Image Processing Case Study: Noise Reduction With the Kuwahara Filter

In modern society, digital images play an important role in the presentation and interpretation of visual information. Inherent to capturing these images is the introduction of undesirable noise due to the imaging system or other environmental factors. This noise presents major challenges for analysis and feature extraction, especially with the increased use of machine learning and artificial intelligence applied to images. Noise reduction techniques take many forms, ranging from simple mean filters to the removal of high-frequency information within the image. In this case study, I present the Kuwahara filter, a nonlinear filter that uses statistical information about surrounding pixels with the goal of preserving edge information while smoothing interior regions.