A dense depth estimation method using superpixels

Feng Jin; Xuefeng Li

doi:10.1109/ICCWAMTIP.2015.7493994

A dense depth estimation method using superpixels

Feng Jin, Xuefeng Li

School of Automation

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

Conventional stereo matching or depth estimation algorithms always provide incomplete disparity map. These pixels without depth estimation in the map are named depth gaps. Weak texture and occluded areas are main source of depth gaps. We propose a novel method to assign good depth estimation on the areas above. Our algorithm combines state-of-art superpixel segmentation approach and linear filter. First we do superpixel segmentation on reference image, after this every pixel has a label determines which superpixel it belongs to. Then merging distance between superpixels in spatial space and color space, we apply linear filter on every superpixel. We evaluate the performance of our algorithm with some classic stereo datasets to show the promotion we obtained.

Original language	English
Title of host publication	2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	290-294
Number of pages	5
ISBN (Electronic)	9781467382663
DOIs	https://doi.org/10.1109/ICCWAMTIP.2015.7493994
Publication status	Published - 16 Jun 2016
Event	12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015 - Chengdu, China Duration: 18 Dec 2015 → 20 Dec 2015

Publication series

Name	2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015

Conference

Conference	12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015
Country/Territory	China
City	Chengdu
Period	18/12/15 → 20/12/15

Keywords

Depth estimation
Stereo matching
linear filter
superpixel segmentation

Access to Document

10.1109/ICCWAMTIP.2015.7493994

Cite this

Jin, F., & Li, X. (2016). A dense depth estimation method using superpixels. In 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015 (pp. 290-294). Article 7493994 (2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCWAMTIP.2015.7493994

Jin, Feng ; Li, Xuefeng. / A dense depth estimation method using superpixels. 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 290-294 (2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015).

@inproceedings{3c76a7871bdb4022a97e9bce799cfe90,

title = "A dense depth estimation method using superpixels",

abstract = "Conventional stereo matching or depth estimation algorithms always provide incomplete disparity map. These pixels without depth estimation in the map are named depth gaps. Weak texture and occluded areas are main source of depth gaps. We propose a novel method to assign good depth estimation on the areas above. Our algorithm combines state-of-art superpixel segmentation approach and linear filter. First we do superpixel segmentation on reference image, after this every pixel has a label determines which superpixel it belongs to. Then merging distance between superpixels in spatial space and color space, we apply linear filter on every superpixel. We evaluate the performance of our algorithm with some classic stereo datasets to show the promotion we obtained.",

keywords = "Depth estimation, Stereo matching, linear filter, superpixel segmentation",

author = "Feng Jin and Xuefeng Li",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015 ; Conference date: 18-12-2015 Through 20-12-2015",

year = "2016",

month = jun,

day = "16",

doi = "10.1109/ICCWAMTIP.2015.7493994",

language = "English",

series = "2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "290--294",

booktitle = "2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015",

address = "United States",

}

Jin, F & Li, X 2016, A dense depth estimation method using superpixels. in 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015., 7493994, 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015, Institute of Electrical and Electronics Engineers Inc., pp. 290-294, 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015, Chengdu, China, 18/12/15. https://doi.org/10.1109/ICCWAMTIP.2015.7493994

A dense depth estimation method using superpixels. / Jin, Feng; Li, Xuefeng.
2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015. Institute of Electrical and Electronics Engineers Inc., 2016. p. 290-294 7493994 (2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A dense depth estimation method using superpixels

AU - Jin, Feng

AU - Li, Xuefeng

PY - 2016/6/16

Y1 - 2016/6/16

N2 - Conventional stereo matching or depth estimation algorithms always provide incomplete disparity map. These pixels without depth estimation in the map are named depth gaps. Weak texture and occluded areas are main source of depth gaps. We propose a novel method to assign good depth estimation on the areas above. Our algorithm combines state-of-art superpixel segmentation approach and linear filter. First we do superpixel segmentation on reference image, after this every pixel has a label determines which superpixel it belongs to. Then merging distance between superpixels in spatial space and color space, we apply linear filter on every superpixel. We evaluate the performance of our algorithm with some classic stereo datasets to show the promotion we obtained.

AB - Conventional stereo matching or depth estimation algorithms always provide incomplete disparity map. These pixels without depth estimation in the map are named depth gaps. Weak texture and occluded areas are main source of depth gaps. We propose a novel method to assign good depth estimation on the areas above. Our algorithm combines state-of-art superpixel segmentation approach and linear filter. First we do superpixel segmentation on reference image, after this every pixel has a label determines which superpixel it belongs to. Then merging distance between superpixels in spatial space and color space, we apply linear filter on every superpixel. We evaluate the performance of our algorithm with some classic stereo datasets to show the promotion we obtained.

KW - Depth estimation

KW - Stereo matching

KW - linear filter

KW - superpixel segmentation

UR - http://www.scopus.com/inward/record.url?scp=84978880862&partnerID=8YFLogxK

U2 - 10.1109/ICCWAMTIP.2015.7493994

DO - 10.1109/ICCWAMTIP.2015.7493994

M3 - Conference contribution

AN - SCOPUS:84978880862

T3 - 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015

SP - 290

EP - 294

BT - 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015

Y2 - 18 December 2015 through 20 December 2015

ER -

Jin F, Li X. A dense depth estimation method using superpixels. In 2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015. Institute of Electrical and Electronics Engineers Inc. 2016. p. 290-294. 7493994. (2015 12th International Computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2015). doi: 10.1109/ICCWAMTIP.2015.7493994

A dense depth estimation method using superpixels

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this