Presentation 2015-03-19
Driving Simulation Systems Based on Real Images Using Real-time Novel View Synthesis
Hiroya MATSUMOTO, Norihiko KAWAI, Tomokazu SATO, Takashi MACHIDA, Naokazu YOKOYA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Driving-simulation systems have been commonly used for training beginner drivers. In common driving-simulation systems, driver's view images are rendered by 3D CG models. One problem here is that rendered images are not photorealistic and thus they cannnot give real sensation to the users. For this problem, a novel view synthesis method, which creates a photorealistic image of virtual viewpoint from real images is expected to be a promising solution. One critical problem of the conventional approaches is that it takes large computational time to create one image. Thus, it is difficult to simply employ them for developing a photorealistic driving simulator. This thesis proposes a method to create a novel view image in real-time by drastically reducing the computational cost using (1) off-line local plane fitting to a point cloud and (2) a coarse-to-fine approach to create view-dependent geometry.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) novel view synthesis / vehicle safety system / 3D point cloud / omni-directional image
Paper # BioX2014-54,PRMU2014-174
Date of Issue

Conference Information
Committee PRMU
Conference Date 2015/3/12(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Driving Simulation Systems Based on Real Images Using Real-time Novel View Synthesis
Sub Title (in English)
Keyword(1) novel view synthesis
Keyword(2) vehicle safety system
Keyword(3) 3D point cloud
Keyword(4) omni-directional image
1st Author's Name Hiroya MATSUMOTO
1st Author's Affiliation Nara Institute of Science and Technology()
2nd Author's Name Norihiko KAWAI
2nd Author's Affiliation Nara Institute of Science and Technology
3rd Author's Name Tomokazu SATO
3rd Author's Affiliation Nara Institute of Science and Technology
4th Author's Name Takashi MACHIDA
4th Author's Affiliation Toyota Central R&D Labs
5th Author's Name Naokazu YOKOYA
5th Author's Affiliation Nara Institute of Science and Technology
Date 2015-03-19
Paper # BioX2014-54,PRMU2014-174
Volume (vol) vol.114
Number (no) 521
Page pp.pp.-
#Pages 6
Date of Issue