Help me in labeling the object detected in image and then traking its position under that label.

edit

Ok, so till now I developed the harr cascade of my object and then using basic image processing functions like contour and edge detection i am able to detect and locate the exact point of interest(POI) in the object, like a tip of the finger. I am using kinect on libfreenect and python on linux with open cv. So now i want to label each point (POI) detected in the video frame and and then show its features in the windows, like POI1=x,y,z. Right now what I am getting is a continuous feed of x,y,z of all the POI detected, but no label, hence cannot know which parameter is for which POI. What m I missing here..

--- :( troubled engel!