Digital image decomposition and contrast enhancement using high-dimensional model representation


Tunga B., KOCANAOĞULLARI A.

SIGNAL IMAGE AND VIDEO PROCESSING, vol.12, no.2, pp.299-306, 2018 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 12 Issue: 2
  • Publication Date: 2018
  • Doi Number: 10.1007/s11760-017-1158-8
  • Journal Name: SIGNAL IMAGE AND VIDEO PROCESSING
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.299-306
  • Istanbul Technical University Affiliated: Yes

Abstract

Contrast is the difference in brightness and color that makes an object distinguishable. Contrast enhancement (CE) is a technique used to improve the visual quality of an image for human recognition. This study proposes a new methodology called high-dimensional model representation (HDMR) for enhancing contrast in digital images. The novelty of HDMR is that the method first decomposes the image into its dimensions, then represents the image using the superposition of decomposed components and finally enhances contrast in the image by adding certain HDMR components to the representation. HDMR has high performance as a CE technique in both grayscale and color images when compared with some state-of-the-art methods.