An Efficient Noise Removal Algorithm Based on the Noise Density
Vishnu Praksh, K Shreedarshan · 2015
Gray scale and color images are affected by salt and pepper (impulse noise) which is encountered frequently in acquisition, transmission and processing of images. Proposed method achieves restoration of noisy image by usage of highly efficient filters which adapt based on the existing noise density in the image. The proposed algorithm involves two stages: noise density calculation of the corrupted image followed by noise detection and filtering. As noise density increases, the window size is increased which gives better results. The proposed algorithm replaces the pixels with values 0 and 255 with the median of the window considered if the window also includes pixel values other than 0 or 255. If the window considered contains pixels with values 0 and 255 only then they are replaced by the mean value of all elements present in the selected window. Since the proposed algorithm chooses the filter based on noise density, it works better than Median filter, Progressive Switched Median Filter (PSMF), Untrimmed Median Filter (UMF) and Adaptive Median Filter (AMF) considered individually. The proposed algorithm is tested against different grayscale and color images and it gives better Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF). The proposed algorithm is tested against different grayscale and color images and it gives better Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF).