Single Image Reflection Removal with Patch Reflectance Prior

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Han, Dongshen, Yoon, Heechan, Kwon, Hyukmin, Kim, Hyun-Cheol, Choo, Hyon-Gon, Lee, Seungkyu, Zhang, Chaoning
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910226644467712
author Han, Dongshen
Yoon, Heechan
Kwon, Hyukmin
Kim, Hyun-Cheol
Choo, Hyon-Gon
Lee, Seungkyu
Zhang, Chaoning
author_facet Han, Dongshen
Yoon, Heechan
Kwon, Hyukmin
Kim, Hyun-Cheol
Choo, Hyon-Gon
Lee, Seungkyu
Zhang, Chaoning
contents Single Image Reflection Removal (SIRR) in real-world images is a challenging task due to diverse image degradations occurring on the glass surface during light transmission and reflection. Many existing methods rely on specific prior assumptions to resolve the problem. In this paper, we propose a general reflection intensity prior that captures the intensity of the reflection phenomenon and demonstrate its effectiveness. To learn the reflection intensity prior, we introduce the Reflection Prior Extraction Network (RPEN). By segmenting images into regional patches, RPEN learns non-uniform reflection prior in an image. We propose Prior-based Reflection Removal Network (PRRN) using a simple transformer U-Net architecture that adapts reflection prior fed from RPEN. Experimental results on real-world benchmarks demonstrate the effectiveness of our approach achieving state-of-the-art accuracy in SIRR.
format Preprint
id arxiv_https___arxiv_org_abs_2312_03798
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Single Image Reflection Removal with Patch Reflectance Prior
Han, Dongshen
Yoon, Heechan
Kwon, Hyukmin
Kim, Hyun-Cheol
Choo, Hyon-Gon
Lee, Seungkyu
Zhang, Chaoning
Computer Vision and Pattern Recognition
Single Image Reflection Removal (SIRR) in real-world images is a challenging task due to diverse image degradations occurring on the glass surface during light transmission and reflection. Many existing methods rely on specific prior assumptions to resolve the problem. In this paper, we propose a general reflection intensity prior that captures the intensity of the reflection phenomenon and demonstrate its effectiveness. To learn the reflection intensity prior, we introduce the Reflection Prior Extraction Network (RPEN). By segmenting images into regional patches, RPEN learns non-uniform reflection prior in an image. We propose Prior-based Reflection Removal Network (PRRN) using a simple transformer U-Net architecture that adapts reflection prior fed from RPEN. Experimental results on real-world benchmarks demonstrate the effectiveness of our approach achieving state-of-the-art accuracy in SIRR.
title Single Image Reflection Removal with Patch Reflectance Prior
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2312.03798