Question about running ImageJ on a multi-CPU workstation....

[Christopher Yip]

Just a quick question that might be naive but

....

How does Image J run on multi-CPU machines? I noticed that on a dual-
Opteron SUSE10 machine, top only seems to show one CPU being tasked  
whenever an ImageJ task is executed.... (BTW - this really isn't that  
big an issue - the program runs great on this machine - I was just  
curious)

Thanks

Chris

[Wayne Rasband]

> How does Image J run on multi-CPU machines? I noticed that on a  
> dual-Opteron SUSE10 machine, top only seems to show one CPU being  
> tasked whenever an ImageJ task is executed.... (BTW - this really  
> isn't that big an issue - the program runs great on this machine -  
> I was just curious)

Each ImageJ command, plugin or macro runs on a separate thread. So  
try starting a second task, operating on a different image, and see  
if both CPUs are used. Also, processing and image display run on  
separate threads, so try running a task that continuously displays  
images and monitor the CPU usage. Such tasks include live video  
capture, stack animation, the built in benchmark  
(Plugins>Utilities>Benchmark), or any of Kai Barthel's plugins (3D  
Color Inspector, 3D Surface Plot, Volume Viewer).

-wayne

[Paul Grimm]

Are there any plugins or implementations of Imagej to control a z  
motor to software autofocus a microscope?
--

Paul C. Grimm
Professor of Clinical Pediatrics
Pediatric Nephrology
University of California at San Diego

Email [hidden email]
Phone [phone]
Fax   [phone]
Snail mail
UCSD Pediatrics
Mail Stop 0831
9500 Gilman Drive,
La Jolla California
92093-0831



"A conclusion is the place where you got tired of thinking"
Steven Wright

[christophe leterrier]

I'll be very interested to know that, too. I heard they were some serial
com plugins to control XY (possibly Z) stages, but I would be very
interested to hear about implementing an autofocus function in ImageJ

Christophe Leterrier

PhD
Neuronal Receptors Dynamics / Zsolt Lenkei
Neurobiology Lab / Jean Rossier
ESPCI-CNRS UMR 7637
10 rue Vauquelin 75231 Paris Cedex 05 France
tel [phone]
fax [phone]
[hidden email]

Paul Grimm a écrit :

> Are there any plugins or implementations of Imagej to control a z
> motor to software autofocus a microscope?
> --
> Paul C. Grimm
> Professor of Clinical Pediatrics
> Pediatric Nephrology
> University of California at San Diego
>
> Email [hidden email]
> Phone [phone]
> Fax   [phone]
> Snail mail
> UCSD Pediatrics
> Mail Stop 0831
> 9500 Gilman Drive,
> La Jolla California
> 92093-0831
>
>
>
> "A conclusion is the place where you got tired of thinking"
> Steven Wright

[Weller Andrew Francis]

Paul and Christophe,

I am currently working with the 'IJSerial v1.0b' plugin for serial port
control of a microscope (including xyz stage) from
http://mitel.dimi.uniud.it/eslide/sw.php. It is a relatively easy task to
send commands through the serial interface to control whatever you want to
control that accept serial commands!

In my example, I control many aspects of the microscope through 5 character
strings. For example, "50010", etc. If the stage you would like to use
accepts similar commands (I know this is the case for Marzhauser and Prior)
then this should not be a problem.

Something that I have been thinking about lately is implementing an autofocus
procedure. In my current macro, it starts by initialising many aspects of the
microscope (upper and lower Z thresholds so to not break any slides, etc).
Something else that can be initialised is a particular Z value that you know
corresponds to the 'near enough' in-focus plane. Once the stage has moved to
a new xy position, you can then run a plugin that tests the sharpness (focal
quality) of the image from this in-focus z-plane and either move the stage up
or down to compensate until you have the optimum plane. (I've used a highpass
filter before to test how many sharp pixels there are.)

Anyway, some rambling thoughts as much as anything. I hope it helps?!

Cheers, Andy

--
Dr. Andy Weller
ETH Zurich
Geological Institute
CAB E 64
Universitaetstrasse 6
8092 Zurich, Switzerland

[hidden email]
www.erdw.ethz.ch/Weller

[phone] office
[phone] lab
[phone] mobile
[phone] fax

[Vincenzo Della Mea]

Not directly in ImageJ, however we are finishing a system for whole  
slide acquisition which includes also an autofocus module (described  
in Della Mea V, Viel F, Beltrami CA. A pixel-based autofocusing  
technique for digital histologic and cytologic slides. Computerized  
Medical Imaging and Graphics 2005; 29(5): 333-341.).
It is somewhat easy to do something similar: you need a focus measure  
function (e.g., standard deviation), and some Z-space visit algorithm  
(e.g., linear, or dichotomic). If you do not mind speed, the best  
thing is to scan the z space at the minimum step possible, and keep  
the image with greatest grey level standard deviation. Some more  
adequate focus measures should be used for fluorescent images.

Vincenzo

Il giorno 16/mag/06, alle ore 15:36, Christophe Leterrier ha scritto:

> I'll be very interested to know that, too. I heard they were some  
> serial com plugins to control XY (possibly Z) stages, but I would  
> be very interested to hear about implementing an autofocus function  
> in ImageJ
>
> Christophe Leterrier

[Manuel Palacios]

Dear All,

I am working in the development of colorimetric assays for inorganics  
in water. The output of the chip-scanner looks pretty much like a dot-
blot assay. I need to write a macro to read the RGB value in each  
spot and report it in a table. So far, I've been doing this manually,  
but each chip has 80 dots and it is to much work when you have 120  
chips to analyze. My question will be: is there any plug-in or macro  
for automatically read micro-arrays and if there is not. Can I make  
and ROI template with 80 selection of the same size that could  
produce an output for each different selection.

Please help me!!!!!!!!!!!!!!! I am a chemist struggling.

Thanks in advance

Best Regards.

Manuel Palacios

---------------------------------------------------------------
Lic. Manuel Alfredo Palacios H.
Doctoral Fellow
[hidden email]

Center for Photochemical Sciences
Bowling Green State University
Bowling Green, OH, 43403
Phone: [phone] (Lab)
http://personal.bgsu.edu/~manuelp

"In theory, there's no difference between theory and practice, but in  
practice there is"
Jan Van Snepscheut
---------------------------------------------------------------