DOI QR코드

DOI QR Code

Research on Infrastructure technology of Stereoscopic Object Expression Utilizing the Grabcut algorithm

Grabcut 알고리즘을 활용한 Stereoscopic 객체표현 기반 기술 연구

  • Lee, Min ho (Dept. of Information Communication & Media Engineering, Graduate School of Nano It Design Fusion, Seoul National University of Science & Technology) ;
  • Choi, Jin yeong (Dept. of Media IT Engineering, Seoul National Univ. of Science and Tech.) ;
  • Lee, Jong hyeok (JJ Company Co., Ltd.) ;
  • Cha, Jae sang (Dept. of Electronics & IT Media Eng., Seoul National University of Science and Technology)
  • 이민호 (서울과학기술대학교 나노IT디자인융합대학원) ;
  • 최진영 (서울과학기술대학교 일반대학원 미디어IT공학과) ;
  • 이종혁 (JJ컴퍼니) ;
  • 차재상 (서울과학기술대학교 전자IT미디어공학과)
  • Received : 2018.10.01
  • Accepted : 2018.10.21
  • Published : 2018.10.31

Abstract

Recently, stereoscopic technology has become a potential for blue ocean as a new growth power industry, and interest in it has been steadily increasing with the development of virtual and augmented reality technologies. Various methods such as binocular parallax and polarized glasses have been developed and used for stereoscopic image expression, but they have limitations such as eye damage, headache, crosstalk and resolution degradation. In this paper, we present a new method of stereoscopic image representation that can overcome the limitations and verify its applicability through basic experiments for object extraction and real - time image representation.

최근 입체영상 기술은 신 성장 동력 산업으로서 잠재성 있는 블루오션으로 자리매김하고 있으며, 가상 및 증강현실 기술의 발전과 함께 관심이 꾸준히 증가하고 있다. 입체영상 표현방식은 양안시차, 편광안경 등 다양한 형태의 기술이 개발되어 활용되고 있으나, 눈의 피해나, 두통, 혼선, 해상도 저하 등 한계점을 지닌다. 본 논문에서는 기존의 한계점을 극복하고자 중첩된 투명 디스플레이를 활용한 새로운 방식의 입체영상 표현 기법을 제시하였으며, Grabcut 알고리즘을 활용한 전경 추출 실험과 아크릴과 스마트 디바이스를 활용한 새로운 방식의 입체감 표현 실험을 통해 적용 가능성을 확인하였다. 또한, 이미지 프로세싱 소요 시간 등 분석을 통해 실시간 스트리밍 적용도 가능함을 검증하였다.

Keywords

References

  1. Kim H.-J., Gwun T.-G., Joo Y.-I. and Seo D.-H. (2015), "A study on implementation of background subtraction algorithm using LMS algorithm and performance comparative analysis," Journal of the Korean Society of Marine Engineering, vol. 39, no. 1 pp.94-98. https://doi.org/10.5916/jkosme.2015.39.1.94
  2. Kwon D.-J. (2010), "A Study on Modified Watershed Algorithm for Object Extraction," Proceedings of KIIS Fall Conference, vol. 20, no. 2.
  3. Park S.-H. (2015), Features of 3D stereoscopic technology and 3D display technology, CCTV news.
  4. Rother C., Blake A. and Kolmogorov V. (2004), "Grabcut; Interactive foreground extraction using iterated graphcuts," ACM Transactions on Graphics (SIGGRAPH), vol. 23, pp.309-314. https://doi.org/10.1145/1015706.1015720
  5. Yun C. O., Yun T. S. and Lee D. H. (2006), "Interactive 3D Stereoscopic Image Editing System using Image-based modeling," KSIAM IT series, vol. 10, no. 2, pp.53-66.
  6. http://www.cs.ru.ac.za/research/g02m1682/, 2018.