How to measure the distance error between local and global trajectories?

I'm looking for a straightforward way to measure the error between samples of global trajectory w.r.t samples of a local planner tracking the global the number of samples are different, and each sample represents a (x,y) location in 2D surface. I thought about calculating the surface between the two curves formed by respective samples set, how to proceed with Matlab or may there is a more adequate method.

Answer 1

[Jacob Wood]

My approach would be: resample red curve onto the blue x values using interp1, find local error in each point, sum together to compute global error. Something like:

red_resample = interp1(red_x,red_y,blue_x)
errors = abs(blue-red_resample)
global_error = sum(errors)

[darova]

I like this approach. Looks smart