GigE Vision cameras in Matlab

[Garry L]

We have several GigE Vision cameras that I am trying to interface to Matlab using the Image Acquisition Toolbox's built-in GigE Vision adapter. The cameras are different models from JAI, Baumer, Basler and Toshiba Teli. So far I have been successful only with JAI. A Baumer camera is not recognized at all. Basler and Teli are recognized, but the video acquisition is very unstable. It would stop in a minute or two by itself. A number of camera parameters can't be adjusted. Also it is impossible to run 2 or more cameras at the same time as the CPU load reaches 100% and the frame rate drops down to 1-2 fps.

I tried several computers with different OS (XP, Windows 7 32-bit, Windows 7 64-bit) with the same result in Matlab. Yet the cameras are running with no issues when using the Vision Acquisition Software from National Instruments.

Has anybody experienced similar problems? Does Mathworks have a list of GigE cameras that have been tested with the IMAQ Toolbox?

Thanks,
Garry

[ImageAnalyst]

Garry:
Often the business of writing "adapters" that allow a camera to work
with MATLAB's Image Acquisition Toolbox is done by the camera
manufacturer. I'd look to them if it's not listed on the Mathworks
web site or in the toolbox documentation.

[Mark Jones]

Hi Garry,

I suspect that your issues may be resolved by installing one of our
patches. The patches are cumulative, so if you are viewing two bug
reports and each has the same files attached, your decision as to which
patches should be installed is quite easy.

Based on what you describe, I think that the following Bug Report
applies to you:

http://www.mathworks.com/support/bugreports/687120

If you want the absolute latest files, you can download the files
attached to:

http://www.mathworks.com/support/bugreports/680567

We have had success using at least one camera from all of the
manufacturers that you have listed, and with several of these vendors we
have had success with several cameras. When we have not had success, we
have investigated the issue and released a patch for the issue.

http://tinyurl.com/4vbopo9

Also, I encourage you to contact technical support when you run into an
issue such as you describe.

http://www.mathworks.com/support/contact_us/index.html

If the above patches do not resolve your issue, please be sure to
contact technical support so that we can work to resolve your issue.

Mark

Mark Jones
Senior Software Engineer
Test and Measurement - Image Acquisition Toolbox
-
MathWorks
3 Apple Hill Drive
Natick, MA 01760 USA
www.mathworks.com

[Mark Jones]

Hi Garry,

http://www.mathworks.com/support/bugreports/687120

http://www.mathworks.com/support/bugreports/680567

http://tinyurl.com/4vbopo9

http://www.mathworks.com/support/contact_us/index.html

Mark

[Garry L]

Hi Mark,

Applying the patches made the Baumer camera recognizable, but the acquisition issues are still there. Have you tested with high-speed cameras? The ones we are using are 200+ fps and we are working in 10-bit modes. The problem is that the CPU load on one core is close to 100%, this is before I apply any processing. I think this is what causing the stability issue. When I try to use two cameras, everything dies.

Again, the NI software works fine even when 3 cameras are running at the same time, so it is not a hardware/system issue.

Thanks,
Garry

Mark Jones <mark....@mathworks.com> wrote in message <4D3452EB...@mathworks.com>...

[Boris N]

Garry,

As far as I know, Matlab's built-in GigE Vision adapter does not contain a system filter driver. Without the filter driver the CPU load gets extensively high, especially when multiple cameras are used. This causes missing frames, streaming stability issues and other problems.

Our ActiveGigE SDK provides a complete interface between Matlab and GigE Vision cameras, and it does come with the filter driver. The ActiveGigE interface is very efficient and easy to use. Unlike Matlab&#8217;s GigE adapter, it does not require purchasing the Image Acquisition Toolbox - the interfacing is done via an ActiveX object directly to your Matlab application. ActiveGigE setup contains several sample projects for Matalb, including Matlab GigE Viewer:
http://www.ab-soft.com/images/matlab.jpg

More information on ActiveGigE and trial version can be found here:
http://www.ab-soft.com/activegige.php

Regards,

Boris
A&B Software

[Garry L]

Boris,

Thank you for pointing me to your SDK. It worked like a charm. Our Matlab application is now running 4 cameras in parallel while doing some image processing in real time - without dropping any frames. I also liked the programming interface. Very easy and intuitive in comparison to what I tried so far.

Garry

[Gino]

Hello Mark,

Can you teach me how to make sure the frame rate of the gige camera?
I use the following two lines to check the constraint of frame rate and change it to the highest one.
FPS_gige = propinfo(src_obj_gige,'AcquisitionFrameRateAbs')
src_obj_gige.AcquisitionFrameRateAbs = FPS_gige.ConstraintValue(2)
And I use TIC & TOC to check the time for couple of snapshot, but it's not corresponding with what I set;(
tic
for i=1:1:90
getsnapshot(obj_gige);
end
T_gige_1514=toc
By the way, my camera is Allied Vision Technologies GE1050C; the specification says the maximum frame rate is 60.

THX:)

Mark Jones <mark....@mathworks.com> wrote in message <4D3452EB...@mathworks.com>...