Trails across map pointing to zero/zero lat/long
Posted: Mon Nov 05, 2012 1:10 pm
I have observed on the web map that there are trails from large cities to the point where the Greenwich Meridian crosses the equator (ie lat/long both=zero). These cross oceans, deserts etc and are clearly erroneous.
It seems to me that these must be caused by points erroneously being reported with zero/zero lat/long and being interpolated with the real positions, with the positions on the trail being dependent on the relative signal strengths. Some GPS chipsets erroneously give zero/zero locations occasionally, so some way of dealing with these is needed.
Can I suggest that all GPS points at zero/zero lat/long be deleted or at least excluded from interpolations?
I have also noticed a few other odd trails across oceans, some of which are curved. Not sure what these are caused by. There are some around Hong Kong, and one from France into the Med Sea.
Also, I have some other suggestions for improving accuracy - if there are positions reported for the same BSSID that are more than an adjustable threshold apart, then some kind of error handling is needed. It may be that someone has moved the Wifi AP, in which case the older locations need to be ignored, or it may be that one of the GPS fixes was widely erroneous. Alternatively it is possible that it is a mobile access point, eg a mobile phone working as a Wifi hotspot. Not sure what the best way of handling this is, but interpolating the positions seems pretty meaningless!
It seems to me that these must be caused by points erroneously being reported with zero/zero lat/long and being interpolated with the real positions, with the positions on the trail being dependent on the relative signal strengths. Some GPS chipsets erroneously give zero/zero locations occasionally, so some way of dealing with these is needed.
Can I suggest that all GPS points at zero/zero lat/long be deleted or at least excluded from interpolations?
I have also noticed a few other odd trails across oceans, some of which are curved. Not sure what these are caused by. There are some around Hong Kong, and one from France into the Med Sea.
Also, I have some other suggestions for improving accuracy - if there are positions reported for the same BSSID that are more than an adjustable threshold apart, then some kind of error handling is needed. It may be that someone has moved the Wifi AP, in which case the older locations need to be ignored, or it may be that one of the GPS fixes was widely erroneous. Alternatively it is possible that it is a mobile access point, eg a mobile phone working as a Wifi hotspot. Not sure what the best way of handling this is, but interpolating the positions seems pretty meaningless!