{"title":"Polarimetric Spatio-Temporal Light Transport Probing","authors":"Seung-Hwan Baek, Felix Heide","doi":"10.1145/3450550.3469737","DOIUrl":null,"url":null,"abstract":"Fig. 1. We propose a computational light transport probingmethod that decomposes transport into full polarization, spatial and temporal dimensions.Wemodel this multi-dimensional light transport as a tensor and analyze low-rank structure in the polarization domain which is exploited by our polarimetric probing method. We instantiate our approach with two imaging systems for spatio-polarimetric and coaxial temporal-polarimetric capture. (a)&(d) Conventional intensity imagers integrate incident light intensity over space and time independently of the polarization states of light, losing geometric and material information encoded in the polarimetric transport. Capturing polarization-resolved spatial transport components of (b) epipolar and (c) non-epipolar dimensions enable fine-grained decomposition of light transport. Combining temporal and polarimetric dimensions, we separate (e) geometry-dependent reflections and (f) direct/indirect reflections that cannot be resolved in the temporal-only measurements.","PeriodicalId":286424,"journal":{"name":"ACM SIGGRAPH 2021 Emerging Technologies","volume":"103 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"11","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"ACM SIGGRAPH 2021 Emerging Technologies","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3450550.3469737","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 11
Abstract
Fig. 1. We propose a computational light transport probingmethod that decomposes transport into full polarization, spatial and temporal dimensions.Wemodel this multi-dimensional light transport as a tensor and analyze low-rank structure in the polarization domain which is exploited by our polarimetric probing method. We instantiate our approach with two imaging systems for spatio-polarimetric and coaxial temporal-polarimetric capture. (a)&(d) Conventional intensity imagers integrate incident light intensity over space and time independently of the polarization states of light, losing geometric and material information encoded in the polarimetric transport. Capturing polarization-resolved spatial transport components of (b) epipolar and (c) non-epipolar dimensions enable fine-grained decomposition of light transport. Combining temporal and polarimetric dimensions, we separate (e) geometry-dependent reflections and (f) direct/indirect reflections that cannot be resolved in the temporal-only measurements.