Automatic SfM-Based 2D-to-3D Conversion for Multi-Object Scenes

Hak Gu KIM
Jin-ku KANG
Byung Cheol SONG

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E97-A    No.5    pp.1159-1161
Publication Date: 2014/05/01
Online ISSN: 1745-1337
DOI: 10.1587/transfun.E97.A.1159
Type of Manuscript: LETTER
Category: Image
2D-to-3D conversion,  structure-from-motion,  multi-objects,  

Full Text: PDF(508.5KB)>>
Buy this Article

This letter presents an automatic 2D-to-3D conversion method using a structure from motion (SfM) process for multi-object scenes. The foreground and background regions may have different depth values in an image. First, we detect the foreground objects and the background by using a depth histogram. Then, the proposed method creates the virtual image by projecting each region with its computed projective matrix. Experimental results compared to previous research show that the proposed method provides realistic stereoscopic images.