-batch and -macro results differ from ImageJ

[Mark J. Chopping]

Greetings,

I have a problem with -batch and -macro modes: the results are very different to when I run almost identical macro code via the GUI (X11).

Using: ImageJ 1.49v Java 1.6.0_24 (64-bit) on a Virtual Machine: Debian GNU/Linux 8.1 (jessie).  Here's the call (all on one line):

~/ImageJ/jre/bin/java -Xmx2048m -jar ~/ImageJ/ij.jar -ijpath ~/ImageJ --headless -batch CANAPI_BG.ijm  /home/mchoppin/testdata/window_0098.tif:/home/mchoppin/testdata/Log0814.txt > doij3_log.txt

The macro opens a 16-bit TIFF and a settings file specified in the argument; various procedures are followed including particle analysis and a TIFF (drawn on) and Results table are written out. They are both different from those I obtain when running ImageJ in the usual (GUI) way. The log output is redirected and shows no error messages (although it does seem to contain an early instance of the Results table that I don't need). I do call setBatchMode(true); in the macro (redundant?). I have tried with --headless and without: no change.

I have to use -batch as I have a large number of large files to process (on a number of VMs, hopefully).

Any ideas? I can post links to the code if that's helpful.

--Mark


Mark Chopping
Department of Earth and Environmental Studies
Montclair State University
Montclair, NJ 07043

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Mark J. Chopping]

I just tested this out under MacOS 10.10.1 with ImageJ 1.48v (64-bit) with:

java -Xmx2048m -jar /Applications/ImageJ_148/ImageJ64_148.app/Contents/Resources/Java/ij.jar -ijpath /Applications/ImageJ_148 -batch CANAPI_BG.ijm window_0098.tif:Log0814.txt > doij3_Log.txt

...and I get exactly the same (incorrect!) results as on the Debian machine (and similarly, no error messages written to the redirected log output).

I think some function must be behaving differently in -batch and-macro modes but I have no idea which.

--Mark

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Mark J. Chopping]

Apologies: it seems this was down an error in my code -- surprise! My bad.  --Mark

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html