Fast Isotropic Median Filtering

TL;DR

This paper introduces a fast, versatile median filtering method capable of handling various kernel shapes, sizes, and image bit depths efficiently, overcoming previous limitations in the field.

Contribution

The authors present a novel median filtering algorithm that works efficiently with arbitrary kernel shapes, sizes, and bit depths, unlike prior methods limited to square kernels and specific data types.

Findings

Supports arbitrary convex kernel shapes including circular

Operates efficiently on images with various bit depths

Overcomes limitations of previous median filtering algorithms

Abstract

Median filtering is a cornerstone of computational image processing. It provides an effective means of image smoothing, with minimal blurring or softening of edges, invariance to monotonic transformations such as gamma adjustment, and robustness to noise and outliers. However, known algorithms have all suffered from practical limitations: the bit depth of the image data, the size of the filter kernel, or the kernel shape itself. Square-kernel implementations tend to produce streaky cross-hatching artifacts, and nearly all known efficient algorithms are in practice limited to square kernels. We present for the first time a method that overcomes all of these limitations. Our method operates efficiently on arbitrary bit-depth data, arbitrary kernel sizes, and arbitrary convex kernel shapes, including circular shapes.