The code implements a simple demo to show how to use ShellPCA given a bunch of meshes (.ply).
The code is implemented based on quocMesh library http://numod.ins.uni-bonn.de/software/quocmesh/ To run the demo, you need to follow instructions there first. Once installation is done, the script could be added and compiled.
The spheres demo uses linear average although an elastic average is expected to perform better You can find the script to compute elastic average under quocMesh repositories easily
- determine the number of data to use;
- provide bendWeight and memWeight, default values are 1.0;
- run and visualize eigen modes
@inproceedings{ZhHeRu15, author = {Zhang, Chao and Heeren, Behrend and Rumpf, Martin and Smith, William}, title = {Shell {PCA}: statistical shape modelling in shell space}, booktitle = {Proc. of IEEE International Conference on Computer Vision}, year = {2015}, pdf = {http://numod.ins.uni-bonn.de/research/papers/public/ZhHeRu15.pdf 1} }