如何在使用OpenCV的Kinect传感器获得的RGB视频中使用Frame-Difference？ [英] How to use Frame-Difference in the RGB video that got from the Kinect sensor with OpenCV?

问题描述

我想使用来自Kinect传感器的RGB视频中的Frame-Difference来实现前景分割

I wanna realize the foreground sementation using Frame-Difference in the RGB video that get from Kinect sensor

但是我对如何获得第1帧感到困惑，第2帧框架和.......

but I am confused with how to get the 1st frame ,2nd frame ,and.......

这里是实现获取RGB视频的主要代码。 

here is the  main code that implement got the RGB video. 

/ ************************************** ****** /

/********************************************/

WaitForSingleObject（hEvents [0]，INFINITE）;

WaitForSingleObject( hEvents[0], INFINITE);

   const NUI_IMAGE_FRAME * RGBFrame; 

   NuiImageStreamGetNextFrame（hStreams [0]，NULL和  RGBFrame）; 

  const NUI_IMAGE_FRAME *RGBFrame; 
  NuiImageStreamGetNextFrame( hStreams[0], NULL, &RGBFrame ); 

   NUI_LOCKED_RECT lockedRGB;

   RGBFrame-> pFrameTexture-> LockRect（NULL，& lockedRGB，NULL，NULL）;  

  NUI_LOCKED_RECT lockedRGB;
  RGBFrame->pFrameTexture->LockRect( NULL, &lockedRGB, NULL, NULL );  

   cvSetData（ cvRGBImage，（unsigned char *）lockedRGB.pBits，lockedRGB.Pitch）; 

  cvSetData( cvRGBImage, (unsigned char*)lockedRGB.pBits, lockedRGB.Pitch ); 

   cvShowImage（" RGB"，cvRGBImage）; 

  cvShowImage( "RGB", cvRGBImage ); 

  

   if（cvWaitKey（1）=='x'）{break; }  

  
  if ( cvWaitKey( 1 ) == 'x' ){ break; }  

   NuiImageStreamReleaseFrame（hStreams [0]，RGBFrame）;

  NuiImageStreamReleaseFrame( hStreams[0], RGBFrame );

/ ******** ************************************ /

/********************************************/

任何人都可以给我一些关于怎么做的建议吗？

Can anyone give me some advises about how to do it ?

（我用Opencv来做Frame-Difference）我只是无法想象如何逐个挑选框架并处理它们

(I am using Opencv to do the Frame-Difference) I just can't figure out how to pick out the frames one by one, and process them

plz help ~~~ thx~

plz help~~~thx~

推荐答案

您需要某种循环来定期检查何时新的框架可用。在Win32中，这通常在主消息循环中完成。以下是如何完成它的一般方法。如果你想要Win32特定的方式（我不是专家）在"Color Basics"中查看
。 SDK附带的示例。

You need some kind of a loop to periodically check when new frames are available. In Win32 this is usually done in the main message loop. Below is a generic way of how it can be done. If you want Win32 specific way (which I'm not an expert of) have a look at "Color Basics" sample which comes with the SDK.

// Declare handles to color stream and next frame event
void *colorStreamHandle;
void *colorNextFrameEvent;

// Initialize sensor
INuiSensor *sensor;
// ...

// Create an event
colorNextFrameEvent = CreateEvent(NULL, TRUE, FALSE, NULL);

// Open color stream and obtain handle to it
sensor->NuiImageStreamOpen(NUI_IMAGE_TYPE_COLOR, NUI_IMAGE_RESOLUTION_640x480, 0, 2, colorNextFrameEvent, &colorStreamHandle);

// Events loop, you may want this in a separate thread
while(running)
{
    if (WaitForSingleObject(nextColorFrameEvent, 0) == WAIT_OBJECT_0)
    {
        // the frame is ready
        // call a function or send a signal, e.g.
        processFrame();
    }
    Sleep(1);
}

void processFrame()
{
    HRESULT hr;
    NUI_IMAGE_FRAME rgbFrame;

    hr = sensor->NuiImageStreamGetNextFrame(colorStreamHandle, 0, &rgbFrame);

// do what you want with the frame

}

这篇关于如何在使用OpenCV的Kinect传感器获得的RGB视频中使用Frame-Difference？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！