On Floating-Point Normal Vectors

dc.contributor.authorMeyer, Quirinen_US
dc.contributor.authorSuessmuth, Jochenen_US
dc.contributor.authorSussner, Gerden_US
dc.contributor.authorStamminger, Marcen_US
dc.contributor.authorGreiner, Guentheren_US
dc.date.accessioned2015-02-23T16:58:12Z
dc.date.available2015-02-23T16:58:12Z
dc.date.issued2010en_US
dc.description.abstractIn this paper we analyze normal vector representations. We derive the error of the most widely used representation, namely 3D floating-point normal vectors. Based on this analysis, we show that, in theory, the discretization error inherent to single precision floating-point normals can be achieved by 250.2 uniformly distributed normals, addressable by 51 bits. We review common sphere parameterizations and show that octahedron normal vectors perform best: they are fast and stable to compute, have a controllable error, and require only 1 bit more than the theoretical optimal discretization with the same error.en_US
dc.description.number4en_US
dc.description.seriesinformationComputer Graphics Forumen_US
dc.description.volume29en_US
dc.identifier.doi10.1111/j.1467-8659.2010.01737.xen_US
dc.identifier.issn1467-8659en_US
dc.identifier.pages1405-1409en_US
dc.identifier.urihttps://doi.org/10.1111/j.1467-8659.2010.01737.xen_US
dc.publisherThe Eurographics Association and Blackwell Publishing Ltden_US
dc.titleOn Floating-Point Normal Vectorsen_US
Files