Skip to Main content Skip to Navigation
Conference papers

Reliable detection of hidden information based on a non-linear local model

Abstract : This paper investigates the reliable detection of information hidden in natural images. It is aimed to design a test with analytically predictable probabilities of error. To this end, the problem of hidden information detection is cast in the framework of hypothesis testing. The optimal test solving the decision problem of steganalysis requires image parameters which are not available in practice. To design a feasible test, a non-linear locally-adapted model of natural images is proposed. This model is linearized to allow an efficient and simple estimation of image parameters which leads to the design of an almost optimal test. Numerical results on a large number of natural images show the relevance of the theoretical findings.
Document type :
Conference papers
Complete list of metadatas

https://hal-utt.archives-ouvertes.fr/hal-02359577
Contributor : Jean-Baptiste Vu Van <>
Submitted on : Tuesday, November 12, 2019 - 2:33:28 PM
Last modification on : Thursday, June 25, 2020 - 12:32:04 PM

Identifiers

Collections

CNRS | ROSAS | UTT

Citation

Rémi Cogranne, Cathel Zitzmann, Lionel Fillatre, Igor Nikiforov, Florent Retraint, et al.. Reliable detection of hidden information based on a non-linear local model. 2011 IEEE Statistical Signal Processing Workshop (SSP), Jun 2011, Nice, France. pp.493-496, ⟨10.1109/SSP.2011.5967740⟩. ⟨hal-02359577⟩

Share

Metrics

Record views

46