[SOLVED] Trying to isolate the silhouette using ofxNI2

beginners
#1

Hi, I want to isolate the user’s silhoutte contour using ofxNI2 (OpenNI2, NiTE2).
ofxNI2 , OpenNI2 and NiTE2 examples are working fine.

Ubuntu 12.04 64 bits
oF 0.8.4

I did read a post about doing the same using ofxOpenNI (OpenNI), but I want to use OpenNI2 advantages over OpenNI.

If anyone has experienced with this addon maybe could help me.

As I understood getUserMap() has the information I need: any value different from 0 is the user ID and otherwise is the background.
I added this function inline nite::UserMap getUserMap() {return user_map;} to the ofxNiTE2::UserTracker class to get the user map (I didn’t see where to get it in the addon)

I noticed that depth.getWidth() = userTracker->getUserMap().getWidth and the it’s the same for the Height, which I think it’s right.

The core code is posted beolw. As a result I get a black screen which it shouldn’t happend, but *userPixel is always 0 (background) that’s the reason for the black screen.

I appreciate any hint of what I am doing wrong. Thanks!

//variables initialized/ within setup()
ofImage img;
ofxNiTE2::UserTracker *userTracker

//variable that correspond to the object
    ofxNI2::DepthStream depth;


void testApp::update()
{

    const short int *userPixels = (userTracker->getUserMap()).getPixels(); 
    ofShortPixels &depthPixels = depth.getPixelsRef(); 
    long total = depthPixels.getWidth()*depthPixels.getHeight();

    unsigned char maskPixels[total];
   //userPixels has more elements than the "total" value calculated above.

       for (long i =0 ;i < total; i++) {

         if (depthPixels[i] != 0){  

            if (*userPixels != 0) {
                  maskPixels[i] = 255;
             else            
                  maskPixels[i] = 0;                 

            userPixels++;
         }

    }

    img.setFromPixels(maskPixels,320,240,OF_IMAGE_GRAYSCALE);
    img.update();

    device->update();
}

//--------------------------------------------------------------
void testApp::draw()
{
	//color.draw();
	img.draw(0,0)
}
#2

Well I solved it, it was quite simple. I deleted the condition that asked if the depth data was different from 0 to manipulate the user data. (it refers to this line in my previous post if (depthPixels[i] != 0){ )

I just have to manipulate the pixels given by userTracker->getUserMap()).getPixels()

I think the reason is because the user data and the depth data have different type of content, so it didn’t make sense to have this condition. Any feedback or a better solution regarding this topic is welcome.

void testApp::update()
{

    const short int *userPixels = (userTracker->getUserMap()).getPixels(); 
   long total = (userTracker->getUserMap()).getWidth()*(userTracker->getUserMap()).getHeight();

    unsigned char maskPixels[total];

       for (long i =0 ;i < total; i++) {
           if (userPixels) {
                if (*userPixels != 0) {
                      maskPixels[i] = 255;

                } else
                {
                      maskPixels[i] = 0;

                      }
            }
            userPixels++;
                  
    }