Interactive Curvature Tensor Visualization on Digital Surfaces H´ el` ene Perrier 1 , J´ er´ emy Levallois 1,2 , David Coeurjolly 1(B ) , Jean-Philippe Farrugia 1 , Jean-Claude Iehl 1 , and Jacques-Olivier Lachaud 2 1 Universit´ e de Lyon, CNRS LIRIS, UMR5205, 69622 Lyon, France david.coeurjolly@liris.cnrs.fr 2 Universit´ e de Savoie Mont Blanc, CNRS LAMA, UMR5127, 73776 Chamb´ ery, France Abstract. Interactive visualization is a very convenient tool to explore complex scientific data or to try different parameter settings for a given processing algorithm. In this article, we present a tool to efficiently analyze the curvature tensor on the boundary of potentially large and dynamic digital objects (mean and Gaussian curvatures, principal cur- vatures, principal directions and normal vector field). More precisely, we combine a fully parallel pipeline on GPU to extract an adaptive triangu- lated isosurface of the digital object, with a curvature tensor estimation at each surface point based on integral invariants. Integral invariants being parametrized by a given ball radius, our proposal allows to explore interactively different radii and thus select the appropriate scale at which the computation is performed and visualized. Keywords: Isosurface visualization · Digital geometry · Curvature estimation · GPU 1 Introduction Volumetric objects are being more and more popular in many applications ranging from object modeling and rendering in Computer Graphics to geometry process- ing in Medical Imaging or Material Sciences. When considering large volumetric data, interactive visualization of those objects (or isosurfaces) is a complex prob- lem. Such issues become even more difficult when dynamic volumetric datasets are considered. Beside visualization, we are also interested in performing geom- etry processing on the digital object and to explore different parameter settings of the geometry processing tool. Here, we focus on curvature tensor estimation (mean/Gaussian curvature, principal curvatures directions. . . ). Most curvature estimators require a parameter fixing the scale at which the computation is per- formed. For short, such parameter (integration radius, convolution kernel size. . . ) This work has been mainly funded by DigitalSnow ANR-11-BS02-009, KIDICO ANR-2010-BLAN-0205 and PRIMES Labex ANR-11-LABX-0063/ANR-11-IDEX- 0007 research grants. c  Springer International Publishing Switzerland 2016 N. Normand et al. (Eds.): DGCI 2016, LNCS 9647, pp. 282–294, 2016. DOI: 10.1007/978-3-319-32360-2 22