Dynamic distance-based shape features for gait recognition

We propose a novel skeleton-based approach
to gait recognition. The core of our method consists of
employing the screened Poisson equation for construct-
ing a family of smooth distance functions associated
with a given shape. The screened Poisson distance func-
tion approximations nicely absorb shape boundary per-
turbations and allow us to dene a rough shape skele-
ton which is relatively stable to such boundary per-
turbations. We demonstrate that pixel-wise variance of
silhouette skeletons is a powerful gait cycle descriptor
leading to a signicant improvement over the existing
state of the art gait recognition rate.