Skip to main navigation Skip to search Skip to main content

Extended difference function and its application in image motion estimation

Research output: Contribution to journalArticlepeer-review

Abstract

Since difference function (DF) based motion estimation methods with translational motion model without magnitude changing limits performance in image estimation with varying illumination, a novel method of translation estimation on the basis of extended difference function in Radon domain was developed after the analysis of difference function, the definition of extended difference function (EDF), and proof of linear relation between zeros and translation vector, which is irrelevant to illumination. Experiments were conducted on this method for different set of image with illumination changes. The results demonstrate that this method is invariance to illumination changes with robustness against noise at low computation cost.

Original languageEnglish
Pages (from-to)897-901
Number of pages5
JournalJilin Daxue Xuebao (Gongxueban)/Journal of Jilin University (Engineering and Technology Edition)
Volume38
Issue number4
StatePublished - Jul 2008

Keywords

  • Extended difference function
  • Information processing
  • Motion estimation
  • Radon transform

Fingerprint

Dive into the research topics of 'Extended difference function and its application in image motion estimation'. Together they form a unique fingerprint.

Cite this