Patella Shape Extraction from 3D Point Cloud Data for Personalized Knee Brace

Abstract

We introduce a novel method to extract the front shape of the patella using 3D point cloud data of the knee obtained with a high resolution scanner or RGB-D camera. The extraction is handled by calculating the angular difference of the normal vector from the center point normal. From the observation of the protrusion of the patella, the outline of the patella can be determined by the starting point at which the deflection of the frontal patella surface increases from the center to the outside. In this paper, we describe the process of patellar shape extraction used for designing a patella protection ring for customized knee braces.