MATLAB Answers

Function 'getdata' from the Image Acquisition Toolbox very slow

6 views (last 30 days)
Tutu
Tutu on 21 Aug 2019
Commented: Tutu on 11 Dec 2019
Hello,
I am using the function 'getdata' from the Image Acquisition Toolbox in a project. However, the data transfer speed is about 1.5 GB/s. It is my understanding that the data are logged in a Buffer localized on the CPU RAM. Hence, when using 'getadata' to bring the data to the CPU RAM as a variable, the speed should live up to the RAM speed - in my case 21,3 GB/s.
How can I increase this speed?
Any help will be appreciated, please.
Thanks

  0 Comments

Sign in to comment.

Answers (1)

Shashwat Bajpai
Shashwat Bajpai on 10 Dec 2019
Edited: Shashwat Bajpai on 10 Dec 2019
Hey,
getdata supports multicore.If you have multiple cores you can try using them to increase the speed.
Please refer the following documentation on getdata for further assistance:
Hope this helps!

  3 Comments

Tutu
Tutu on 10 Dec 2019
Hi Shashwat Bajpai,
Thank you very much for your answer.
Forwarding the structure 'videoinput' to a 'parfeval' execution - with a 'getdata' executed in it - does not seem to be working for me.
Are you rather suggesting that a multiple call of getdata will be automatically executed on different cores? Do I need to extend the number of pools accordingly?
The link you gave to me does not talk about multicore.
Thanks,
Tual
Shashwat Bajpai
Shashwat Bajpai on 11 Dec 2019
I believe you'll need to assign cores explicitly because in multi core computers some cores remain idle to save resources.
Tutu
Tutu on 11 Dec 2019
Great! Could you point toward ressources to dig this matter, and check that this suggestion is working for my issue, please?
Or better, if you could write a bit of code to demonstrate the methodology with 'getdata', I would be very grateful.
Thanks,
Tual

Sign in to comment.

Sign in to answer this question.