A new framework of removing salt and pepper impulse noise for the noisy image including many noise-free white and black pixels
Song Li, Caizhu Wang, Yeqiu Li, Ling Wang, Shiro Sakata, Hiroo Sekiya, and Shingo Kuroiwa
電気学会論文誌C, vol.129, no.7, pp.1427-1434, July, 2009. [pdf document]

<Abstract>

In this paper, we propose a new framework of removing salt and pepper impulse noise. In our proposed framework, the most important point is that the number of noise-free white and black pixels in a noisy image can be determined by using the noise rates estimated by Fuzzy Impulse Noise Detection and Reduction Method (FINDRM) and Efficient Detail-Preserving Approach (EDPA). For the noisy image includes many noise-free white and black pixels, the detected noisy pixel from the FINDRM is re-checked by using the alpha-trimmed mean. Finally, the impulse noise filtering phase of the FINDRM is used to restore the image. Simulation results show that for the noisy image including many noise-free white and black pixels, the proposed framework can decrease the False Hit Rate (FHR) efficiently compared with the FINDRM. Therefore, the proposed framework can be used more widely than the FINDRM.