Edge Preservation of Enhanced Fuzzy Median Mean Filter Using Decision Based Median Filter

Image noise refers to random variations in the basic characteristics of image like brightness, intensity or color difference. These variations are not present in the image which is captured but may occur due to environmental conditions like sensor temperature or due to circuit of the scanner or other similar issues. Basically noise means unwanted signals in the image. Various filters have been designed for removal of almost all types of noise. It has been seen in most of the cases that as a result of high amount of filtering or repetitive filtering of image for the removal of noise, edges of images mostly get distorted or smeared out. It means that most of the filtering techniques lead to loss of fine edges of the images which needs to be preserved in order to enhance the quality of image. This paper has focused on to improve the enhanced fuzzy median mean filter so that fine edges get preserved in a better way. Experiments have been performed in MATLAB. Comparative analysis have been done on the basis of PSNR, MSE, BER and RMSE and it has shown that border correction applied on images improves the results of enhanced fuzzy median mean filter.