evo
evo copied to clipboard
Add error metrics for scale error
Another useful measure, I think, is scale, that is the ratio of traveled distance of estimate over reference:
and derived thereof the scale error
The implementation is part of the relative error measures. An example looks like this: