Speed Imagej possible?

I have a set of images, which was taken from soil pores with CT technology. about 4 Gb for one soil pore.
When I process it with 3D object, it takes me more 48 hours before it finished, now it is moving on .
my conputer’s parameter is 64Gb RAM and 24 processors.

Is it possible to speed up the process?
Besides, it it possible to auto process the images with command line? As the computer environment is Centos.
Thanks very much for your help.
Best regards

Hi @zhiying_Guo,

To understand better why the processing takes so long, it would be great if you could explain what you are actually doing. Which kinds of techniques do you use during your analysis? Some algorithms are simply inherently slow (computational complexity) but maybe we can suggest alternatives.



1 Like

Hello, @dietzc,

Thanks very much for your prompt response. I am a newer to ImageJ.
I can clarify my workflow. But, If it is not enough, I am sorry that maybe I will go back to learn some basic knowledge.
1 import Image sequence from a file folder
2 Image-Adjust-Brightness/Contrast, with default parater
3 change the preference: set inches to pixel
4 Process -filter- median, change the default value to 3.0
5 Select an area of interest with Rectangle on the menu
6 Plugins-Object Counter3D with all the parameter default
then wait until the statistic table, which is we want.
Image sequence is obtained by scanning one soil pore with CT technology.

The six step takes long time, and we have many sets of image sequence. Can I do it in batch mode?
Thanks for your help

Best regards,
Zhiying Guo

Since the 6th step is crucial eventually the Particle Analyzer of the BoneJ plugin can speed up your process because it is an optimized version of the 3D objects counter:



Hello @Bio7,
Thanks very much, I am trying this improved plugin.
BTW, can I run ImageJ or its plugins with Linux shell?
Best regards,
Zhiying Guo

Yes, you can run ImageJ/FIJI in headless mode, see , e.g.:


Answer from StackOverflow:


Hello @Bio7
Great ! I will explore this mode.
Thanks very much for so much information.

Best regards,
Zhiying Guo