Gökrem, LeventErkan, Uğur2025-01-212025-01-2120181308-5514https://dergipark.org.tr/tr/download/article-file/507177https://dergipark.org.tr/tr/pub/umagd/issue/38203/376895https://doi.org/10.29137/umagd.376895https://hdl.handle.net/20.500.12587/20241In this study,a new method has been developed on noise removal, one of the most important partsin image processing. In particular, a new noise removal filter has beendeveloped to removes noise from images the mix of µ and ? values. In thismethod, a new approach is proposed to remove noise when ? value increases while? value is constant. The filter has particularly proven to be more successfulon all of µ values. PSNR have been used to compare the results of the study.The newly developed method has been compared with the median, wiener2, Bayesianshrink, bilateral, median+bilateral, BM3D, KSVD methods. For example, whenµ-0.10 and ?-0.01 added to Lena image, other algorithms’ PSNR results are19.23-19.35, 18.90, 18.73, 19.60, 19.81, 19.70 while they are 27.06 in our newmethod.eninfo:eu-repo/semantics/openAccessGaussian NoiseImage DenoisingNoise RemovalImage EnhancementArticle12-13514210.29137/umagd.376895376895