This paper covers the need to reduce the requirement for disparity generation for real-time video processing. The motivation for the change and the methodology are discussed. The presented algorithm, a dilated Laplacian approach to guiding regenerations, reduces the requirement by an average of 49% but maintains a map accuracy of 89%.

Original languageEnglish
Title of host publication2015 INTERNATIONAL CONFERENCE ON SYSTEMS, SIGNALS AND IMAGE PROCESSING (IWSSIP 2015)
EditorsP Liatsis, A Uus, S Miah
PublisherPubl by IEEE
Pages101-104
Number of pages4
DOIs
Publication statusPublished - 2015
Event22nd International Conference on Systems, Signals and Image Processing (IWSSIP) - London
Duration: 10 Sep 201512 Sep 2015

Publication series

NameInternational Conference on Systems Signal and Image Processing
PublisherIEEE
ISSN (Print)2157-8672

Conference

Conference22nd International Conference on Systems, Signals and Image Processing (IWSSIP)
CityLondon
Period10/09/1512/09/15

    Research areas

  • Computer Vision, Disparity Generation, Laplacian

ID: 1162834