.png){kind=icon}
Return
Single image deraining using local rain distribution map
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-023-16972-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Existing image deraining methods work on the entire domain of images causing the deraining results to contain rain streaks or over-smoothing. To alleviate these problems, we propose a local pixel rain streak removal method based on the observation that rain streaks are sparsely distributed in images when rain levels are slight and middle. We first propose a discriminative anisotropic gradient prior to efficiently extract rain streaks from the background and generate a rain streak distribution (RSD) mask. Then, we recover the information in the rain-distorted regions with the RSD mask using a simple multi-layer image inpainting method. Experiments on both synthesized and real-world images demonstrate that our method outperforms the state-of-the-art methods in terms of rain streak detection, removal, and information restoration.
Keywords:
Single image deraining
Rain streak detection
Local rain streak removal
Sparse distribution of rain streak
Journal
IF:
3
Papers:
2.0W
Citations:
3.2W
Organization
Cited Papers
.png){kind=icon}
Modelling of grain boundary misorientation spectrum in polycrystals with crystallographic texture
PREAI
ShareSave
ShareSaveShareSaveShareSave
View more