topic Quote:samontab wrote: in Software Archive

sm.close() does not close successfully

Xin_X_2 — Fri, 25 Sep 2015 03:19:38 GMT

Hi,guys. It is me again ><

Here is the SDK example.

pxcmStatus OnNewSample(int mid, PXCMCapture.Sample sample) 

{

   // return NO ERROR to continue, 

   // or any ERROR to exit the loop

   return pxcmStatus.PXCM_STATUS_NO_ERROR;

}

 

 

...

// Initialize my event handler

PXCMSenseManager::Handler handler=

       new PXCMSenseManager.Handler();

handler.onNewSample=OnNewSample;

 

// Stream depth samples

sm.Init(handler);

sm.StreamFrames(true);

sm.Close();

And, I know sm.StreamFrames(true); will open a new thread to work.

But, what does sm.close() mean? Just to Close sm.StreamFrames? in other word, that is to close this created thread?

But in my simple WPF example, i add a new button to do sm.close(), the result is that the program crashed, and UI "frozen", i even cant close the program...

So, why ?

And how to use sm.close() correctly?

Thank you very much!

sm.Init() should be paired

samontab — Fri, 25 Sep 2015 04:02:47 GMT

sm.Init() should be paired with sm.Release() instead of sm.Close()

Quote:samontab wrote:

Xin_X_2 — Fri, 25 Sep 2015 12:29:56 GMT

samontab wrote:

sm.Init() should be paired with sm.Release() instead of sm.Close()

Thank you very much for your help.

But, in C#， i did not find sm.Release(), while there is sm.Releaseframe(). But i do not think they are same.

By the way, how do you close the thread? Or if you use callback way to get the sample from camera, how do you stop camera?

{ My poor English may be more difficult to understanding than algorithm...LOL }

In C++ this is how it works:

samontab — Fri, 25 Sep 2015 13:51:14 GMT

In C++ this is how it works:

You need to first create an instance of the session with PXCSenseManager::CreateInstance();

The resources of this session need to be released with PXCSenseManager::Release();

In between those calls, and after you call Init(), you can access the frames by pairing AcquireFrame/ReleaseFrame calls.

If, instead of calling them you use StreamFrames(), then you need to call Close() to terminate that thread.

Now, in C# everything should be the same, except that Release() is called Dispose();

Quote:samontab wrote:

Xin_X_2 — Fri, 25 Sep 2015 14:15:57 GMT

samontab wrote:

In C++ this is how it works:

You need to first create an instance of the session with PXCSenseManager::CreateInstance();

The resources of this session need to be released with PXCSenseManager::Release();

In between those calls, and after you call Init(), you can access the frames by pairing AcquireFrame/ReleaseFrame calls.

If, instead of calling them you use StreamFrames(), then you need to call Close() to terminate that thread.

Now, in C# everything should be the same, except that Release() is called Dispose();

Yes, i read the same things from the SDK. And now i am sure that the function Close() is to kill/terminate the thread which is created by StreamFrames().

Just in my simple demo, I use one button to start the camera, it works and gets the sample from the camera; therefore i want to stop this camera by adding a new button Button2. When clicking Button2, it just run the sm.Close().

However, ALL the UI are frozen...

My program is very like it http://www.i-programmer.info/programming/hardware/8496-real-realsense-in-c-event-streams.html

Also in this article, the author said

How to stop the streaming?

Simple just add another button and call the SenseManager's Close method. To make this work we need to keep a reference to the SenseManager.

private void Button_Click_1(object sender, RoutedEventArgs e) { sm.Close(); }

So i am very confused, as even i copy the same code, my UI is also frozen...

Thank you so much for your value time and patience.

When you work with a GUI, you

samontab — Fri, 25 Sep 2015 14:54:16 GMT

When you work with a GUI, you need to make sure that the main thread remains as responsive as possible.

So, when you click the button, you could for example set a variable in another thread that eventually will close the camera. It is a bad idea to communicate directly with the camera from a GUI thread. That's probably why you see that problem.

I recommend you to read more about GUI programming. You can also start with a command line program to make it easier, and then build a GUI on top of your core program. It's always good to have an independent GUI anyway.