The trick is to find the center of mass of an object.
To do this, one isolates the object by looking for zones with approximately the same distance.
If a person is standing in front of the sensor all his distances will be 50cm +/- 10, for example.
Then look at the number of zones, and the shape. Most people have heads over shoulders, but they might be turned sideways, so that's not always a clue. But at 50cm, a person would occupy 20 zones, at 1M, maybe 10 zones. (I'm guessing a bit here. I've kind of forgotten our results.) A hand at 50cm, would cover far fewer zones.
But you can develop a table of how many zones a person is likely to cover at a few distances. Use that.
But don't forget the value of signal strength. A zone fully covered would have a stronger signal return than would a partially covered one. One can refine which zone is the 'center' using this technique.
Once you decide an object is a person, and you have his (or her) center of mass, you can update your screen.
