Wood Classification

Wood (1996) proposed a point classification based on the quadratic coefficients in a region about the point of interest.

For best use, this technique uses three variables:

• Slope tolerance
• Convexity tolerance
• Window (region) size
• We have added a minimum slope that can be classified.

We modified the Wood's (1996) color scheme, to make our maps more closely resemble topography. Ridges and peaks are green to simulate the forest covered uplands, and the streams/valleys and pits are blue to simulate the stream drainages. Passes are in yellow.

The best values for the parameters depend on the DEM and regional characteristics.

MICRODEM also implements two other algorithms.

The Wood algorithm:

if point elevation is missing, PointType := MissingPoint

if neighboring point elevation is missing, PointType := EdgePoint

if (SlopeDeg < MDDef.MinSlopeToClassify) then PointType := FlatPoint
else if (slopeDeg > MDDef.SlopeClassTolerance) then begin
        if (CrossC > MDDef.ConvexClassTolerance) then PointType := RidgePoint
        else if (CrossC < -MDDef.ConvexClassTolerance) then PointType := ValleyPoint
        else PointType := OtherPoint;
end
else if (MaxCurve > MDDef.ConvexClassTolerance) then begin
        if (MinCurve > MDDef.ConvexClassTolerance) then PointType := PeakPoint
        else if (MinCurve < -MDDef.ConvexClassTolerance) then PointType := PassPoint
        else PointType := RidgePoint;
end
else if (MinCurve < -MDDef.ConvexClassTolerance) then begin
        if (MaxCurve < -MDDef.ConvexClassTolerance) then PointType := PitPoint
        else PointType := ValleyPoint
end
else PointType := OtherPoint;
 

Last revision 8/18/2008