear @mingxxu,

The problem is when I run main.java, It’s still running through Fiji. How should I do to make the trackmate running on Java rather than Fiji?

I assume that you are executing a class’ main method which will in most TrackMate cases fire up an ImageJ/Fiji instance.

You could check if you still run into the memory issue when you operate on a Virtual Stack. I am not entirely sure if TrackMate triggers loading of a large dataset into memory when the computations are done.

Also, it might be worth to check out:

Tracking of Large Image Data

Hello, I would like to ask which tools you guys know for tracking of nuclei in data that does not fit into the RAM of the computer. Personally, I know the following: Arivis Imaris (I think…) TrackMate, loading the data in ImageJ as VirtualStack However, I tried this once and maybe there is room to optimize the code for disk-resident data? Anything else? All the Best, Tischi

Best,
Stefan