Proceedings of the Korean Society of Broadcast Engineers Conference (한국방송∙미디어공학회:학술대회논문집)

The Korean Institute of Broadcast and Media Engineers (한국방송∙미디어공학회)
권호 표지

Comparison of Filter Selection for Compressed Sensing

압축센싱을 위한 필터선택 비교

  • Pham, Phuong Minh (School of Electronic and Electrical Engineering, Sungkyunkwan University) ;
  • Shim, Hiuk Jae (School of Electronic and Electrical Engineering, Sungkyunkwan University) ;
  • Jeon, Byeungwoo (School of Electronic and Electrical Engineering, Sungkyunkwan University)
  • Published : 2012.11.03
Download PDF
⟨ Previous Next ⟩

Abstract

Compressed Sensing (CS) has been developed for several years. Among many CS algorithms for image, the Block-based Compressed Sensing with Smoothed Projected Landweber (BCS-SPL) demonstrates its excellent performance in low-complexity and near-optimal reconstruction. Several noise filtering algorithms of image reconstruction have been introduced such as the Wiener or the median filters, etc. In general, each filter has its own advantages and disadvantages depending on specific coding scheme. In this paper, we show that reconstruction performance can be varied according to the choice of filter. When a sub-rate value is changed, different filter causes different effect as well. Concerning the sub-rate, an inner filter can be chosen to improve the reconstructed image quality.

Keywords