Portal Extraction Based on an Opening Labeling for Ray Tracing

• Main Authors: Noël, Laurent, Biri, Venceslas
• Other Authors: Laboratoire d'Informatique Gaspard-Monge (LIGM), Université Paris-Est Marne-la-Vallée (UPEM)-École des Ponts ParisTech (ENPC)-ESIEE Paris-Fédération de Recherche Bézout (BEZOUT), Centre National de la Recherche Scientifique (CNRS)-Centre National de la Recherche Scientifique (CNRS)-Centre National de la Recherche Scientifique (CNRS)
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2015
• Subjects: ray tracing; opening map; labeling; portals; [INFO.INFO-GR]Computer Science [cs]/Graphics [cs.GR]; [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV]; Iceland
• Online Access: https://hal.science/hal-01159795; https://hal.science/hal-01159795/document; https://hal.science/hal-01159795/file/ismm2015.pdf; https://doi.org/10.1007/978-3-319-18720-4_3

International audience Rendering photo-realistic images from a 3D scene description remains a challenging problem when processing complex geometry exposing many occlusions. In that case, simulating light propagation requires hours to produce a correct image. An opening map can be used to extract information from the geometry of the empty space of a scene, where light travels. It describes local thickness and allows to identify narrow regions that are difficult to traverse with ray tracing. We propose a new method to extract portals in order to improve rendering algorithms based on ray tracing. This method is based on the opening map, which is used to define a labeling of the empty space. Then portals-2D surfaces embedded in empty space-are extracted from labeled regions. We demonstrate that those portals can be sampled in order to explore the scene efficiently with ray tracing.