Gradient-based synthesized multiple exposure time color HDR image

Annamária R. Varkónyi-Kóczy, András Rövid, Takeshi Hashimoto

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

High dynamic range (HDR) of illumination may cause serious distortion and information loss in the viewing and further processing of digital images. On the other hand, digital processing can often improve the visual quality of real-world photographs or views. Recently, HDR imaging techniques have become the focus of much research because of their high theoretical and practical importance. If the scene to be viewed or processed has HDR (dark and bright) illumination, these methods are able to help to detect details that a bright environment washes out, and they miss fewer details in a dark environment. By applying HDR techniques, the performance of different image processing and computer vision algorithms, information enhancement, and object and pattern recognition can also be improved. In this paper, a new tone reproduction algorithm is introduced, which may help to develop hard-to-view or nonviewable features and content of color images. The method applies a multiple exposure time image synthesization technique, where the red, green, and blue (RGB) color components of the pixels are separately handled. In the output, the corresponding (modified) color components are blended. As a result, a high-quality color HDR image is obtained, which contains the maximum level of details and color information.

Original languageEnglish
Pages (from-to)1779-1785
Number of pages7
JournalIEEE Transactions on Instrumentation and Measurement
Volume57
Issue number8
DOIs
Publication statusPublished - Aug 1 2008

    Fingerprint

Keywords

  • Color image reproduction
  • Digital image processing
  • High dynamic range (HDR) images
  • Information enhancement
  • Multiple exposure time image synthesization
  • Tone reproduction

ASJC Scopus subject areas

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this