Exploring the JTS.smooth() method

While experimenting with implementing an irregular point contour process for GeoTools this week I came across the new (at least to me) method JTS.smooth() method. This gets a throw away line at the bottom of the documentation that tells me it takes a geometry and a fit value between 0 and 1, and then uses splines to smooth the geometry.

I delighted as I had put off implementing smoothing to the end as I thought it would be hard work, but this was easy and it worked exactly as described. But I was left wondering about what the fit value did, the docs are a little light on explanation and to be honest I get scared looking at Martin’s code too closely. So I built a small test program so I could see what happened as I varied the fit value.

As you can see I quickly discovered that lower values of fit give “curvier” shapes. I also experimented with densification of the lines, which constrains the curve to more of the straight sections of the line.