Transformation of image-signal-dependent noise into image-signal-independent noise.

A point transformation, the normalizing transform, is presented, which, when applied to a measured noisy image, renders its noise signal independent. The transform is suitable for arbitary noise-to-signal dependence. We demonstrate its applicability and its limitations by using, as an example, noisy signals that belong to a family of gamma-distributed random variables with power-law variance-to-mean dependence. Its normalizing and variance-stabilizing properties are studied.