Removing fences from sweep motion videos using global 3D reconstruction and fence-aware light field rendering

Chanya Lueangwattana, Shohei Mori, Hideo Saito

Research output: Contribution to journalArticle

Abstract

Diminishing the appearance of a fence in an image is a challenging research area due to the characteristics of fences (thinness, lack of texture, etc.) and the need for occluded background restoration. In this paper, we describe a fence removal method for an image sequence captured by a user making a sweep motion, in which occluded background is potentially observed. To make use of geometric and appearance information such as consecutive images, we use two well-known approaches: structure from motion and light field rendering. Results using real image sequences show that our method can stably segment fences and preserve background details for various fence and background combinations. A new video without the fence, with frame coherence, can be successfully provided.

Original languageEnglish
JournalComputational Visual Media
DOIs
Publication statusAccepted/In press - 2018 Jan 1

Fingerprint

Fences
Restoration
Textures

Keywords

  • diminished reality (DR)
  • fence
  • light field rendering (LFR)
  • structure from motion (SfM)
  • video
  • video repair

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Computer Graphics and Computer-Aided Design
  • Artificial Intelligence

Cite this

Removing fences from sweep motion videos using global 3D reconstruction and fence-aware light field rendering. / Lueangwattana, Chanya; Mori, Shohei; Saito, Hideo.

In: Computational Visual Media, 01.01.2018.

Research output: Contribution to journalArticle

@article{ec0a8edafc7f4af0ac2b54b622a526df,
title = "Removing fences from sweep motion videos using global 3D reconstruction and fence-aware light field rendering",
abstract = "Diminishing the appearance of a fence in an image is a challenging research area due to the characteristics of fences (thinness, lack of texture, etc.) and the need for occluded background restoration. In this paper, we describe a fence removal method for an image sequence captured by a user making a sweep motion, in which occluded background is potentially observed. To make use of geometric and appearance information such as consecutive images, we use two well-known approaches: structure from motion and light field rendering. Results using real image sequences show that our method can stably segment fences and preserve background details for various fence and background combinations. A new video without the fence, with frame coherence, can be successfully provided.",
keywords = "diminished reality (DR), fence, light field rendering (LFR), structure from motion (SfM), video, video repair",
author = "Chanya Lueangwattana and Shohei Mori and Hideo Saito",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/s41095-018-0126-8",
language = "English",
journal = "Computational Visual Media",
issn = "2096-0433",
publisher = "Tsinghua University Press",

}

TY - JOUR

T1 - Removing fences from sweep motion videos using global 3D reconstruction and fence-aware light field rendering

AU - Lueangwattana, Chanya

AU - Mori, Shohei

AU - Saito, Hideo

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Diminishing the appearance of a fence in an image is a challenging research area due to the characteristics of fences (thinness, lack of texture, etc.) and the need for occluded background restoration. In this paper, we describe a fence removal method for an image sequence captured by a user making a sweep motion, in which occluded background is potentially observed. To make use of geometric and appearance information such as consecutive images, we use two well-known approaches: structure from motion and light field rendering. Results using real image sequences show that our method can stably segment fences and preserve background details for various fence and background combinations. A new video without the fence, with frame coherence, can be successfully provided.

AB - Diminishing the appearance of a fence in an image is a challenging research area due to the characteristics of fences (thinness, lack of texture, etc.) and the need for occluded background restoration. In this paper, we describe a fence removal method for an image sequence captured by a user making a sweep motion, in which occluded background is potentially observed. To make use of geometric and appearance information such as consecutive images, we use two well-known approaches: structure from motion and light field rendering. Results using real image sequences show that our method can stably segment fences and preserve background details for various fence and background combinations. A new video without the fence, with frame coherence, can be successfully provided.

KW - diminished reality (DR)

KW - fence

KW - light field rendering (LFR)

KW - structure from motion (SfM)

KW - video

KW - video repair

UR - http://www.scopus.com/inward/record.url?scp=85057309626&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057309626&partnerID=8YFLogxK

U2 - 10.1007/s41095-018-0126-8

DO - 10.1007/s41095-018-0126-8

M3 - Article

AN - SCOPUS:85057309626

JO - Computational Visual Media

JF - Computational Visual Media

SN - 2096-0433

ER -