Morphological statistics of natural images and consequences

Research output: Contribution to journalConference articlepeer-review

Abstract

We present statistics on natural images dealing with homogeneous and connected regions, defined as connected components of differences of level sets. We show that their area is distributed according to a power law. From these experimental results, we show precisely how the modeling of images as functions of bounded variation, widely used in image restoration methods, is incompletely adapted to natural images. For this purpose we use a model of convolution and sampling for the formation of images. We also use our experimental results to confirm the scale invariance of natural images.

Original languageEnglish
Pages (from-to)208-214
Number of pages7
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume4119
DOIs
Publication statusPublished - 1 Dec 2000
Externally publishedYes
EventWavelet Applications in Signal and Image Processing VIII - San Diego, CA, USA
Duration: 31 Jul 20004 Aug 2000

Fingerprint

Dive into the research topics of 'Morphological statistics of natural images and consequences'. Together they form a unique fingerprint.

Cite this