… as long as you use a single circle around the exact address to hide.
It was already raised elsewhere, that the so called private zones around an address to hide is misleading. As long as they have the form of a regular circle, it needs only three entry/exit points from this “private zone” to calculate the position of the hidden address.
Assume two differing running/cycling tracks published on any of the social web pages applying private zones to your track for “protection” of your privacy, you can already have four points on the border of the private zone, which is already one more than required to calculate the center.
Of course the accuracy is depending on GPS precision, radius size and distance of the points on circle from each other (triangle angle). The method is not applicable in geographic longitude/latitude projection directly and needs to be transformed to an equidistant projection first. That’s also why the equidistant circle is an ellipse on the map in geographic projection above.
The simplest solution would be not to publish any tracks. However, if your ego requires you to publish your tracks, you should use a private zone around a public address (e.g. police station nearby), an unfriendly neighbor or define three overlapping privacy zones with unequal distances from address to hide. Or don’t use privacy zones at all and press the start stop buttons several meters away from home.
I have written a short python script with some artificial data, which can be found here. The unterlying math is explained in the documentation. You can download the scripts to test your own data with it.
| Longitude | Latitude | Radius | |
|---|---|---|---|
| Defined | 8.273765 | 49.999 | 500.0 |
| Calculated | 8.27377 | 49.999 | 500.523 |