tracking in ch5 with "reuse"

[yincai]

Hi!

I was processing some MD data using 1.5.2 (haven't tried 1.6 because someone in the group has reported memory issue). The data has been first processed once using the complete pipeline and the output was saved in ch5. I have re-processed the data after having changed some parameters (no segmentation parameter) by having the "reuse" on. Then, the tracking idx was not all right any more (at least by reading using MATLAB): instead of having the cumulative increasing object index, it used (I guess) the object index/frame. By restarting cellcognition and off click "reuse", the output went back normal by processing the data completely new.

Greetings from Heidelberg ;)
Yin

[rhoef]

Hi, you did what I would have suggested as a workaround.

Do you know anything about the Memory issue. No one reported it to us. We know that memory consumption is growing during runtime since certain data has to be keept alive for event detection. This causes troubles for images where cells are quite dense and long timelapse movies.

-- Rudi

[yincai]

Hi Rudi!

As I said, I haven't tried 1.6 myself yet and the colleague who did that is currently on holiday. I will forward your message to him. But as I know, he has probably observed what you described. And as he has MD data for 48h, I guess this was the issue for him to keep the processing pace.

Best, Yin