An Inaccurate Median Filter Architecture for Salt and Pepper Noise Removal

Abstract

In this paper, an inaccurate median filter is proposed to eliminate the salt and pepper noise in the grayscale image. The corrupted pixel is detected and replaced using the Inaccurate Median Filter (IAMF). The IAMF excludes the corrupted pixel while computing the median value of the provided pixels in the 3×3 window. The input data bits are divided into different segments whose Partial Median (PM) are computed and concatenated to obtain the final median. The PM for each segment is computed without any comparators. The Peak Signal to Noise Ratio (PSNR) and Structural Similarity (SSIM) are used to analyze the image quality of the proposed architecture. The synthesis results are obtained using GPDK 90nm library at a clock frequency of 500MHz. These results shows that the proposed architecture consumes less area and produces good image quality compared to the existing architectures.