• DocumentCode
    2006788
  • Title

    Median filter on FPGAs

  • Author

    Gundam, Madhuri ; Charalampidis, Dimitrios

  • Author_Institution
    Dept. of Electr. Eng., Univ. of New Orleans, New Orleans, LA, USA
  • fYear
    2012
  • fDate
    11-13 March 2012
  • Firstpage
    83
  • Lastpage
    87
  • Abstract
    Median filters have been shown to be effective in removing impulsive noise from images. This paper proposes three different directional median filter implementations on FPGA. The techniques are capable of performing median filtering operations for four different directions simultaneously. These implementations are extensions to an existing cumulative histogram-based median filtering technique. Yet, directional processing was not addressed in previous work. Directional median filters process images along the direction of the filtering window, and have been shown to be more effective in removing impulsive noise compared to traditional median filters. All techniques aimed to decrease processing time, while reducing the resource utilization on hardware. Experimental results confirmed that all three techniques are synthesizable and implementable on Xilinx Virtex 5. There is a trade-off between resource utilization and processing time among the three techniques.
  • Keywords
    field programmable gate arrays; image denoising; median filters; FPGA; Xilinx Virtex 5; cumulative histogram-based median filtering operation; directional median filter; directional processing; image impulsive noise removal; image processing time; resource utilization; Clocks; Field programmable gate arrays; Filtering; Histograms; Noise; Registers; Resource management; Directional median filtering; FPGA; cumulative histogram;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    System Theory (SSST), 2012 44th Southeastern Symposium on
  • Conference_Location
    Jacksonville, FL
  • ISSN
    0094-2898
  • Print_ISBN
    978-1-4577-1492-4
  • Type

    conf

  • DOI
    10.1109/SSST.2012.6195129
  • Filename
    6195129