Skip to main navigation Skip to search Skip to main content

Removing rain and snow in a single image using guided filter

  • School of Computer Science and Technology, Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Dynamic weather conditions, such as rain and snow, degrade the performance of vision algorithms, which is served for video surveillance and analysis tasks. Many methods have been proposed to reduce the effects of raindrops and snow flakes that are regarded as noises in each frame of a certain video. However, the removing of rain-drop or snow-flake is quite challenging since each of them affects only on a very small region of an image, thus, leading us a rub to determine which region should be smoothed and which should not. In this paper, a guided filter, by which no pixel-based statistical information for detecting rain and snow are needed, is introduced in the removal stage. The results show that it may achieve better performance. In addition, the degradation-restoring algorithm is suitable for both single images and videos.

Original languageEnglish
Title of host publicationCSAE 2012 - Proceedings, 2012 IEEE International Conference on Computer Science and Automation Engineering
Pages304-307
Number of pages4
DOIs
StatePublished - 2012
Externally publishedYes
Event2012 IEEE International Conference on Computer Science and Automation Engineering, CSAE 2012 - Zhangjiajie, China
Duration: 25 May 201227 May 2012

Publication series

NameCSAE 2012 - Proceedings, 2012 IEEE International Conference on Computer Science and Automation Engineering
Volume2

Conference

Conference2012 IEEE International Conference on Computer Science and Automation Engineering, CSAE 2012
Country/TerritoryChina
CityZhangjiajie
Period25/05/1227/05/12

Keywords

  • bad weather
  • outdoor vision
  • rain removal
  • snow removal

Fingerprint

Dive into the research topics of 'Removing rain and snow in a single image using guided filter'. Together they form a unique fingerprint.

Cite this