Depth Map Estimation Using Census Transform for Light Field Cameras

Takayuki TOMIOKA; Kazu MISHIBA; Yuji OYAMADA; Katsuya KONDO

doi:10.1587/transinf.2017EDP7052

Depth Map Estimation Using Census Transform for Light Field Cameras

Takayuki TOMIOKA, Kazu MISHIBA, Yuji OYAMADA, Katsuya KONDO

Full Text Views

0

Share
Cite this

Summary :

Depth estimation for a lense-array type light field camera is a challenging problem because of the sensor noise and the radiometric distortion which is a global brightness change among sub-aperture images caused by a vignetting effect of the micro-lenses. We propose a depth map estimation method which has robustness against sensor noise and radiometric distortion. Our method first binarizes sub-aperture images by applying the census transform. Next, the binarized images are matched by computing the majority operations between corresponding bits and summing up the Hamming distance. An initial depth obtained by matching has ambiguity caused by extremely short baselines among sub-aperture images. After an initial depth estimation process, we refine the result with following refinement steps. Our refinement steps first approximate the initial depth as a set of depth planes. Next, we optimize the result of plane fitting with an edge-preserving smoothness term. Experiments show that our method outperforms the conventional methods.

Publication: IEICE TRANSACTIONS on Information Vol.E100-D No.11 pp.2711-2720

Publication Date: 2017/11/01

Publicized: 2017/08/02

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2017EDP7052

Type of Manuscript: PAPER

Category: Image Recognition, Computer Vision

Authors

Takayuki TOMIOKA
  Tottori University
Kazu MISHIBA
  Tottori University
Yuji OYAMADA
  Tottori University
Katsuya KONDO
  Tottori University

Keyword

light field camera, Lytro, depth map, census transform

Cite this

Copy

Takayuki TOMIOKA, Kazu MISHIBA, Yuji OYAMADA, Katsuya KONDO, "Depth Map Estimation Using Census Transform for Light Field Cameras" in IEICE TRANSACTIONS on Information, vol. E100-D, no. 11, pp. 2711-2720, November 2017, doi: 10.1587/transinf.2017EDP7052.
Abstract: Depth estimation for a lense-array type light field camera is a challenging problem because of the sensor noise and the radiometric distortion which is a global brightness change among sub-aperture images caused by a vignetting effect of the micro-lenses. We propose a depth map estimation method which has robustness against sensor noise and radiometric distortion. Our method first binarizes sub-aperture images by applying the census transform. Next, the binarized images are matched by computing the majority operations between corresponding bits and summing up the Hamming distance. An initial depth obtained by matching has ambiguity caused by extremely short baselines among sub-aperture images. After an initial depth estimation process, we refine the result with following refinement steps. Our refinement steps first approximate the initial depth as a set of depth planes. Next, we optimize the result of plane fitting with an edge-preserving smoothness term. Experiments show that our method outperforms the conventional methods.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.2017EDP7052/_p

Copy

@ARTICLE{e100-d_11_2711,
author={Takayuki TOMIOKA, Kazu MISHIBA, Yuji OYAMADA, Katsuya KONDO, },
journal={IEICE TRANSACTIONS on Information},
title={Depth Map Estimation Using Census Transform for Light Field Cameras},
year={2017},
volume={E100-D},
number={11},
pages={2711-2720},
abstract={Depth estimation for a lense-array type light field camera is a challenging problem because of the sensor noise and the radiometric distortion which is a global brightness change among sub-aperture images caused by a vignetting effect of the micro-lenses. We propose a depth map estimation method which has robustness against sensor noise and radiometric distortion. Our method first binarizes sub-aperture images by applying the census transform. Next, the binarized images are matched by computing the majority operations between corresponding bits and summing up the Hamming distance. An initial depth obtained by matching has ambiguity caused by extremely short baselines among sub-aperture images. After an initial depth estimation process, we refine the result with following refinement steps. Our refinement steps first approximate the initial depth as a set of depth planes. Next, we optimize the result of plane fitting with an edge-preserving smoothness term. Experiments show that our method outperforms the conventional methods.},
keywords={},
doi={10.1587/transinf.2017EDP7052},
ISSN={1745-1361},
month={November},}

Copy

TY - JOUR
TI - Depth Map Estimation Using Census Transform for Light Field Cameras
T2 - IEICE TRANSACTIONS on Information
SP - 2711
EP - 2720
AU - Takayuki TOMIOKA
AU - Kazu MISHIBA
AU - Yuji OYAMADA
AU - Katsuya KONDO
PY - 2017
DO - 10.1587/transinf.2017EDP7052
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E100-D
IS - 11
JA - IEICE TRANSACTIONS on Information
Y1 - November 2017
AB - Depth estimation for a lense-array type light field camera is a challenging problem because of the sensor noise and the radiometric distortion which is a global brightness change among sub-aperture images caused by a vignetting effect of the micro-lenses. We propose a depth map estimation method which has robustness against sensor noise and radiometric distortion. Our method first binarizes sub-aperture images by applying the census transform. Next, the binarized images are matched by computing the majority operations between corresponding bits and summing up the Hamming distance. An initial depth obtained by matching has ambiguity caused by extremely short baselines among sub-aperture images. After an initial depth estimation process, we refine the result with following refinement steps. Our refinement steps first approximate the initial depth as a set of depth planes. Next, we optimize the result of plane fitting with an edge-preserving smoothness term. Experiments show that our method outperforms the conventional methods.
ER -