Share Email Print

Proceedings Paper

Compression scheme by use of object-segmented sub-image array transformed from computational elemental image array based on multiple objects in 3D integral imaging
Author(s): Hyoung-Woo Lee; Ju-Han Lee; Ho-Hyun Kang; Eun-Soo Kim
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

In this paper, we address a highly enhanced compression scheme in the condition of multiple objects in Integral Imaging (InIm) by use of sub-images (SIs) to segment each object and to remove the Motion Vector (MV) of residual image array transformed from Sub-Image Array (SIA). In the pick-up process, SIA is generated from EIA after the perspectives passing through virtual pinhole array is recorded as Elemental Image Array (EIA). The similarity enhancement among SIs expects compression efficiency to improve, but the compression efficiency of the EIA in the picked-up condition of multiple objects does not correspond to that of the picked-up condition of a simplified object. In the proposed scheme, the depth of objects is computed by two adaptive SIs located at horizontal left and right side from the reference SI positioned to the center of the SIA. A depth map image generated from two adaptive the SIs and a reference SI is applied to segment each object considering to the distance between those. Therefore, an adaptive objectsegmented SI is obtained and, which is motion-estimated from the original SIA based on MSE to generate the motioncompensated object-segmented SIA and which SIAs from each segmented object are finally combined as the motioncompensated SIA, and which based on multiple objects is transformed to residual SIA to minimize the spatial redundancy and which SIA is compressed by MPEG-4. The proposed algorithm shows the enhanced compression efficiency than that of the baseline JPEG and the conventional EIA compression scheme.

Paper Details

Date Published: 15 October 2012
PDF: 13 pages
Proc. SPIE 8498, Optics and Photonics for Information Processing VI, 84980B (15 October 2012);
Show Author Affiliations
Hyoung-Woo Lee, Kwangwoon Univ. (Korea, Republic of)
Ju-Han Lee, Kwangwoon Univ. (Korea, Republic of)
Ho-Hyun Kang, Kwangwoon Univ. (Korea, Republic of)
Eun-Soo Kim, Kwangwoon Univ. (Korea, Republic of)

Published in SPIE Proceedings Vol. 8498:
Optics and Photonics for Information Processing VI
Abdul A. S. Awwal; Khan M. Iftekharuddin, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?