DefinePK

Abstract

The evaluation of mesh segmentation has received a great deal of attention since 3D mesh segmentation is an essential step in many mesh operations. For this reason, notable efforts have been made towards a better evaluation of mesh segmentation methods, and one of the most popular works is the benchmark of Chen et al., which allows a quantitative evaluation of mesh segmentation algorithms. Based on the given data sets, which comprise manual and automatic segmentations, many evaluation metrics have been proposed recently. In this context, we present in this study an overview of the existing similarity metrics and new ones proposed in our previous works addressing the problem of evaluating 3D mesh segmentation by describing each method and giving an extensive study and experimental comparison of them.