Hi Group,
i've set up a VideoFrameGrabber class according to the examples in Dr. Andrew Davison: Kinect Open Source Programming Secrets
Now my Context-Generating takes up to 6 seconds. Any Ideas, before i let a profiler run over it?
Here's some source:
private VideoFrameGrabber()
{
System.out.println(System.currentTimeMillis());
configOpenNI();
System.out.println(System.currentTimeMillis());
}
gives me the following output:
1338455441777
Started context generating...
1338455447421
Image dimensions (640, 480)
configOpenNI is as follows:
private void configOpenNI()
// create context, depth generator, image generator
{
try {
context = new Context();
// add the NITE Licence
License licence = new License("PrimeSense", "0KOIk2JeIBYClPWVnMoRKn5cdY4="); // vendor, key
context.addLicense(licence);
depthGen = DepthGenerator.create(context);
imageGen = ImageGenerator.create(context);
// if possible, set the viewpoint of the DepthGenerator to match the ImageGenerator
// depthGen.getAlternativeViewpointCapability().setViewpoint(imageGen);
MapOutputMode mapMode = new MapOutputMode(640, 480, 30); // xRes, yRes, FPS
depthGen.setMapOutputMode(mapMode);
imageGen.setMapOutputMode(mapMode);
imageGen.setPixelFormat(PixelFormat.RGB24);
// set Mirror mode for all
context.setGlobalMirror(false);
context.startGeneratingAll();
System.out.println("Started context generating...");
depthMD = depthGen.getMetaData();
// use depth metadata to access depth info (avoids bug with DepthGenerator)
}
catch (Exception e) {
System.out.println(e);
System.exit(1);
}
} // end of configOpenNI()
Any ideas greatly appreciated.
Daniel