Features Identification Module

Features Identification Module

The Features Identification Module is designed for extracting relevant features, which could be anatomical features (such as eyes and/or the mouth) or decorative features (such as rosettes and/or repeated circles) from the 3D models of CH collections. The process of feature extraction is built upon an algorithm that takes as input a mesh in .ply format, a shape chosen from within an atlas of available plane curves (such as circles, citrus curves, spirals, geometric petal curves) and various other technical parameters. The module can also, optionally, work on the input model in .fct format, used in the faceting of fragments. The algorithm returns a list of vertices indices, corresponding to the region of each identified feature, and the parameters associated with the fitting curves. The features identification algorithm contains the routines used to identify, to locate and to evaluate the size of relevant features of the input 3D model.

The routine is TRL-4, it is integrated in the dashboard platform of GRAVITATE as well and can be used as a stand-alone MATLAB function for identifying and locating features in 3D meshes and point clouds.

There are some technical limitations due to the development of the algorithms in a MATLAB environment, although it is envisaged that, fostered by the Centre of Excellence the component will be further developed in the next future.

Release

User Manual

License