cartagen.smooth_topographic

smooth_topographic(geometry, iterations=2, angle=10.0, weights=[1.0, 2.0, 1.0])

Smooth a line or polygon and mimic hand-made cartographic generalization.

This algorithm described in Müller and Wang [1] combines both a low-pass filter using weighted moving average (WMA), and a high-pass filter using an angular threshold. It is suitable to smooth topographic features such as rivers, coastlines, lakes, etc. and simplify it to mimic hand-made generalization.

Accept Multi geometries. If a polygon is provided, it also applies the smoothing to its holes using the same parameters.

Parameters:

• geometry (LineString, MultiLineString, Polygon, MultiPolygon, LinearRing) – The geometry to smooth. If an open line is provided, the endpoints are preserved. If a closed ring or polygon is provided, the smoothing wraps around.

• iterations (int, optional) – Number of low-pass filter passes before thinning. Default is 2.

• angle (float, optional) – Turning-angle threshold in degrees for the high-pass step. Default is 10.0.

• weights (sequence of odd length, optional) – Weights for the moving average window. Default is [1, 2, 1].

Returns:

LineString, MultiLineString, Polygon, MultiPolygon, LinearRing – Simplified geometry of the same type as input.

See also

smooth_gaussian

Smooth a line or a polygon and attenuate its inflexion points.

smooth_platre

Smooth a line and preserve the integrity of sharp turns.

smooth_snake

Smooth a line or polygon using the snake method.

smooth_taubin

Smooth a line or polygon and prevent shrinkage.

smooth_wma

Smooth a line or polygon using a low-pass filter.

Examples

>>> line = LineString([(0, 0), (1, 0.9), (2, 0.1), (3, 1), (4, 0)])
>>> simplify_topographic(line, 1, 15.0)
<LINESTRING (0 0, 2 0.5, 4 0)>

(Source code)