A simple algorithm for efficient piecewise linear approximation of space curves

An on-line method for piecewise linear approximation of open or closed space curves is described. The algorithm guarantees approximation within a deviation threshold and is offered as an efficient, on-line alternative to the split and merge approach. Other efficient methods operate only on planar curves, whereas the approach we offer is also appropriate for space curves. A simple function of chord and arc length is used to form the initial set of approximating points. Preliminary Gaussian smoothing, posterior merging and least squares fitting are optional and can be done depending on the application. The algorithm performance has been tested on a variety of planar curves and comparisons made with other piecewise linear curve approximation algorithms