Abstract: 3D mesh segmentation is a fundamental problem of computer graphics with a wide range of applications.In recent years,various approaches have emerged and this paper surveys methods and algorithms based on clustering analysis.We provide the preliminaries to mesh segmentation and identify two distinct types of mesh segmentation,namely patch segmentation and part segmentation.We summarize the shape descriptors in use and classify the state-of-art methods into five categories methodologically,i.e.region growth,hierarchical clustering,iterative clustering and spectral clustering.Four evaluation criteria are given to compare the methods discussed under different application scenarios.Applications and prospects of 3D mesh segmentation are also discussed to shed light on the challenges that future work should face.