Computing Minimum Complexity 1D Curve Simplifications under the Fréchet Distance

Abstract

We consider the problem of simplifying curves under the Fréchet distance. Let P be a curve and ε ≥ 0 be a distance threshold. An ε-simplification is a curve within Fréchet distance ε of P . We consider ε-simplifications of minimum complexity (i.e. minimum number of vertices). Parameterized by ε, we define a continuous family of minimum complexity ε-simplifications P ε of a curve P in one dimension. We present a data structure that after linear preprocessing time can report the ε-simplification in linear output-sensitive time. Moreover, for k ≥ 1, we show how this data structure can be used to report a simplification P ε with at most k vertices that is closest to P in O(k) time.