Multilevel median filter using Harmonic mean ...

Multilevel median filter using Harmonic mean filter for Gaussian Noise

Author Name : Narendra Kumar, Krishna Kumar

ABSTRACT

Many applications in today's digital age are built on photos, and so the ensuing successes must reflect on their quality. Because pictures are not always of high quality due to numerous sources of noise (natural noise, sensor faults, transmission issues, etc.), it is critical to reducing noise automatically and effectively. Digital image processing techniques are currently being employed to tackle a wide range of issues. Although these issues are frequently unconnected, they all necessitate strategies for improving graphical resources for better interpretation and analysis. In this paper, we proposed a multilevel filter for Gaussian noise. In the first level, we are using a wiener filter for Gaussian noise. The output of the Wiener filter is used as an input of the geometric mean filter and the output of the geometric mean filter is used as an input of the median filter. We are using different parameters to show the comparison with different filters, like a median filter, harmonic mean filter, and geometric filter.

Keywords: Median filter, Harmonic mean filter, geometric mean filter, multilevel filter, Gaussian noise, wiener filter, peak to noise ratio.