Improved Image Expansion for Preserving Sharpness Without Jagged Edges

Damian Tohl, Jim Siu Jimmy Li

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

1 Citation (Scopus)

Abstract

Image expansion is an interpolation problem which has been known to cause blurred and jagged edges. A number of methods have been developed to preserve sharp edges while preventing jaggedness, but the results are often unsatisfactory. In this paper, we propose an image expansion technique which can preserve sharpness of an image without jagged edges, while maintaining smoothness in other parts of the interpolated image. The input image is first expanded by the nearest neighbour interpolation method in order to preserve the original edges. It is also interpolated by another interpolation algorithm with a different degree of edge preservation. The absolute difference between the two outputs produces a residual image of edges with a magnitude which increases with the sharpness of an original edge. This edge residual image is then used to adaptively adjust the gamma value of a novel modified gamma contrast enhancement algorithm so that the edges can regain their original sharpness. A halo effect suppressor has also been introduced to suppress halo effects by avoiding over-sharpening of the edges. It has been shown that our algorithm can produce sharp images without jagged edges and outperform the other benchmarking algorithms visually and quantitatively.

Original languageEnglish
Title of host publicationProceedings of TENCON 2018 - 2018 IEEE Region 10 Conference
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1263-1268
Number of pages6
ISBN (Electronic)9781538654576
ISBN (Print)9781538654583
DOIs
Publication statusPublished - 22 Feb 2019
Event2018 IEEE Region 10 Conference, TENCON 2018 - Jeju, Korea, Republic of
Duration: 28 Oct 201831 Oct 2018

Publication series

NameIEEE Region 10 Annual International Conference, Proceedings/TENCON
Volume2018-October
ISSN (Print)2159-3442
ISSN (Electronic)2159-3450

Conference

Conference2018 IEEE Region 10 Conference, TENCON 2018
CountryKorea, Republic of
CityJeju
Period28/10/1831/10/18

Keywords

  • Image edge detection
  • Interpolation
  • Benchmark testing
  • IEEE Regions
  • Conferences
  • Microsoft Windows
  • Australia
  • image enhancement
  • Edge preservation

Fingerprint Dive into the research topics of 'Improved Image Expansion for Preserving Sharpness Without Jagged Edges'. Together they form a unique fingerprint.

Cite this