Machine Vision

取消
显示结果 
搜索替代 
您的意思是: 

ADV Toolkit (Imageprocessing on NVidia GPU)

 


@rfdfdgf wrote:

 

I want to congratulate you for this great toolkit, is exactly what I needed.

I have just bought an GTS450 videocard to use it in my image processing projects.

I have checked the NVidia website for how many Processor Cores this videocard has, it says 192.

But when I try you're example it says number of streaming processors = 4 !!!

 

 


Hi, Andrej,

 

Its seems to be OK. You probably confused SMs (streaming multiprocessors) with SPs (shader processors or CUDA Cores). As far as I know, your GTS 450 based on GF 106 architecture and on this architecture each streaming multiprocessor contains 48 SPs. So, 192/48 = 4. SMs are really the basic execution unit in NVIDIA hardware.

 

Hope it helps for understanding,

 

Andrey.

 

11 条消息(共 34 条)
5,168 次查看

 

Andrey I must confess you that I simply love a straight answer.

 

Thank you very much.

 

I hope the shop will accept the return in exchange of a Quadro FX380 video card 🙂

 

If you're kind how many SMs and SPs have the 9600 video card? Where can I find all this data about CUDA video cards?

 

Once again thank you for the great job.

 

Best Regards,

  

Andrei 

 

0 项奖励
12 条消息(共 34 条)
5,165 次查看

 


@rfdfdgf wrote:

 

Andrey I must confess you that I simply love a straight answer.

 

Thank you very much.

 

I hope the shop will accept the return in exchange of a Quadro FX380 video card 🙂

 

If you're kind how many SMs and SPs have the 9600 video card? Where can I find all this data about CUDA video cards?

 

Once again thank you for the great job.



 

On all cards other than Fermi, a streaming multiprocessor is a cluster of 8 scalar cores. Therefore I have 8 SMs (and totally 64 SPs). But from performance point of view GTS 450 should be much faster than 9600 GT (even with less SMs).

 

You can get tech specs from nvidia web site (search for each card), or from wikipedia:

 

http://en.wikipedia.org/wiki/GeForce_400_Series

http://en.wikipedia.org/wiki/Nvidia_Quadro

http://en.wikipedia.org/wiki/GeForce_9_Series

http://en.wikipedia.org/wiki/CUDA

 

Andrey.

 

13 条消息(共 34 条)
5,162 次查看

Hello everyone,

                         I am facing an issue when I am trying to build an executable with the toolkit on my LabVIEW 2010. It builds the executable successfully, but the only problem being when I run the exe. The VI runs perfectly fine with now warnings/ errors. But the exe throws out dll's missing errors namely freeimage.dll and advision.dll missing. I was able to find the feeimage.dll but unable to find the later one. Could someone help me out with this if there is a workaround for the same and if any one has encountered the same problem?

 

Thanking you in anticipation.

 

Regards,

Shreyas H

Shreyas Hebbare
Shreyas Technologies
India
0 项奖励
14 条消息(共 34 条)
4,962 次查看

I got your code to work! Finally a simple way of using IMAQ with the GPU. If you could post the source code to your DLL files, that would help me a lot in my goal to start writing my own image processing filters.

 

Cheers,

 

Bob

0 项奖励
15 条消息(共 34 条)
4,895 次查看

Is it possible that this toolset could be extended to stream to video formats like avi?  One of the great bottlenecks I have found

in IMAQ is the performance limitations of the AVI Write frame.

0 项奖励
16 条消息(共 34 条)
4,861 次查看

Yes, there als AVI functions based on OpenCV VideoWriter, which can be better in some cases.

You must know FourCC code for your codec.

For example, writing video with xvid codec looks like that:

 

Vid-Snippet.png

 

(Some users reported that this code caused troubles on exit)

0 项奖励
17 条消息(共 34 条)
4,850 次查看

Great Job,

How about missing AVI libraries? It is possible I might missing someting...

Looking forward to test it with high speed GigE cams.

-Artur

0 项奖励
18 条消息(共 34 条)
4,679 次查看

AVI's library identified,

No more help required, thank you

-Artur

0 项奖励
19 条消息(共 34 条)
4,661 次查看

Congratulation for this toolkit! This is exactly what I needed and it's, in my opinion, easier to use than the GPU analysis toolkit!

 

Any chance you've been updating it since 2010? 

 

regards,

0 项奖励
20 条消息(共 34 条)
4,619 次查看