Most image denoising algorithms are designed and optimized for additive white Gaussian (AWG) noise. The noise characteristics in a real image could be very different from the AWG noise model, especially at high ISO speeds. In this project, we present multiresolution bilateral filter to reduce noise in digital images.

Related Publications:

A new image denoising method based on the bilateral filter, Ming Zhang and Bahadir K. Gunturk, IEEE Int. Conf. on Acoustics, Speech, Signal Processing (ICASSP), 2008. [pdf]

Sample Results:   

Left: Input image; Right: Result of our algorithm

In raster scan order: Input image, result of the wavelet based BLS-GSM algorithm [Portilla 03], result of the bilateral filter [Tomasi 98], result of our algorithm.