Salient spectral features for points detection

In this paper we introduce a novel method for detecting salient features points on 3D meshes. The contribution of the proposed method is to detect salient feature points in the dual graph spectral domain instead of spatial one. A dual graph Laplacian spectrum of 3D shape is firstly computed for each triangles of the shape. Then we compute the geometric energy of the surface shape using eigenfunctions of the proposed Laplacian matrix. Maxima are finally picked with a scale factor allowing the salient triangles identification and salient points are detected from the triangles centroids. An evaluation of the experimental results and a comparison with the state of the art methods shows the effectiveness of our algorithm.