We present examples of how to automate contour extraction from images, where the desired contour is the outline of the largest object in an image. We also focus on contours of vases and demonstrate how we can symmetrize vases using Karcher means.
There is also code from the library fdasrsf to show how we can compute pairwise geodesic distances in the SRVF framework, as well as Karcher means and tangent PCA.
Finally, there are also examples of how one can implement a KNN classifier using only a distance matrix, to perform classification.