Hey guys,
i've just wondered why the more advanced (not the basic version) algorithms for lane detection uses a planar homography to get a "bird view". What advantages does that have? I can calculate a polynomial function for the lane as well without the "bird view", can't I?