Reflection Removal from Multiple Images
- 주제(키워드) reflection removal , layer separation , motion field estimation , illumination chage , brightness constancy constraint , gradient constancy constraint
- 발행기관 서강대학교 일반대학원
- 지도교수 박래홍
- 발행년도 2018
- 학위수여년월 2018. 2
- 학위명 석사
- 학과 및 전공 일반대학원 전자공학과
- 실제URI http://www.dcollection.net/handler/sogang/000000063064
- 본문언어 영어
- 저작권 서강대학교 논문은 저작권보호를 받습니다.
초록/요약
This thesis proposes a method to separate the background and reflection layers from multiple images. The proposed method uses the motion parallax, which is caused by the difference of the distance between the camera and each layer. Each motion field is estimated using the brightness constancy and the gradient constancy that is robust to brightness change. To separate the background and reflection layers, the two data terms are combined with the sparsity constraint and independence of the gradients of two layers. Experiments show that the reconstructed images using both the brightness constancy constraint and the gradient constancy constraint have smaller errors than those with the brightness constancy constraint only. Also, the proposed method and two state-of-the-art algorithms are compared and evaluated qualitatively and quantitatively on real-world images.
more