Comparism of different statistical models used in shape index calculation
| Authors | |
|---|---|
| Year of publication | 2019 |
| Type | Appeared in Conference without Proceedings |
| Citation | |
| Description | Shape index, the measure of local surface topology, is calculated using several different linear statistical models of z coordinates on x and y coordinates, i.e. quadratic with interaction without and with intercept, cubic with interaction of x and y without and with intercept (with and without other interactions), and similar models of higher order. The estimates of regression coefficients related to the quadratic terms and their interaction are elements of Weingarten matrix from which the principal curvatures are calculated. Our data contain stereo-photogrammetry images of human faces. We used one control (healthy) patient and one facial palsy patient to demonstrate differenties of selected linear models. The results are compared numerically and visually by static images of the human face in different views, i.e. frontal, lateral and vertical. All statistical analyses and visualisations are performed in R. |
| Related projects: |