Douglas-Peucker Generalization Algorithm for simplifying 2D & 3D polylines and polygons

The process of generalization is normally executed when the map scale has to be reduced. Another usage is in cleaning up noisy representations of polylines, possibly generated by a raster-to-vector conversion (tracing). Generalization can be seen as a form of lossy data compression; given a large and complicated object, we seek to simplify it by reducing detail. Ideally, we obtain a polyline/polygon with far fewer vertices that looks essentially the same at a target scale. This can be a big win in computer graphics, where replacing a large model with a smaller one might have little visual impact but require significantly less storage space and rendering time.