Sensor Programming

Question

I am doing kinect sensor project to generate point cloud And
I have to get the shortest distance (value of z in point cloud) between object and Sensor.
I am having problem in getting the right values of z means ditance of object from the camera.
I got some values in milimeter but those are coming .09 like that .That is wrong.
I have used the function NuiTransformDepthImageToSkeleton() for getting the image depth in front of the camera.
This is the method I am getting the data
long* depth2rgb = (long*) depthToRgbMap;
NUI_IMAGE_FRAME imageFrame;
NUI_LOCKED_RECT LockedRect;
if (sensor->NuiImageStreamGetNextFrame(depthStream, 0, &imageFrame) < 0) return;
INuiFrameTexture* texture = imageFrame.pFrameTexture;
texture->LockRect(0, &LockedRect, NULL, 0);
if (LockedRect.Pitch != 0) {
const USHORT* curr = (const USHORT*) LockedRect.pBits;
for (int j = 0; j < height; ++j)
std::ofstream fout3("XYZW_Value_MM.txt");
if (fout3.is_open())
fout3 << "X" << "\t\t" << "Y"<<"\t\t"<< "Z" << "\t\t" << "W" <<
std::endl;
for (int i = 0; i < width; ++i)
// Get depth of pixel in millimeters
USHORT depth = NuiDepthPixelToDepth(*curr++);
// Store coordinates of the point corresponding to this pixel
Vector4 pos1;
Vector4 pos = NuiTransformDepthImageToSkeleton(i, j, depth << 3, NUI_IMAGE_RESOLUTION_640x480);
long double x, y, z, w;
x = pos.x ;
y = pos.y ;
z = pos.z ;
w = pos.w ;
fout3 << x << "\t";
fout3 << y << "\t";
fout3 << z << "\t";
fout3 << w << std::endl;
// Store the index into the color array corresponding to this pixel
NuiImageGetColorPixelCoordinatesFromDepthPixelAtResolution(
NUI_IMAGE_RESOLUTION_640x480, NUI_IMAGE_RESOLUTION_640x480, NULL,
i, j, depth << 3, depth2rgb, depth2rgb + 1);
depth2rgb += 2;
texture->UnlockRect(0);
sensor->NuiImageStreamReleaseFrame(depthStream, &imageFrame);
It is not giving right value .Please suggst.

Answer 1

All Kinect distances are in millimeters. You will need to convert the depth values to inches yourself. Review the depth basics samples for the code needed to pull the depth value point from the depthframe. If you are using the packed version of the depth
frame, you need to ensure you right shift by 3 to remove the lower 3 bits for the player mask.
To get real world coordinates, you need to use the coordinate mapper and map the depth frame to SkeletonSpace. The Depth3D samples demonstrates this.
Carmine Sirignano - MSFT