Peak noise removal by a facet model

摘要

A peak noise removal method based on local gray tone statistics from the facet model is introduced. Each pixel in an image is statistically tested to determine whether it belongs to the same gray tone intensity surface as its neighborhood pixels. If its gray tone is outside the 95 per cent confidence interval estimated from the neighborhood gray tones, it is judged as peak noise and its value is replaced by an average of the gray tone values of the neighborhood pixels. In order to estimate the local gray tone statistics, an assumption is made that the neighborhood region is described by a linear or quadratic facet surface model. It is also shown that this method can be successfully applied to scan line noise removal by using a one-dimensional (horizontal or vertical) neighborhood.