I came across this concept recently--it is laid out pretty well in a Wikipedia article if you want some more details. Basically, you use it to represent a latitude/longitude pair as a single set of characters. There are two basic features to it that are pretty cool:
This is a two-way hash...a point can be encoded and decoded. There is a loss of precision...the final representation does not match the input value exactly, but is within a defined margin of error.
It looks much better than latitude and longitude values when passed in a URL, and it is easy enough to calculate that it is amost cost-free to use.
- Each successive bit of information provides additional accuracy so that the location of the point is gradually refined. This also allows the same formula to handle various levels of precision.
- The bits that represent latitude and longitude are interpolated with one another. That means that a single short piece of information represents the values in two dimensions. This is especially useful for matching and storing unique point values in a database
This is a two-way hash...a point can be encoded and decoded. There is a loss of precision...the final representation does not match the input value exactly, but is within a defined margin of error.
It looks much better than latitude and longitude values when passed in a URL, and it is easy enough to calculate that it is amost cost-free to use.
No comments:
Post a Comment