I found a paper using a similar idea (but using a different method to step the parameter) to find the orthogonal projection onto curves. It also converges quadraticaly, and seems a little faster through my experiments. http://www.geometrie.tugraz.at/wallner/sproj.pdf