{"title":"Variational model for robust computation of optical flow under varying illumination","authors":"M. Kharbat, N. Aouf","doi":"10.1109/UKRICIS.2010.5898120","DOIUrl":null,"url":null,"abstract":"We present a variational optical flow method to robustly estimate motion under varying illumination. The corresponding energy functional adopts two constancy assumptions including moments-based descriptors as well as the gradient of the brightness. We illustrate how the proposed assumptions are designed to compensate for the multiplicative and additive components of the brightness variation. The method is described within a combined local-global framework which preserves discontinuity over spatiotemporal domain. Comprehensive evaluations demonstrate the reliable results of our method under strong varying illumination as well as its robustness against random noise.","PeriodicalId":359942,"journal":{"name":"2010 IEEE 9th International Conference on Cyberntic Intelligent Systems","volume":"57 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2010 IEEE 9th International Conference on Cyberntic Intelligent Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/UKRICIS.2010.5898120","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
We present a variational optical flow method to robustly estimate motion under varying illumination. The corresponding energy functional adopts two constancy assumptions including moments-based descriptors as well as the gradient of the brightness. We illustrate how the proposed assumptions are designed to compensate for the multiplicative and additive components of the brightness variation. The method is described within a combined local-global framework which preserves discontinuity over spatiotemporal domain. Comprehensive evaluations demonstrate the reliable results of our method under strong varying illumination as well as its robustness against random noise.