Restoring chopped and nodded images by tight frames

Raymond CHAN, Lixin SHEN*, Zuowei SHEN

*Corresponding author for this work

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

Abstract

In infrared astronomy, the observed chopped and nodded image g can be viewed as the image obtained by passing the true image f through a highpass filter. Here we propose an iterative restoration algorithm by building up a tight frame wavelet system from a multiresolution analysis that has the highpass filter as one of the wavelet filters. To recover f, the low frequency information of f hidden in g is unfolded by a wavelet decomposition and reconstruction algorithm and combined with the given high frequency information in g. The main advantage of using our method to restore chopped and nodded images is that there are fewer artifacts as compared to the well-known projected Landweber method. Also the noise in the restored image is significantly reduced. Simulated and real images are tested to illustrate the efficiency of our method. Here, we briefly describe the main ideas of our recent paper1 and the details can be found there.

Original languageEnglish
Title of host publicationProceedings Volume 5205, Advanced Signal Processing Algorithms, Architectures, and Implementations XIII
EditorsFranklin T. LUK
PublisherSPIE
Pages310-319
Number of pages10
Volume5205
DOIs
Publication statusPublished - 24 Dec 2003
Externally publishedYes
EventOptical Science And Technology, SPIE's 48th Annual Meeting - San Diego, USA, United States
Duration: 6 Aug 20038 Aug 2003

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
PublisherSPIE
ISSN (Print)0277-786X

Conference

ConferenceOptical Science And Technology, SPIE's 48th Annual Meeting
Country/TerritoryUnited States
CitySan Diego, USA
Period6/08/038/08/03

Keywords

  • Chopped and nodded image
  • Tight frame

Fingerprint

Dive into the research topics of 'Restoring chopped and nodded images by tight frames'. Together they form a unique fingerprint.

Cite this