Scale invariance and noise in natural images
Authors:
- 1 September 2009
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15505499,p. 2209-2216
Abstract
Natural images are known to have scale invariant statistics. While some eariler studies have reported the kurtosis of marginal bandpass filter response distributions to be constant throughout scales, other studies have reported that the kurtosis values are lower for high frequency filters than for lower frequency ones. In this work we propose a resolution for this discrepancy and suggest that this change in kurtosis values is due to noise present in the image. We suggest that this effect is consistent with a clean, natural image corrupted by white noise. We propose a model for this effect, and use it to estimate noise standard deviation in corrupted natural images. In particular, our results suggest that classical benchmark images used in low-level vision are actually noisy and can be cleaned up. Our results on noise estimation on two sets of 50 and a 100 natural images are significantly better than the state-of-the-art.Keywords
This publication has 15 references indexed in Scilit:
- What makes a good model of natural images?2007
- On the Spatial Statistics of Optical FlowInternational Journal of Computer Vision, 2007
- Noise estimation for video processing based on spatio-temporal gradientsIEEE Signal Processing Letters, 2006
- Training Methods for Image Noise Level Estimation on Wavelet ComponentsEURASIP Journal on Advances in Signal Processing, 2004
- Comparison of graph cuts with belief propagation for stereo, using identical MRF parameters2003
- On Advances in Statistical Modeling of Natural ImagesJournal of Mathematical Imaging and Vision, 2003
- A mathematical analysis of the DCT coefficient distributions for imagesIEEE Transactions on Image Processing, 2000
- Estimation of image noise varianceIEE Proceedings - Vision, Image, and Signal Processing, 1999
- Independent component filters of natural images compared with simple cells in primary visual cortexProceedings Of The Royal Society B-Biological Sciences, 1998
- Origins of scaling in natural imagesVision Research, 1997