MATLAB: How to sample a contour to a fixed length whilst keeping global shape

image analysisimage processingImage Processing Toolbox

I'm trying to normalize the length of a contour to a fixed length but keep the smooth global shape of the contour. My aim is to then extract Fourier Descriptors from this set of boundary points.
For example, the figure below (taken from http://sconseil.fr/static/files/conseil_eusipco07.pdf) shows how 64 boundary points can be used to represent the contour of a given hand shape. It mentions that "the sampling is done interpolating points which are at an equal arc length".
Currently, given an input image (such as the one below), I can obtain boundary points using the Matlab function 'bwboundaries'. My goal is to then normalize the points it gives me to a fixed set of points.

Best Answer

Perhaps you could take a 2D spline http://www.mathworks.com/help/curvefit/examples/constructing-spline-curves-in-2d-and-3d.html and interpolate it at 64 points?
Possibly a smoothing spline might make sense ?
Related Question