Digital Foot Models from Sensor Data for Improved Orthopedic Aid Fitting

This report describes and develops different methods for converting 3D time series data to surface representations. The considered methods contain public domain mesh generation software, as well as linear regression models and surface representations using signed distance functions. We provide a simple code base for the latter two methods such that they can be used for further research in a simple manner. We apply and test the algorithms on a point cloud for a foot model. All methods yield good representations of the underlying foot geometry. Such algorithms can therefore have a big impact on handling foot problems.