A Rank Based Edge Enhancement Filter

Edge enhancement is an operation commonly employed in image processing. Usually the edge slope is steepened by the applying a linear filter that amplifies the high spatial frequency content of an image. This often results in the amplification of noise and in ringing or over-enhancement of edges that are already sharp. The rank filter is a nonlinear filter. The output value for a pixel is obtained by ranking the pixel values within a neighbourhood in the input image and using the value from a specified rank position. Near an edge, low and high rank values represent of the dark and light sides of the edge respectively. Edges may be enhanced by specifying two rank positions and assigning the output value to that of the rank position which is closer in intensity to the centre pixel. Strongest edge enhancement results from using extreme rank positions while noise sensitivity may be reduced by using rank positions close to the median. Excessive ringing is avoided since no new intensity values are generated. Successive applications of a rank-based edge enhancement filter converge to a stable root image.