Hierarchical haze removal using dark channel prior

Jin Hwan Kim, Chang Su Kim

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The haze removal algorithm using dark channel prior, which was proposed by He et al., is an efficient algorithm and presents impressive results. But its high memory and computational requirements limit its applications. In this paper, we propose a method to improve the memory usage and calculation speed. We notice that the matting process accounts for most calculation time, so we replace the matting process with a fast bilateral filtering scheme. Using the bilateral filter, we can reduce the memory usage, but its computational complexity is still high. To reduce the computational complexity as well, we adapt a hierarchical structure for the bilateral filtering. Experimental results show that the proposed algorithm can remove haze in a picture effectively, while requiring much less computations than the He et al.'s method.

Original languageEnglish
Pages (from-to)457-464
Number of pages8
JournalTransactions of the Korean Institute of Electrical Engineers
Volume59
Issue number2
Publication statusPublished - 2010 Feb
Externally publishedYes

Keywords

  • Haze removal
  • Image enhancement

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Hierarchical haze removal using dark channel prior'. Together they form a unique fingerprint.

  • Cite this