Implementation of Multiple Threshold MDBUT Filter for Salt-Pepper Noise Removal in Low Resolution Images

Abstract

It is important to enhance the quality of blurred monochrome images captured using low resolution monitoring or Surveillance cameras. The images, thus captured suffer from noise and degradation leading to loss or misconception of information. The images are usually available in fixed and variable scale. In this paper, we propose a technique for digital image restoration for variable scale low resolution images. Median filtering is applied to multiple threshold value of images. The result of filtering is noisy and noiseless pixels. The noiseless pixels are preserved first and restoration is applied only to the noisy pixels. Hardware implementation of Multiple Threshold MDBUT filter demonstrates that this method performs well than many existing decision-based methods. The analysis of FPGA implementation of the filter clearly shows that this algorithm is effective for blurred images captured by low resolution monitoring cameras corrupted by salt-pepper noise.