On scale-invariant properties in natural images and their simulations
This work provides incremental insights into statistical properties of natural images, potentially benefiting image processing and computational vision researchers.
The authors analyzed natural images from the van Hateren dataset and computationally demonstrated scale-invariant properties, observing a power-law decaying power spectrum. They proposed a dynamic model that qualitatively reproduced the observed slope and numerically confirmed scale-invariance for two source types.
We study samples of natural images for which a set of statistical characteristics is computed and scale-invariant properties of samples are demonstrated computationally. Computations of the power spectrum are carried out and a power-law decaying power spectrum is observed on samples taken from van Hateren images of natural scenes. We propose a dynamic model to reproduce the observed slope in the power spectrum qualitatively. For two types of sources for this model the behaviour of power spectrum is investigated and scale-invariance confirmed numerically. We then discuss potential applications of scale-invariant properties of natural images.